My computers were connected to an UPS surge projector. Various kitchen machinery and extension sockets that was not doesn't work anymore. The neighbor's TV blew.
There was a bright white flash inside of my living room, just under an open window and directly above the my UPS surge protector and then all my devices turned off including the surge projector. The PSU in one of the computers is gone, but not the motherboard and rest.
I was sitting on the concrete floor in the middle of the room watching it happen.
I live in Thailand in the middle of a big flat coconut groove with only two houses. Except for the palm trees, my house is the tallest structure in a 200 x 200 meters (300 feet)+ area and we have frequent huge thunder storms, but this one was in a category on its own. It made my concrete walls tremble. It was so intense. Spectacular show by nature.
Only two months ago, I bought a socket outlet tester. It reported that my sockets was not grounded. I got an electricians out to fix that. I suspect this is the reason why the lightning chooses me. We have frequent big thunderstorms and haven't had this issue before, but now there is a ground connection.
I don't know if there is a proper grounding rod. There is a cable going from the house into the ground, but I don't know what it is attached too. The rest of the electric installation is amateur's business so could be connected to nothing. The house is a long-term rental and the electric installation came with it, but I can change it regardless of it's is a rental.
The electricity doesn't even come directly from the power company's cables. It goes in to the owners house, where the meter is. From there they have pulled a 200 meter cable into ours. There is a fuse box in my house.
Any suggestions on what to do next? A big metal stick somewhere in the garden?"},"title":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["compute"],"value":"Ask HN: My house just got hit by lightning and it fried the PSU in my computer"}},"_tags":["story","author_TowerTall","story_31270349","ask_hn"],"author":"TowerTall","children":[31270509,31270564,31270586,31270601,31270615,31270632,31270633,31270638,31270653,31270679,31270685,31270691,31270694,31270754,31270864,31270987,31271156,31271402,31271552,31271990,31272071,31274070,31274338,31279488,31280930],"created_at":"2022-05-05T06:47:29Z","created_at_i":1651733249,"num_comments":79,"objectID":"31270349","points":74,"story_id":31270349,"story_text":"If everything is wired correctly in my house, should all equipment attached to my UPS have survived this with no impact or is a direct lightning strike so powerful that even high-quality home-grade UPS must give up?
My computers were connected to an UPS surge projector. Various kitchen machinery and extension sockets that was not doesn't work anymore. The neighbor's TV blew.
There was a bright white flash inside of my living room, just under an open window and directly above the my UPS surge protector and then all my devices turned off including the surge projector. The PSU in one of the computers is gone, but not the motherboard and rest.
I was sitting on the concrete floor in the middle of the room watching it happen.
I live in Thailand in the middle of a big flat coconut groove with only two houses. Except for the palm trees, my house is the tallest structure in a 200 x 200 meters (300 feet)+ area and we have frequent huge thunder storms, but this one was in a category on its own. It made my concrete walls tremble. It was so intense. Spectacular show by nature.
Only two months ago, I bought a socket outlet tester. It reported that my sockets was not grounded. I got an electricians out to fix that. I suspect this is the reason why the lightning chooses me. We have frequent big thunderstorms and haven't had this issue before, but now there is a ground connection.
I don't know if there is a proper grounding rod. There is a cable going from the house into the ground, but I don't know what it is attached too. The rest of the electric installation is amateur's business so could be connected to nothing. The house is a long-term rental and the electric installation came with it, but I can change it regardless of it's is a rental.
The electricity doesn't even come directly from the power company's cables. It goes in to the owners house, where the meter is. From there they have pulled a 200 meter cable into ours. There is a fuse box in my house.
Any suggestions on what to do next? A big metal stick somewhere in the garden?","title":"Ask HN: My house just got hit by lightning and it fried the PSU in my computer","updated_at":"2024-09-20T11:06:31Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"kylequest"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["thunder","compute"],"value":"123456\npassword\nphpbb\nqwerty\n12345\njesus\n12345678\n1234\nabc123\nletmein\ntest\nlove\n123\npassword1\nhello\nmonkey\ndragon\ntrustno1\n111111\niloveyou\n1234567\nshadow\n123456789\nchrist\nsunshine\nmaster\ncomputer\nprincess\ntigger\nfootball\nangel\njesus1\n123123\nwhatever\nfreedom\nkiller\nasdf\nsoccer\nsuperman\nmichael\ncheese\ninternet\njoshua\nfuckyou\nblessed\nbaseball\nstarwars\n000000\npurple\njordan\nfaith\nsummer\nashley\nbuster\nheaven\npepper\n7777777\nhunter\nlovely\nandrew\nthomas\nangels\ncharlie\ndaniel\n1111\njennifer\nsingle\nhannah\nqazwsx\nhappy\nmatrix\npass\naaaaaa\n654321\namanda\nnothing\nginger\nmother\nsnoopy\njessica\nwelcome\npokemon\niloveyou1\n11111\nmustang\nhelpme\njustin\njasmine\norange\ntesting\napple\nmichelle\npeace\nsecret\n1\ngrace\nwilliam\niloveyou2\nnicole\n666666\nmuffin\ngateway\nfuckyou1\nasshole\nhahaha\npoop\nblessing\nblahblah\nmyspace1\nmatthew\ncanada\nsilver\nrobert\nforever\nasdfgh\nrachel\nrainbow\nguitar\npeanut\nbatman\ncookie\nbailey\nsoccer1\nmickey\nbiteme\nhello1\neminem\ndakota\nsamantha\ncompaq\ndiamond\ntaylor\nforum\njohn316\nrichard\nblink182\npeaches\ncool\nflower\nscooter\nbanana\njames\nasdfasdf\nvictory\nlondon\n123qwe\n123321\nstartrek\ngeorge\nwinner\nmaggie\ntrinity\nonline\n123abc\nchicken\njunior\nchris\npassw0rd\naustin\nsparky\nadmin\nmerlin\ngoogle\nfriends\nhope\nshalom\nnintendo\nlooking\nharley\nsmokey\n7777\njoseph\nlucky\ndigital\na\nthunder\nspirit\nbandit\nenter\nanthony\ncorvette\nhockey\npower\nbenjamin\niloveyou!\n1q2w3e\nviper\ngenesis\nknight\nqwerty1\ncreative\nfoobar\nadidas\nrotimi\nslayer\nwisdom\npraise\nzxcvbnm\nsamuel\nmike\ndallas\ngreen\ntesttest\nmaverick\nonelove\ndavid\nmylove\nchurch\nfriend\ngod\ndestiny\nnone\nmicrosoft\n222222\nbubbles\n11111111\ncocacola\njordan23\nilovegod\nfootball1\nloving\nnathan\nemmanuel\nscooby\nfuckoff\nsammy\nmaxwell\njason\njohn\n1q2w3e4r\nbaby\nred123\nblabla\nprince\nqwert\nchelsea\n55555\nangel1\nhardcore\ndexter\nsaved\n112233\nhallo\njasper\ndanielle\nkitten\ncassie\nstella\nprayer\nhotdog\nwindows\nmustdie\ngates\nbillgates\nghbdtn\ngfhjkm\n1234567890"},"title":{"matchLevel":"none","matchedWords":[],"value":"Pony password stealer own password list"},"url":{"matchLevel":"none","matchedWords":[],"value":""}},"_tags":["story","author_kylequest","story_6857010","ask_hn"],"author":"kylequest","created_at":"2013-12-05T20:32:27Z","created_at_i":1386275547,"num_comments":0,"objectID":"6857010","points":2,"story_id":6857010,"story_text":"123456\npassword\nphpbb\nqwerty\n12345\njesus\n12345678\n1234\nabc123\nletmein\ntest\nlove\n123\npassword1\nhello\nmonkey\ndragon\ntrustno1\n111111\niloveyou\n1234567\nshadow\n123456789\nchrist\nsunshine\nmaster\ncomputer\nprincess\ntigger\nfootball\nangel\njesus1\n123123\nwhatever\nfreedom\nkiller\nasdf\nsoccer\nsuperman\nmichael\ncheese\ninternet\njoshua\nfuckyou\nblessed\nbaseball\nstarwars\n000000\npurple\njordan\nfaith\nsummer\nashley\nbuster\nheaven\npepper\n7777777\nhunter\nlovely\nandrew\nthomas\nangels\ncharlie\ndaniel\n1111\njennifer\nsingle\nhannah\nqazwsx\nhappy\nmatrix\npass\naaaaaa\n654321\namanda\nnothing\nginger\nmother\nsnoopy\njessica\nwelcome\npokemon\niloveyou1\n11111\nmustang\nhelpme\njustin\njasmine\norange\ntesting\napple\nmichelle\npeace\nsecret\n1\ngrace\nwilliam\niloveyou2\nnicole\n666666\nmuffin\ngateway\nfuckyou1\nasshole\nhahaha\npoop\nblessing\nblahblah\nmyspace1\nmatthew\ncanada\nsilver\nrobert\nforever\nasdfgh\nrachel\nrainbow\nguitar\npeanut\nbatman\ncookie\nbailey\nsoccer1\nmickey\nbiteme\nhello1\neminem\ndakota\nsamantha\ncompaq\ndiamond\ntaylor\nforum\njohn316\nrichard\nblink182\npeaches\ncool\nflower\nscooter\nbanana\njames\nasdfasdf\nvictory\nlondon\n123qwe\n123321\nstartrek\ngeorge\nwinner\nmaggie\ntrinity\nonline\n123abc\nchicken\njunior\nchris\npassw0rd\naustin\nsparky\nadmin\nmerlin\ngoogle\nfriends\nhope\nshalom\nnintendo\nlooking\nharley\nsmokey\n7777\njoseph\nlucky\ndigital\na\nthunder\nspirit\nbandit\nenter\nanthony\ncorvette\nhockey\npower\nbenjamin\niloveyou!\n1q2w3e\nviper\ngenesis\nknight\nqwerty1\ncreative\nfoobar\nadidas\nrotimi\nslayer\nwisdom\npraise\nzxcvbnm\nsamuel\nmike\ndallas\ngreen\ntesttest\nmaverick\nonelove\ndavid\nmylove\nchurch\nfriend\ngod\ndestiny\nnone\nmicrosoft\n222222\nbubbles\n11111111\ncocacola\njordan23\nilovegod\nfootball1\nloving\nnathan\nemmanuel\nscooby\nfuckoff\nsammy\nmaxwell\njason\njohn\n1q2w3e4r\nbaby\nred123\nblabla\nprince\nqwert\nchelsea\n55555\nangel1\nhardcore\ndexter\nsaved\n112233\nhallo\njasper\ndanielle\nkitten\ncassie\nstella\nprayer\nhotdog\nwindows\nmustdie\ngates\nbillgates\nghbdtn\ngfhjkm\n1234567890","title":"Pony password stealer own password list","updated_at":"2024-09-19T20:14:13Z","url":""},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"pritambarhate"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["thunder","compute"],"value":"Hi,
While reading the BlackBlaze Compute Story I came across, Packet's ARM Offering. (https://www.packet.net/bare-metal/servers/compute/)
Their c1.large.arm offers ARM Cavium ThunderX with 96 Physical Cores running at 2.0 Ghz.
Has anybody used such a configuration for Web Application Servers? Possibly something running Go?
Since many web applications need high concurrency and wait on DB queries, I think having these large number of cores might be beneficial.
Any thoughts and inputs on this will be well appreciated."},"title":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["thunder"],"value":"Ask HN: ARM Cavium ThunderX (96 Physical Cores) for Web Application Servers?"}},"_tags":["story","author_pritambarhate","story_16752184","ask_hn"],"author":"pritambarhate","children":[16752453],"created_at":"2018-04-04T05:22:53Z","created_at_i":1522819373,"num_comments":1,"objectID":"16752184","points":2,"story_id":16752184,"story_text":"Hi,
While reading the BlackBlaze Compute Story I came across, Packet's ARM Offering. (https://www.packet.net/bare-metal/servers/compute/)
Their c1.large.arm offers ARM Cavium ThunderX with 96 Physical Cores running at 2.0 Ghz.
Has anybody used such a configuration for Web Application Servers? Possibly something running Go?
Since many web applications need high concurrency and wait on DB queries, I think having these large number of cores might be beneficial.
Any thoughts and inputs on this will be well appreciated.","title":"Ask HN: ARM Cavium ThunderX (96 Physical Cores) for Web Application Servers?","updated_at":"2024-09-20T02:15:22Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"flatroze"},"title":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["compute"],"value":"ARM Desktop Tower Computer"},"url":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["thunder"],"value":"https://www.avantek.co.uk/store/avantek-32-core-cavium-thunderx-arm-desktop.html"}},"_tags":["story","author_flatroze","story_16069989"],"author":"flatroze","created_at":"2018-01-04T13:00:01Z","created_at_i":1515070801,"num_comments":0,"objectID":"16069989","points":2,"story_id":16069989,"title":"ARM Desktop Tower Computer","updated_at":"2024-09-20T01:54:49Z","url":"https://www.avantek.co.uk/store/avantek-32-core-cavium-thunderx-arm-desktop.html"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"rbanffy"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["thunder","compute"],"value":"ThunderX2 Arms Hyperscale and HPC Compute"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["thunder","compute"],"value":"https://www.nextplatform.com/2018/05/07/thunderx2-arms-hyperscale-and-hpc-compute/"}},"_tags":["story","author_rbanffy","story_17030046"],"author":"rbanffy","created_at":"2018-05-09T14:23:21Z","created_at_i":1525875801,"num_comments":0,"objectID":"17030046","points":3,"story_id":17030046,"title":"ThunderX2 Arms Hyperscale and HPC Compute","updated_at":"2024-09-20T02:23:42Z","url":"https://www.nextplatform.com/2018/05/07/thunderx2-arms-hyperscale-and-hpc-compute/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"timthorn"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["thunder","compute"],"value":"ThunderX2 Arms Hyperscale and HPC Compute"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["thunder","compute"],"value":"https://www.nextplatform.com/2018/05/07/thunderx2-arms-hyperscale-and-hpc-compute/"}},"_tags":["story","author_timthorn","story_17020265"],"author":"timthorn","created_at":"2018-05-08T11:45:20Z","created_at_i":1525779920,"num_comments":0,"objectID":"17020265","points":2,"story_id":17020265,"title":"ThunderX2 Arms Hyperscale and HPC Compute","updated_at":"2024-09-20T02:22:45Z","url":"https://www.nextplatform.com/2018/05/07/thunderx2-arms-hyperscale-and-hpc-compute/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"MediaSquirrel"},"story_text":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["compute"],"value":"About six months ago, I started working on a project to fine-tune Whisper locally on my M2 Ultra Mac Studio with a limited compute budget. I got into it. The problem I had at the time was I had 15,000 hours of audio data in Google Cloud Storage, and there was no way I could fit all the audio onto my local machine, so I built a system to stream data from my GCS to my machine during training.
Gemma 3n came out, so I added that. Kinda went nuts, tbh.
Then I put it on the shelf.
When Gemma 4 came out a few days ago, I dusted it off, cleaned it up, broke out the Gemma part from the Whisper fine-tuning and added support for Gemma 4.
I'm presenting it for you here today to play with, fork and improve upon.
One thing I have learned so far: It's very easy to OOM when you fine-tune on longer sequences! My local Mac Studio has 64GB RAM, so I run out of memory constantly.
Anywho, given how much interest there is in Gemma 4, and frankly, the fact that you can't really do audio fine-tuning with MLX, that's really the reason this exists (in addition to my personal interest). I would have preferred to use MLX and not have had to make this, but here we are. Welcome to my little side quest.
And so I made this. I hope you have as much fun using it as I had fun making it.
-Matt"},"title":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["thunder"],"value":"Show HN: Gemma 4 Multimodal Fine-Tuner for Apple Silicon"},"url":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["thunder"],"value":"https://github.com/mattmireles/gemma-tuner-multimodal"}},"_tags":["story","author_MediaSquirrel","story_47680309","show_hn"],"author":"MediaSquirrel","children":[47680482,47680630,47680929,47681012,47681304,47682463,47683355,47683578,47687221,47689925,47690986,47691011,47713340,47762949],"created_at":"2026-04-07T19:37:05Z","created_at_i":1775590625,"num_comments":28,"objectID":"47680309","points":235,"story_id":47680309,"story_text":"About six months ago, I started working on a project to fine-tune Whisper locally on my M2 Ultra Mac Studio with a limited compute budget. I got into it. The problem I had at the time was I had 15,000 hours of audio data in Google Cloud Storage, and there was no way I could fit all the audio onto my local machine, so I built a system to stream data from my GCS to my machine during training.
Gemma 3n came out, so I added that. Kinda went nuts, tbh.
Then I put it on the shelf.
When Gemma 4 came out a few days ago, I dusted it off, cleaned it up, broke out the Gemma part from the Whisper fine-tuning and added support for Gemma 4.
I'm presenting it for you here today to play with, fork and improve upon.
One thing I have learned so far: It's very easy to OOM when you fine-tune on longer sequences! My local Mac Studio has 64GB RAM, so I run out of memory constantly.
Anywho, given how much interest there is in Gemma 4, and frankly, the fact that you can't really do audio fine-tuning with MLX, that's really the reason this exists (in addition to my personal interest). I would have preferred to use MLX and not have had to make this, but here we are. Welcome to my little side quest.
And so I made this. I hope you have as much fun using it as I had fun making it.
-Matt","title":"Show HN: Gemma 4 Multimodal Fine-Tuner for Apple Silicon","updated_at":"2026-05-03T22:28:33Z","url":"https://github.com/mattmireles/gemma-tuner-multimodal"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"nottoday88"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["thunder","compute"],"value":"One thing I can say in my favour is that I'm a deep thinker, I try to understand and grok what is needed. What kind of computer related job would benefit from a deep thinker?,
How could I sell or promote this ability?"},"title":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["compute"],"value":"I'm 50, can I find a computer job? Some Elixir for getting Agile or Lean?"}},"_tags":["story","author_nottoday88","story_12031205","ask_hn"],"author":"nottoday88","children":[12031392,12032241,12032346,12032670,12034026,12035692,12036181],"created_at":"2016-07-04T15:13:06Z","created_at_i":1467645186,"num_comments":17,"objectID":"12031205","points":19,"story_id":12031205,"story_text":"One thing I can say in my favour is that I'm a deep thinker, I try to understand and grok what is needed. What kind of computer related job would benefit from a deep thinker?,
How could I sell or promote this ability?","title":"I'm 50, can I find a computer job? Some Elixir for getting Agile or Lean?","updated_at":"2024-09-19T23:24:15Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"hcrisp"},"title":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["thunder"],"value":"Apple Plans Thinner, Foldable iPhones to Revive Growth"},"url":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["compute"],"value":"https://www.wsj.com/tech/apple-thin-iphone-foldable-computer-plans-dbb4286c"}},"_tags":["story","author_hcrisp","story_42423508"],"author":"hcrisp","children":[42423556,42424350,42424835,42438823],"created_at":"2024-12-15T14:19:20Z","created_at_i":1734272360,"num_comments":3,"objectID":"42423508","points":4,"story_id":42423508,"title":"Apple Plans Thinner, Foldable iPhones to Revive Growth","updated_at":"2024-12-19T06:39:43Z","url":"https://www.wsj.com/tech/apple-thin-iphone-foldable-computer-plans-dbb4286c"},{"_highlightResult":{"author":{"fullyHighlighted":true,"matchLevel":"partial","matchedWords":["thunder"],"value":"thinxer"},"story_text":{"matchLevel":"none","matchedWords":[],"value":""},"title":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["compute"],"value":"Cuberox \u2013 six-screen waterproof linux-powered computer"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://www.kickstarter.com/projects/1426448868/cuberox-six-screen-waterproof-linux-powered-comput"}},"_tags":["story","author_thinxer","story_9127333"],"author":"thinxer","created_at":"2015-03-01T15:26:08Z","created_at_i":1425223568,"num_comments":0,"objectID":"9127333","points":1,"story_id":9127333,"story_text":"","title":"Cuberox \u2013 six-screen waterproof linux-powered computer","updated_at":"2023-09-07T02:12:14Z","url":"https://www.kickstarter.com/projects/1426448868/cuberox-six-screen-waterproof-linux-powered-comput"},{"_highlightResult":{"author":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["thunder"],"value":"ill_thinker"},"story_text":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["compute"],"value":"Today I was told by my manager that my arguments come across jumbled and incoherent.
This is not news to me. Since a young age I've found it incredibly difficult to reason about and form an argument for or against a topic. I first encountered this in my high school design and technology exam when I was asked to articulate why I would choose a particular material over another. No matter how much I tried, I simply could not reason about the different materials and which one would be better.
12 years on and I now have a PhD in computer science. I find software development extremely straightforward and the accompanying decision making easy. But during design discussions before development takes place, I still find it very difficult to present a reasoned argument as to why I'd design a piece of software one way over another. I mostly rely on an internal feeling that either provides positive or negative emotions (intuition). Project level decisions are also affected by this.
The absence of reasoned argument is not due to a lack of trying. When I attempt to reason about a problem I struggle to think of the questions I should be asking. Thoughts about a topic do not come naturally, and often they are very cyclic and disjointed. It is only through writing down my thoughts that my argument gains clarity. You can imagine how problematic this thought process is for active oral discussions, my discourse appears sporadic and poorly thought-out.
Is there anything I can do to improve my reasoning skills so I can participate coherently in oral discussions?"},"title":{"matchLevel":"none","matchedWords":[],"value":"Ask HN: How do I get better at creating well-reasoned arguments?"}},"_tags":["story","author_ill_thinker","story_24703551","ask_hn"],"author":"ill_thinker","children":[24703695,24703703,24705516,24709341,24710104,24710920,24712980,24713235,24714340,24728782],"created_at":"2020-10-06T23:14:28Z","created_at_i":1602026068,"num_comments":9,"objectID":"24703551","points":16,"story_id":24703551,"story_text":"Today I was told by my manager that my arguments come across jumbled and incoherent.
This is not news to me. Since a young age I've found it incredibly difficult to reason about and form an argument for or against a topic. I first encountered this in my high school design and technology exam when I was asked to articulate why I would choose a particular material over another. No matter how much I tried, I simply could not reason about the different materials and which one would be better.
12 years on and I now have a PhD in computer science. I find software development extremely straightforward and the accompanying decision making easy. But during design discussions before development takes place, I still find it very difficult to present a reasoned argument as to why I'd design a piece of software one way over another. I mostly rely on an internal feeling that either provides positive or negative emotions (intuition). Project level decisions are also affected by this.
The absence of reasoned argument is not due to a lack of trying. When I attempt to reason about a problem I struggle to think of the questions I should be asking. Thoughts about a topic do not come naturally, and often they are very cyclic and disjointed. It is only through writing down my thoughts that my argument gains clarity. You can imagine how problematic this thought process is for active oral discussions, my discourse appears sporadic and poorly thought-out.
Is there anything I can do to improve my reasoning skills so I can participate coherently in oral discussions?","title":"Ask HN: How do I get better at creating well-reasoned arguments?","updated_at":"2025-10-06T21:07:41Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"krrishd"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["thunder","compute"],"value":"I have a friend, who has an uncanny talent for anything mathematical. He can write formulas and functions to solve the most obscure math problems, and he has the mind of an algorithmic thinker that I and probably many wish they had. Think project Euler, he's really good at those types of problems.
He doesn't actually know any code, but I feel like he would be able to accomplish so many things with it. I've talked to him about it, and he's really interested, but he doesn't know where to start. Due to financial issues, he occasionally does not have Internet, but he does have a decent and functional computer to work with. What books/tutorials/resources should I give him so that he can put his mathematical and problem-solving talent to good use?"},"title":{"matchLevel":"none","matchedWords":[],"value":"Ask HN:Friend is math/algorithmic genius, wants to learn programming. Resources?"},"url":{"matchLevel":"none","matchedWords":[],"value":""}},"_tags":["story","author_krrishd","story_6854663","ask_hn"],"author":"krrishd","children":[6854848,6856927,6856941],"created_at":"2013-12-05T15:00:25Z","created_at_i":1386255625,"num_comments":4,"objectID":"6854663","points":8,"story_id":6854663,"story_text":"I have a friend, who has an uncanny talent for anything mathematical. He can write formulas and functions to solve the most obscure math problems, and he has the mind of an algorithmic thinker that I and probably many wish they had. Think project Euler, he's really good at those types of problems.
He doesn't actually know any code, but I feel like he would be able to accomplish so many things with it. I've talked to him about it, and he's really interested, but he doesn't know where to start. Due to financial issues, he occasionally does not have Internet, but he does have a decent and functional computer to work with. What books/tutorials/resources should I give him so that he can put his mathematical and problem-solving talent to good use?","title":"Ask HN:Friend is math/algorithmic genius, wants to learn programming. Resources?","updated_at":"2024-09-19T20:13:55Z","url":""},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"BobbyLLM"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["thunder","compute"],"value":"As a handsome local AI enjoyer\u2122 you\u2019ve probably noticed one of the big flaws with LLMs:
It lies. Confidently. ALL THE TIME.
I\u2019m autistic and extremely allergic to vibes-based tooling, so \u2026 I built a thing. Maybe it\u2019s useful to you too.
The thing: llama-conductor
llama-conductor is a router that sits between your frontend (eg: OWUI) & backend (llama.cpp + llama-swap). Local-first but it should talk to anything OpenAI-compatible if you point it there (note: experimental so YMMV).
LC is a glass-box that makes the stack behave like a deterministic system, instead of a drunk telling a story about the fish that got away.
TL;DR: \u201cIn God we trust. All others must bring data.\u201d
Three examples:
1. KB mechanics (markdown, JSON, checksums)
You keep \u201cknowledge\u201d as dumb folders on disk. Drop docs (.txt, .md, .pdf`) in them. Then:
>>attach <kb> - attaches a KB folder
>>summ new - generates SUMM_.md files with SHA-256 provenance baked in + moves the original to a sub-folder
Now, when you ask something like:
> \u201cyo, what did the Commodore C64 retail for in 1982?\u201d
..it answers from the attached KBs only.
If the fact isn\u2019t there, it tells you - explicitly - instead of winging it. Eg:
"The provided facts state the Commodore 64 launched at $595 and was reduced to $250, but do not specify a 1982 retail price. The Amiga\u2019s pricing and timeline are also not detailed in the given facts.
Missing information includes the exact 1982 retail price for Commodore\u2019s product line and which specific model(s) were sold then."
[Confidence: medium | Source: Mixed]
No vibes. Just: here\u2019s what\u2019s in your docs, here\u2019s what\u2019s missing, don't GIGO yourself into stupid.
Then, if you're happy with the summary, you can:
>>move to vault
2. Mentats: proof-or-refusal mode (Vault-only)
Mentats is the \u201cdeep think\u201d pipeline against your curated sources.
* no chat history
* no filesystem KBs
* no Vodka
* Vault-only grounding (Qdrant)
It runs a triple-pass (thinker \u2192 critic \u2192 thinker). It\u2019s slow on purpose. You can audit it. And if the Vault has nothing relevant? It refuses and tells you to go pound sand:
FINAL_ANSWER:
The provided facts do not contain information about the Acorn computer or its 1995 sale price.
Sources: Vault
FACTS_USED: NONE
[ZARDOZ HATH SPOKEN]
Also yes, it writes a mentats_debug.log. Go look at it any time you want.
The flow is basically:
Attach KBs \u2192 SUMM \u2192 Move to Vault \u2192 Mentats.
No mystery meat. No \u201ctrust me bro, embeddings.\u201d
3. Vodka: deterministic memory on a potato budget
Potato PCs have two classic problems: goldfish memory + context bloat that murders your VRAM.
Vodka fixes both without extra model compute.
* !! stores facts verbatim (JSON on disk)
* ?? recalls them verbatim (TTL + touch limits so memory doesn\u2019t become landfill)
* CTC (Cut The Crap)* hard-caps context (last N messages + char cap) and creates a concatenated summary (not LLM) so you don\u2019t get VRAM spikes after 400 messages
So instead of:
\u201cRemember my server is 203.0.113.42\u201d \u2192 \u201cGot it!\u201d \u2192 [100 msgs later] \u2192 \u201c127.0.0.1\u201d
you get:
!! my server is 203.0.113.42`\n?? server ip \u2192 203.0.113.42 (with TTL/touch metadata)
And because context stays bounded: stable KV cache, stable speed, your potato PC stops crying.
There\u2019s more (a lot more) in the README, but I\u2019ve already over-autism\u2019ed this post.
TL;DR:
If you want your local LLM to shut up when it doesn\u2019t know and show receipts when it does, come poke it:
Primary (Codeberg)\nhttps://codeberg.org/BobbyLLM/llama-conductor
Mirror (GitHub):\nhttps://github.com/BobbyLLM/llama-conductor
PS: Sorry about the AI slop image. I can't draw for shit.
PPS: A human with ASD wrote this using Notepad++. If it the formatting or language are weird, now you know why."},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: I'm tired of my LLM bullshitting. So I fixed it"}},"_tags":["story","author_BobbyLLM","story_46721773","show_hn"],"author":"BobbyLLM","children":[46722917,46723460,46726285,46730841,46730929],"created_at":"2026-01-22T16:50:09Z","created_at_i":1769100609,"num_comments":9,"objectID":"46721773","points":5,"story_id":46721773,"story_text":"As a handsome local AI enjoyer\u2122 you\u2019ve probably noticed one of the big flaws with LLMs:
It lies. Confidently. ALL THE TIME.
I\u2019m autistic and extremely allergic to vibes-based tooling, so \u2026 I built a thing. Maybe it\u2019s useful to you too.
The thing: llama-conductor
llama-conductor is a router that sits between your frontend (eg: OWUI) & backend (llama.cpp + llama-swap). Local-first but it should talk to anything OpenAI-compatible if you point it there (note: experimental so YMMV).
LC is a glass-box that makes the stack behave like a deterministic system, instead of a drunk telling a story about the fish that got away.
TL;DR: \u201cIn God we trust. All others must bring data.\u201d
Three examples:
1. KB mechanics (markdown, JSON, checksums)
You keep \u201cknowledge\u201d as dumb folders on disk. Drop docs (.txt, .md, .pdf`) in them. Then:
>>attach <kb> - attaches a KB folder
>>summ new - generates SUMM_.md files with SHA-256 provenance baked in + moves the original to a sub-folder
Now, when you ask something like:
> \u201cyo, what did the Commodore C64 retail for in 1982?\u201d
..it answers from the attached KBs only.
If the fact isn\u2019t there, it tells you - explicitly - instead of winging it. Eg:
"The provided facts state the Commodore 64 launched at $595 and was reduced to $250, but do not specify a 1982 retail price. The Amiga\u2019s pricing and timeline are also not detailed in the given facts.
Missing information includes the exact 1982 retail price for Commodore\u2019s product line and which specific model(s) were sold then."
[Confidence: medium | Source: Mixed]
No vibes. Just: here\u2019s what\u2019s in your docs, here\u2019s what\u2019s missing, don't GIGO yourself into stupid.
Then, if you're happy with the summary, you can:
>>move to vault
2. Mentats: proof-or-refusal mode (Vault-only)
Mentats is the \u201cdeep think\u201d pipeline against your curated sources.
* no chat history
* no filesystem KBs
* no Vodka
* Vault-only grounding (Qdrant)
It runs a triple-pass (thinker \u2192 critic \u2192 thinker). It\u2019s slow on purpose. You can audit it. And if the Vault has nothing relevant? It refuses and tells you to go pound sand:
FINAL_ANSWER:
The provided facts do not contain information about the Acorn computer or its 1995 sale price.
Sources: Vault
FACTS_USED: NONE
[ZARDOZ HATH SPOKEN]
Also yes, it writes a mentats_debug.log. Go look at it any time you want.
The flow is basically:
Attach KBs \u2192 SUMM \u2192 Move to Vault \u2192 Mentats.
No mystery meat. No \u201ctrust me bro, embeddings.\u201d
3. Vodka: deterministic memory on a potato budget
Potato PCs have two classic problems: goldfish memory + context bloat that murders your VRAM.
Vodka fixes both without extra model compute.
* !! stores facts verbatim (JSON on disk)
* ?? recalls them verbatim (TTL + touch limits so memory doesn\u2019t become landfill)
* CTC (Cut The Crap)* hard-caps context (last N messages + char cap) and creates a concatenated summary (not LLM) so you don\u2019t get VRAM spikes after 400 messages
So instead of:
\u201cRemember my server is 203.0.113.42\u201d \u2192 \u201cGot it!\u201d \u2192 [100 msgs later] \u2192 \u201c127.0.0.1\u201d
you get:
!! my server is 203.0.113.42`\n?? server ip \u2192 203.0.113.42 (with TTL/touch metadata)
And because context stays bounded: stable KV cache, stable speed, your potato PC stops crying.
There\u2019s more (a lot more) in the README, but I\u2019ve already over-autism\u2019ed this post.
TL;DR:
If you want your local LLM to shut up when it doesn\u2019t know and show receipts when it does, come poke it:
Primary (Codeberg)\nhttps://codeberg.org/BobbyLLM/llama-conductor
Mirror (GitHub):\nhttps://github.com/BobbyLLM/llama-conductor
PS: Sorry about the AI slop image. I can't draw for shit.
PPS: A human with ASD wrote this using Notepad++. If it the formatting or language are weird, now you know why.","title":"Show HN: I'm tired of my LLM bullshitting. So I fixed it","updated_at":"2026-04-29T08:12:32Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"palish"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["thunder","compute"],"value":"This was a post on Reddit that grew into a little essay. What do you think of it? I'm trying to find the correct answers, so if something is incorrect then please rub it in my face. :)
--
The second system effect is when a designer isn't new on the job anymore (he's made at least one system) and he has an overconfidence in his abilities. Because of that, the second system he designs is usually an overcomplicated mess. There are exceptions. Woz may not have designed an overcomplicated second system, for example. And if you understand how little you understand when you're first starting out, you're more likely to build small, stable systems instead of large, brittle systems. But history has pretty much proven that inexperience + overconfidence = mess, regardless of your profession.
The first system usually goes well because you're careful. You want your first system to go well since you will be judged by it. You keep your opinions to yourself and use proven methods that you fully understand to design your system. After you've designed this beautiful little gem (or not-so-beautiful but working, so the creator may think it is beautiful), you're left with a feeling of accomplishment. In fact, that first system wasn't really hard at all! You read the requirements and delivered a system that did exactly that. But look at how little that first system accomplished. In fact, it was even rigid; it did exactly what was needed without an eye towards extensibility. You know you can do better.
But.. It seems to take a lot of experience to correctly design a large, extensible system. I really hate to say the e-word since (being 19) I hate the idea that there's this magical quality, experience, that you'll only get after a certain amount of time has passed, regardless of anything you do to try to accelerate the process. But it seems true. It seems like to be good designers, we need to immerse ourselves in as many projects as possible; understand that we know very little; and most importantly, get things delivered, even if they are only at the 80% mark.
The Mythical Man Month is a good read. Honestly, it's extremely.. well.. boring. The advice was penned for a very different time, so a lot of it doesn't apply today (a programmer needs a personal secretary? That would be nice, but probably not required). But if you read it through the shades of, \"Wow, he is a really creative thinker!\" then you'll enjoy it a lot. He talks about why the best designers have usually created at least two systems. The first gives them experience, and the second gives them humility.
Choose two: beautiful, meets all requirements, stable.
\"Beautiful, and meets all requirements\" is the second system effect. It's okay to choose this if you're designing a small system.
\"Meets all requirements, and stable\" seems to be how most systems are designed. These types of systems are the moneymakers, so if you're doing a startup, it's probably best to choose this option. As Trevor says, it's more like a model train set instead of an elegant mathematical proof. There can be bumps on the tracks, but they won't stop the train from going 'round. Windows is a good example of this. You might think, \"Windows? Stable? What an idiot.\" But look at its size. It's humongous. It's impressive how much it has accomplished, even if it's far from perfect.
\"Beautiful, and stable\" seems to be the best choice to strive for, once you have the experience. It'll turn out that some of the features you desperately need aren't so critical after all. The Apple II is a great example of this. When it was released, everyone else thought that you needed to sell a monitor with a computer, and that a TV was inadequate. By cutting the need to design a whole monitor, Woz was able to produce something much cheaper and fabulously stable. Also, choosing \"beautiful and stable\" results in a system that \"is beautiful, is stable, and meets all requirements\", since you've just trimmed half of your requirements.
From my very limited two years of experience, this seems to be a decent design process for complicated systems:
1. Accept that you don't know as much as you need to know. Look for battle-tested components that other people have already designed, and use those. Copying is not cheating in the world of design. (Give credit where credit is due, though. If you created all your components, fabulous. That's incredibly impressive. But don't claim to have designed everything even if you've made them all work together.)
2. Cut as much from the requirements as possible, and even a little bit beyond what's comfortable. Think to yourself, \"How can I avoid designing a new monitor?\"
3. Focus on simplicity. Keep the adage \"form follows function\" in mind. If the implementation is small and simple, it will be far more stable than a big, extensible system that does a lot more. And people will probably like the smaller system better anyway.
4. These are only guidelines. Just focus on doing what makes the most sense, regardless of any rules. (For C++ programmers, this includes rules against using Boost or STL because \"it's not fast enough\". Take the time to learn your containers, then choose to not use them once you understand them.)
5. Deliver a system when it's 80% complete. (That's 80% complete, not 20% buggy.) Software is powerful because it can be improved later. Draw a sketch, don't paint a painting. The result can always be transformed into a painting later.
What do you think? It's always great to find new things that help designers, so please share your wisdom on all of this."},"title":{"matchLevel":"none","matchedWords":[],"value":"How to Design a System"},"url":{"matchLevel":"none","matchedWords":[],"value":""}},"_tags":["story","author_palish","story_74725","ask_hn"],"author":"palish","children":[74732,74809,74868,76781],"created_at":"2007-10-31T17:27:45Z","created_at_i":1193851665,"num_comments":5,"objectID":"74725","points":5,"story_id":74725,"story_text":"This was a post on Reddit that grew into a little essay. What do you think of it? I'm trying to find the correct answers, so if something is incorrect then please rub it in my face. :)
--
The second system effect is when a designer isn't new on the job anymore (he's made at least one system) and he has an overconfidence in his abilities. Because of that, the second system he designs is usually an overcomplicated mess. There are exceptions. Woz may not have designed an overcomplicated second system, for example. And if you understand how little you understand when you're first starting out, you're more likely to build small, stable systems instead of large, brittle systems. But history has pretty much proven that inexperience + overconfidence = mess, regardless of your profession.
The first system usually goes well because you're careful. You want your first system to go well since you will be judged by it. You keep your opinions to yourself and use proven methods that you fully understand to design your system. After you've designed this beautiful little gem (or not-so-beautiful but working, so the creator may think it is beautiful), you're left with a feeling of accomplishment. In fact, that first system wasn't really hard at all! You read the requirements and delivered a system that did exactly that. But look at how little that first system accomplished. In fact, it was even rigid; it did exactly what was needed without an eye towards extensibility. You know you can do better.
But.. It seems to take a lot of experience to correctly design a large, extensible system. I really hate to say the e-word since (being 19) I hate the idea that there's this magical quality, experience, that you'll only get after a certain amount of time has passed, regardless of anything you do to try to accelerate the process. But it seems true. It seems like to be good designers, we need to immerse ourselves in as many projects as possible; understand that we know very little; and most importantly, get things delivered, even if they are only at the 80% mark.
The Mythical Man Month is a good read. Honestly, it's extremely.. well.. boring. The advice was penned for a very different time, so a lot of it doesn't apply today (a programmer needs a personal secretary? That would be nice, but probably not required). But if you read it through the shades of, \"Wow, he is a really creative thinker!\" then you'll enjoy it a lot. He talks about why the best designers have usually created at least two systems. The first gives them experience, and the second gives them humility.
Choose two: beautiful, meets all requirements, stable.
\"Beautiful, and meets all requirements\" is the second system effect. It's okay to choose this if you're designing a small system.
\"Meets all requirements, and stable\" seems to be how most systems are designed. These types of systems are the moneymakers, so if you're doing a startup, it's probably best to choose this option. As Trevor says, it's more like a model train set instead of an elegant mathematical proof. There can be bumps on the tracks, but they won't stop the train from going 'round. Windows is a good example of this. You might think, \"Windows? Stable? What an idiot.\" But look at its size. It's humongous. It's impressive how much it has accomplished, even if it's far from perfect.
\"Beautiful, and stable\" seems to be the best choice to strive for, once you have the experience. It'll turn out that some of the features you desperately need aren't so critical after all. The Apple II is a great example of this. When it was released, everyone else thought that you needed to sell a monitor with a computer, and that a TV was inadequate. By cutting the need to design a whole monitor, Woz was able to produce something much cheaper and fabulously stable. Also, choosing \"beautiful and stable\" results in a system that \"is beautiful, is stable, and meets all requirements\", since you've just trimmed half of your requirements.
From my very limited two years of experience, this seems to be a decent design process for complicated systems:
1. Accept that you don't know as much as you need to know. Look for battle-tested components that other people have already designed, and use those. Copying is not cheating in the world of design. (Give credit where credit is due, though. If you created all your components, fabulous. That's incredibly impressive. But don't claim to have designed everything even if you've made them all work together.)
2. Cut as much from the requirements as possible, and even a little bit beyond what's comfortable. Think to yourself, \"How can I avoid designing a new monitor?\"
3. Focus on simplicity. Keep the adage \"form follows function\" in mind. If the implementation is small and simple, it will be far more stable than a big, extensible system that does a lot more. And people will probably like the smaller system better anyway.
4. These are only guidelines. Just focus on doing what makes the most sense, regardless of any rules. (For C++ programmers, this includes rules against using Boost or STL because \"it's not fast enough\". Take the time to learn your containers, then choose to not use them once you understand them.)
5. Deliver a system when it's 80% complete. (That's 80% complete, not 20% buggy.) Software is powerful because it can be improved later. Draw a sketch, don't paint a painting. The result can always be transformed into a painting later.
What do you think? It's always great to find new things that help designers, so please share your wisdom on all of this.","title":"How to Design a System","updated_at":"2024-09-19T16:23:24Z","url":""},{"_highlightResult":{"author":{"fullyHighlighted":true,"matchLevel":"partial","matchedWords":["thunder"],"value":"twunde"},"story_text":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["compute"],"value":"I left my last job because of burnout but now after 2 years here I'm picking up the same bad habits, checking email and doing extra work on weekends. I know one of the triggers was getting a work laptop, which I now use as my primary computer. How do you keep a balance between life and work?"},"title":{"matchLevel":"none","matchedWords":[],"value":"How do you keep a good work/life balance"}},"_tags":["story","author_twunde","story_11604325","ask_hn"],"author":"twunde","children":[11605423,11606238,11613692,11684464],"created_at":"2016-05-01T00:32:09Z","created_at_i":1462062729,"num_comments":4,"objectID":"11604325","points":3,"story_id":11604325,"story_text":"I left my last job because of burnout but now after 2 years here I'm picking up the same bad habits, checking email and doing extra work on weekends. I know one of the triggers was getting a work laptop, which I now use as my primary computer. How do you keep a balance between life and work?","title":"How do you keep a good work/life balance","updated_at":"2024-09-19T23:06:03Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"brian68"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["thunder","compute"],"value":"As a person with vision problems I sometimes struggle with certain documents that are difficult to read (my computer fonts are usually bold with sizes ~14). I needed to read some RFCs and discovered that the PDF versions use a font smaller and "thinner" (Courier 10?) than what I'm used to. I tried some options to modify the original plain text documents and discovered that modifying their fonts was not as easy as I thought. Finally I created a utility to convert RFCs to PDF allowing the user to change fonts and margins and I wanted to share it. There is also a utility to "bulk convert" a folder with RFCs in plain text.I have converted RFCs to PDFs with two free monospaced fonts: Liberation Mono and Ubuntu Mono, check the releases. Hope you find it useful!."},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: Rfc2pdf"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://github.com/OscarMaestre/rfc2pdf"}},"_tags":["story","author_brian68","story_34700470","show_hn"],"author":"brian68","children":[34704356],"created_at":"2023-02-07T22:06:31Z","created_at_i":1675807591,"num_comments":2,"objectID":"34700470","points":3,"story_id":34700470,"story_text":"As a person with vision problems I sometimes struggle with certain documents that are difficult to read (my computer fonts are usually bold with sizes ~14). I needed to read some RFCs and discovered that the PDF versions use a font smaller and "thinner" (Courier 10?) than what I'm used to. I tried some options to modify the original plain text documents and discovered that modifying their fonts was not as easy as I thought. Finally I created a utility to convert RFCs to PDF allowing the user to change fonts and margins and I wanted to share it. There is also a utility to "bulk convert" a folder with RFCs in plain text.I have converted RFCs to PDFs with two free monospaced fonts: Liberation Mono and Ubuntu Mono, check the releases. Hope you find it useful!.","title":"Show HN: Rfc2pdf","updated_at":"2024-09-20T13:09:25Z","url":"https://github.com/OscarMaestre/rfc2pdf"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"wwwlouishinofun"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["thunder","compute"],"value":"1.safari
Frankly, I only care if the new system is faster chrome, and Safari currently has only two advantages for me:
A. Live Text is really easy to use, I think Apple's server cost is high for this OCR feature, which is not supported by Chrome on Windows or macOS (pausing to copy video text on YouTube is very useful)
B. Compact layout and UI design really increase the vertical space, it can integrate tabs, address bar, and extension icons on a single bar, which is great, I use trackpad gestures to switch and close tabs, so it doesn't increase the difficulty of operation, which I have reported to Google many times but their UI design is slow to improve, for the screen of the first resolution, I believe that Chrome's UI will take up a huge space
In addition to this, there are too many disadvantages of Safari:
The speed of switching between multiple tabs is not comparable to chrome at all,Safari will drop frames when youtube videos are played at twice the speed of 4k,Insufficient extension,I'm a senior browser enthusiast,I still chose chrome for speed,Chrome's current problem is that two-finger zoom is not sensitive
2.iphone
I've long since stopped using an iPhone, so I don't care about the phone mirroring feature that I've experienced for a long time, in essence, the wireless network they use to transmit data is always not the ultimate smooth experience
I'm using a folding screen phone, believe me, a 220G, thinner than the iPhone folding machine is the ideal Internet terminal, watching the content is very comfortable, unless you like the phone to record videos, otherwise I can't think of a reason to continue using the iPhone in 2024
3.ai features and interactions
I'm very tired of making wheels repeatedly, I believe that Apple AI 10,000% is not as good as my GPT4 is a hundred times better, but Apple's advantage is that it can be linked to the hardware of the computer, which has the potential, but it is not enough for me to pay for it at present
4. So far, Apple has few advantages left
UI and other aesthetic design, I give this one full marks
The leader of the ARM chip PC, such as the MacBookM1, the peripheral power consumption is 5W (screen and the like), and the video SOC is only 1W, so that the power consumption of the whole machine is 7-8W, and the 49Wh battery can also play 7 hours of continuous video
Below is the original document, I am afraid that the machine translation is not accurate, I am currently also studying IELTS, but English is really difficult for me, I am now working in Viet Nam, I am very much looking forward to studying in English-speaking countries, no matter what, I expect Apple to continue to improve in software and hardware, do not squeeze toothpaste1.safari
Frankly, I only care if the new system is faster chrome, and Safari currently has only two advantages for me:
A. Live Text is really easy to use, I think Apple's server cost is high for this OCR feature, which is not supported by Chrome on Windows or macOS (pausing to copy video text on YouTube is very useful)
B. Compact layout and UI design really increase the vertical space, it can integrate tabs, address bar, and extension icons on a single bar, which is great, I use trackpad gestures to switch and close tabs, so it doesn't increase the difficulty of operation, which I have reported to Google many times but their UI design is slow to improve, for the screen of the first resolution, I believe that Chrome's UI will take up a huge space
In addition to this, there are too many disadvantages of Safari:
The speed of switching between multiple tabs is not comparable to chrome at all,Safari will drop frames when youtube videos are played at twice the speed of 4k,Insufficient extension,I'm a senior browser enthusiast,I still chose chrome for speed,Chrome's current problem is that two-finger zoom is not sensitive"},"title":{"matchLevel":"none","matchedWords":[],"value":"Let's talk in great detail about whether macOS 15 Sequoia is worth upgrading"}},"_tags":["story","author_wwwlouishinofun","story_41564250","ask_hn"],"author":"wwwlouishinofun","children":[41564270,41564561,41564568,41566889],"created_at":"2024-09-17T04:53:10Z","created_at_i":1726548790,"num_comments":7,"objectID":"41564250","points":1,"story_id":41564250,"story_text":"1.safari
Frankly, I only care if the new system is faster chrome, and Safari currently has only two advantages for me:
A. Live Text is really easy to use, I think Apple's server cost is high for this OCR feature, which is not supported by Chrome on Windows or macOS (pausing to copy video text on YouTube is very useful)
B. Compact layout and UI design really increase the vertical space, it can integrate tabs, address bar, and extension icons on a single bar, which is great, I use trackpad gestures to switch and close tabs, so it doesn't increase the difficulty of operation, which I have reported to Google many times but their UI design is slow to improve, for the screen of the first resolution, I believe that Chrome's UI will take up a huge space
In addition to this, there are too many disadvantages of Safari:
The speed of switching between multiple tabs is not comparable to chrome at all,Safari will drop frames when youtube videos are played at twice the speed of 4k,Insufficient extension,I'm a senior browser enthusiast,I still chose chrome for speed,Chrome's current problem is that two-finger zoom is not sensitive
2.iphone
I've long since stopped using an iPhone, so I don't care about the phone mirroring feature that I've experienced for a long time, in essence, the wireless network they use to transmit data is always not the ultimate smooth experience
I'm using a folding screen phone, believe me, a 220G, thinner than the iPhone folding machine is the ideal Internet terminal, watching the content is very comfortable, unless you like the phone to record videos, otherwise I can't think of a reason to continue using the iPhone in 2024
3.ai features and interactions
I'm very tired of making wheels repeatedly, I believe that Apple AI 10,000% is not as good as my GPT4 is a hundred times better, but Apple's advantage is that it can be linked to the hardware of the computer, which has the potential, but it is not enough for me to pay for it at present
4. So far, Apple has few advantages left
UI and other aesthetic design, I give this one full marks
The leader of the ARM chip PC, such as the MacBookM1, the peripheral power consumption is 5W (screen and the like), and the video SOC is only 1W, so that the power consumption of the whole machine is 7-8W, and the 49Wh battery can also play 7 hours of continuous video
Below is the original document, I am afraid that the machine translation is not accurate, I am currently also studying IELTS, but English is really difficult for me, I am now working in Viet Nam, I am very much looking forward to studying in English-speaking countries, no matter what, I expect Apple to continue to improve in software and hardware, do not squeeze toothpaste1.safari
Frankly, I only care if the new system is faster chrome, and Safari currently has only two advantages for me:
A. Live Text is really easy to use, I think Apple's server cost is high for this OCR feature, which is not supported by Chrome on Windows or macOS (pausing to copy video text on YouTube is very useful)
B. Compact layout and UI design really increase the vertical space, it can integrate tabs, address bar, and extension icons on a single bar, which is great, I use trackpad gestures to switch and close tabs, so it doesn't increase the difficulty of operation, which I have reported to Google many times but their UI design is slow to improve, for the screen of the first resolution, I believe that Chrome's UI will take up a huge space
In addition to this, there are too many disadvantages of Safari:
The speed of switching between multiple tabs is not comparable to chrome at all,Safari will drop frames when youtube videos are played at twice the speed of 4k,Insufficient extension,I'm a senior browser enthusiast,I still chose chrome for speed,Chrome's current problem is that two-finger zoom is not sensitive","title":"Let's talk in great detail about whether macOS 15 Sequoia is worth upgrading","updated_at":"2024-09-20T17:50:56Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"dennis_jeeves"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["thunder","compute"],"value":"Looking to freeload (in a manner speaking ). I'm seeking existing established businesses where the people at the helm of affairs are searching to augment their key members with level headed, non-political individuals. (Well, it's lonely up there isn't it?). \nCertainly I'm open to people who are wage slaves like me who would like to get off the wage slavery bandwagon for any future potential partnerships.
I'm 43 years old, a computer programmer by profession and surrounded by regular (and younger) folks who "just don't get it". Never been an employee material, but being an entrepreneur or starting a business is fraught with risks that could leave one devastated in an irrecoverable manner, hence this post. Essentially I'm trying to ensure my long term livelihood, survival, and sanity.
What's in for you? Essentially you get a 'critical' thinker who upholds and enforces ( if I have the authority) some values that are not obvious to most people in a common workplace. For example:\n- Mindless pursuit of profits/sales figures at the expense of other important things, for example health and well being.\n- Way too many meetings.\n- promotion of talkers over thinkers. ( there are exceptions to this)\n- Frequently promoting short term fixes over long term ones.\n- Falling for hype and buzzwords( for example most of the "big data" topics/implementations are BS).\n- confusing overwork with productivity.
etc...
It is important to note that I can actually do work ( like programming) as opposed to be an armchair philosopher."},"title":{"matchLevel":"none","matchedWords":[],"value":"Businesses looking for partners/successors?"}},"_tags":["story","author_dennis_jeeves","story_15424617","ask_hn"],"author":"dennis_jeeves","children":[15425631],"created_at":"2017-10-07T17:12:07Z","created_at_i":1507396327,"num_comments":2,"objectID":"15424617","points":1,"story_id":15424617,"story_text":"Looking to freeload (in a manner speaking ). I'm seeking existing established businesses where the people at the helm of affairs are searching to augment their key members with level headed, non-political individuals. (Well, it's lonely up there isn't it?). \nCertainly I'm open to people who are wage slaves like me who would like to get off the wage slavery bandwagon for any future potential partnerships.
I'm 43 years old, a computer programmer by profession and surrounded by regular (and younger) folks who "just don't get it". Never been an employee material, but being an entrepreneur or starting a business is fraught with risks that could leave one devastated in an irrecoverable manner, hence this post. Essentially I'm trying to ensure my long term livelihood, survival, and sanity.
What's in for you? Essentially you get a 'critical' thinker who upholds and enforces ( if I have the authority) some values that are not obvious to most people in a common workplace. For example:\n- Mindless pursuit of profits/sales figures at the expense of other important things, for example health and well being.\n- Way too many meetings.\n- promotion of talkers over thinkers. ( there are exceptions to this)\n- Frequently promoting short term fixes over long term ones.\n- Falling for hype and buzzwords( for example most of the "big data" topics/implementations are BS).\n- confusing overwork with productivity.
etc...
It is important to note that I can actually do work ( like programming) as opposed to be an armchair philosopher.","title":"Businesses looking for partners/successors?","updated_at":"2024-09-20T01:27:32Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"weakgeekinseek"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["thunder","compute"],"value":"i am an entrepreneurial sharp business mind in search of a technologically savvy counterpart. i have many ideas including a previous sold at in excess of $200 million. i am a thinker with a knack for financial gains but very limited to my software development knowledge. i am looking for a computer wiz to help develop my ideas and form partnerships for all possible ventures possible with your skills. my email is 150mikkimase@gmail.com and my cell is 6467097799. please contact me, Mikki, at any time suitable for you. i look forward to your inquiries and am excited to move forward with my new partner(s)."},"title":{"matchLevel":"none","matchedWords":[],"value":"Entrepreneur searching for developer counterpart"},"url":{"matchLevel":"none","matchedWords":[],"value":""}},"_tags":["story","author_weakgeekinseek","story_9620290","ask_hn"],"author":"weakgeekinseek","created_at":"2015-05-28T18:54:42Z","created_at_i":1432839282,"num_comments":0,"objectID":"9620290","points":1,"story_id":9620290,"story_text":"i am an entrepreneurial sharp business mind in search of a technologically savvy counterpart. i have many ideas including a previous sold at in excess of $200 million. i am a thinker with a knack for financial gains but very limited to my software development knowledge. i am looking for a computer wiz to help develop my ideas and form partnerships for all possible ventures possible with your skills. my email is 150mikkimase@gmail.com and my cell is 6467097799. please contact me, Mikki, at any time suitable for you. i look forward to your inquiries and am excited to move forward with my new partner(s).","title":"Entrepreneur searching for developer counterpart","updated_at":"2023-09-07T02:51:02Z","url":""},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"dennis_jeeves"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["thunder","compute"],"value":"\tLooking to freeload (in a manner speaking ). I'm seeking existing established businesses where the people at the helm of affairs are searching to augment their key members with level headed, non-political individuals. (Well, it's lonely up there isn't it?). Certainly I'm open to people who are wage slaves like me who would like to get off the wage slavery bandwagon for any future potential partnerships.\nI'm 43 years old, a computer programmer by profession and surrounded by regular (and younger) folks who "just don't get it". Never been an employee material, but being an entrepreneur or starting a business is fraught with risks that could leave one devastated in an irrecoverable manner, hence this post. Essentially I'm trying to ensure my long term livelihood, survival, and sanity.
What's in for you? Essentially you get a 'critical' thinker who upholds and enforces ( if I have the authority) some values that are not obvious to most people in a common workplace. For example: - Mindless pursuit of profits/sales figures at the expense of other important things, for example health and well being. - Way too many meetings. - promotion of talkers over thinkers. ( there are exceptions to this) - Frequently promoting short term fixes over long term ones. - Falling for hype and buzzwords( for example most of the "big data" topics/implementations are BS). - confusing overwork with productivity.
etc...
It is important to note that I can actually do work ( like programming) as opposed to be an armchair philosopher."},"title":{"matchLevel":"none","matchedWords":[],"value":"Businesses looking for partners/successors?"}},"_tags":["story","author_dennis_jeeves","story_15465563","ask_hn"],"author":"dennis_jeeves","created_at":"2017-10-13T13:46:23Z","created_at_i":1507902383,"num_comments":0,"objectID":"15465563","points":1,"story_id":15465563,"story_text":"\tLooking to freeload (in a manner speaking ). I'm seeking existing established businesses where the people at the helm of affairs are searching to augment their key members with level headed, non-political individuals. (Well, it's lonely up there isn't it?). Certainly I'm open to people who are wage slaves like me who would like to get off the wage slavery bandwagon for any future potential partnerships.\nI'm 43 years old, a computer programmer by profession and surrounded by regular (and younger) folks who "just don't get it". Never been an employee material, but being an entrepreneur or starting a business is fraught with risks that could leave one devastated in an irrecoverable manner, hence this post. Essentially I'm trying to ensure my long term livelihood, survival, and sanity.
What's in for you? Essentially you get a 'critical' thinker who upholds and enforces ( if I have the authority) some values that are not obvious to most people in a common workplace. For example: - Mindless pursuit of profits/sales figures at the expense of other important things, for example health and well being. - Way too many meetings. - promotion of talkers over thinkers. ( there are exceptions to this) - Frequently promoting short term fixes over long term ones. - Falling for hype and buzzwords( for example most of the "big data" topics/implementations are BS). - confusing overwork with productivity.
etc...
It is important to note that I can actually do work ( like programming) as opposed to be an armchair philosopher.","title":"Businesses looking for partners/successors?","updated_at":"2024-09-20T01:31:25Z"},{"_highlightResult":{"author":{"fullyHighlighted":true,"matchLevel":"partial","matchedWords":["thunder"],"value":"thund"},"title":{"matchLevel":"none","matchedWords":[],"value":"Uses for an old Android device"},"url":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["compute"],"value":"https://www.computerworld.com/article/2487680/20-great-uses-for-an-old-android-device.html"}},"_tags":["story","author_thund","story_32383835"],"author":"thund","children":[32384804,32395035,32395142,32395147,32395171,32395339,32395378,32395426,32395460,32395472,32395616,32395688,32395716,32395848,32395850,32395921,32396016,32396050,32396231,32396859,32397077,32397117,32397174,32397313,32399512,32399687,32400050,32400100,32400218,32401615],"created_at":"2022-08-08T10:00:08Z","created_at_i":1659952808,"num_comments":103,"objectID":"32383835","points":138,"story_id":32383835,"title":"Uses for an old Android device","updated_at":"2026-07-27T04:54:45Z","url":"https://www.computerworld.com/article/2487680/20-great-uses-for-an-old-android-device.html"},{"_highlightResult":{"author":{"fullyHighlighted":true,"matchLevel":"partial","matchedWords":["thunder"],"value":"thund3rp"},"story_text":{"matchLevel":"none","matchedWords":[],"value":""},"title":{"matchLevel":"none","matchedWords":[],"value":"Al-Qaida comunications hacked.."},"url":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["compute"],"value":"http://www.computerworld.com/s/article/9218050/Al_Qaida_communications_knocked_offline_says_expert_"}},"_tags":["story","author_thund3rp","story_2713224"],"author":"thund3rp","created_at":"2011-06-30T09:16:40Z","created_at_i":1309425400,"num_comments":0,"objectID":"2713224","points":2,"story_id":2713224,"story_text":"","title":"Al-Qaida comunications hacked..","updated_at":"2024-09-19T17:47:33Z","url":"http://www.computerworld.com/s/article/9218050/Al_Qaida_communications_knocked_offline_says_expert_"},{"_highlightResult":{"author":{"fullyHighlighted":true,"matchLevel":"partial","matchedWords":["thunder"],"value":"thinxer"},"story_text":{"matchLevel":"none","matchedWords":[],"value":""},"title":{"matchLevel":"none","matchedWords":[],"value":"Deep Learning vs. Big Data: Who Owns What?"},"url":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["compute"],"value":"http://www.computervisionblog.com/2015/05/deep-learning-vs-big-data-who-owns-what.html"}},"_tags":["story","author_thinxer","story_9568263"],"author":"thinxer","created_at":"2015-05-19T05:26:29Z","created_at_i":1432013189,"num_comments":0,"objectID":"9568263","points":1,"story_id":9568263,"story_text":"","title":"Deep Learning vs. Big Data: Who Owns What?","updated_at":"2023-09-07T02:47:03Z","url":"http://www.computervisionblog.com/2015/05/deep-learning-vs-big-data-who-owns-what.html"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"de6u99er"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["thunder","compute"],"value":"Since the recent updates, I have the feeling that Claude Opus is becoming dumber on complex tasks.
Things that used to work cleanly in a single prompt now fail completely. In coding workflows, it constantly ignores mandatory CLAUDE.md project rules, makes unsolicited edits to unrelated files, and breaks working code. It starts arguing based on stale comments, fails to update documentation when code changes, and edits based on blind guesswork instead of actually verifying the codebase first.
I even caught myself using Fable for tasks that I always used Opus for in the past, just to get decent results.
It feels like Anthropic is shifting their technical problems onto paying users. Either they are using strict safety filters that silently fall back to cheaper models without telling us, or they are downgrading the compute under heavy load to save money.
The result is the same kind of shrinkflation: we pay the same subscription price, but we either burn way more Opus tokens on endless retries, or we are forced to spend money on more expensive Fable tokens to get the quality we used to have.
Has anyone here successfully moved away from Claude for complex engineering workflows? What are you replacing it with, and how are you handling the transition?
Disclaimer: As a native German speaker, I used Gemini to clean up and polish the English text for this post."},"title":{"matchLevel":"none","matchedWords":[],"value":"Is it just me, or has Claude Opus gotten worse recently?"}},"_tags":["story","author_de6u99er","story_49519639","ask_hn"],"author":"de6u99er","children":[49519705,49519710,49519739,49519781,49519959,49519987,49520067,49520445,49520885,49521030,49521097,49521696,49524250,49525033,49526617,49531112,49546397,49547246],"created_at":"2026-09-01T08:58:16Z","created_at_i":1788253096,"num_comments":21,"objectID":"49519639","points":12,"story_id":49519639,"story_text":"Since the recent updates, I have the feeling that Claude Opus is becoming dumber on complex tasks.
Things that used to work cleanly in a single prompt now fail completely. In coding workflows, it constantly ignores mandatory CLAUDE.md project rules, makes unsolicited edits to unrelated files, and breaks working code. It starts arguing based on stale comments, fails to update documentation when code changes, and edits based on blind guesswork instead of actually verifying the codebase first.
I even caught myself using Fable for tasks that I always used Opus for in the past, just to get decent results.
It feels like Anthropic is shifting their technical problems onto paying users. Either they are using strict safety filters that silently fall back to cheaper models without telling us, or they are downgrading the compute under heavy load to save money.
The result is the same kind of shrinkflation: we pay the same subscription price, but we either burn way more Opus tokens on endless retries, or we are forced to spend money on more expensive Fable tokens to get the quality we used to have.
Has anyone here successfully moved away from Claude for complex engineering workflows? What are you replacing it with, and how are you handling the transition?
Disclaimer: As a native German speaker, I used Gemini to clean up and polish the English text for this post.","title":"Is it just me, or has Claude Opus gotten worse recently?","updated_at":"2026-09-04T19:19:52Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"section33"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["thunder","compute"],"value":"I might be the only SRE on Earth with his own bowling center. It's a more in-depth gig than you'd think.
My family and I bought an abandoned 8-lane bowling center in the rural mid-west. In our small town there weren't many recreation options for families. You've heard of a food desert? This is an R&R desert.
It had been abandoned for a good reason. The roof leaks, the electrical system was constantly surging, and my 70-year-old bowling equipment (still) doesn't work perfectly. The system that keeps your score is particularly interesting to me. It's the thing you watch during your game, but it fades into the background beyond that. Turns out these things are really cool, but absurdly expensive.
Ours was installed in 2008 and cost six figures. It's calculating ball speed and trajectory, camera-based pin detection (object detection and trig, on ICs!), runs the fouling, the animations, the pinsetting machine and ball return. Very cool stuff for its age.
From the business perspective, my facility only cost me $105k. To forklift-replace the score keeping system runs anywhere between $80-$120k, depending on features, vendor, and unit age. No upgrades or service contracts, mind you, and every feature and customization is a new line item. That's for a 1:1 replacement on a system installed in 2008. Incredible, given how fast the tech world moves.
Replacement parts cost a shocking $4000 per pair of lanes. But wait, the bowling machines themselves are 70 years old, so what's this "advanced" system actually doing back there? Actuating a single relay to trigger that big old machine. Everything else is strictly mechanical. Meanwhile I've got a six-figure invoice in my hand. I'm upset.
Given the state of open hardware, computer vision, real-time event streaming, and open source running megascale products worldwide, there had to be a way to do this myself.
So far I've built an equivalent prototype for about $200 per lane-pair, $400 if you're fancy. ESP32 and ESPNow with an RS485 fallback, reporting to a raspberry pi lane computer that's really just redis and a state machine bolted to an ESP32 gateway for the mesh.
Since it's all ESP32, I've got a fistful of spare controllers in a drawer, pre-flashed or waiting to be. All common off-the-shelf hardware: microcontrollers wired to relays, optocouplers, and IR-break-beam sensors, each running slightly different firmware. Writing the firmware and protocol is the actual hard part.
It's an ESPNow star-topology mesh: each node emits events from its sensors and accepts commands for its controls, reporting to a gateway node connected to the raspi over UART. From there it's event streaming: RX packets get translated and tossed into redis, commands relay back out to the mesh as needed. RS485 sits underneath as a wired fallback for noisy RF environments.
Once the data's in redis, it's familiar middleware/React/websocket/pub-sub stuff. Any React dev can build their own UI and bowling animations. Since it all runs on commodity hardware, I can do legit anything I want as the proprietor, and I own all my data. Repairs take five minutes; I can swap the rig on a lane pair in under 10. I'd bet a house like mine could go from zero to running in an hour or two.
We're calling it OpenLaneLink, and I plan to open source the hardware, firmware, and software stack when it's ready. Bowling is fun, and I want to help keep it affordable for alleys like mine.\nI hate vendor lock-in. I'm not a fan of closed systems, calling support for every hiccup, or paying to "white label" my own equipment. Want to go Tron-themed for a night? Good luck finding a neon neumorphic theme in something bought at the turn of the century.
All that bugged me. Sure, bowling equipment is niche, but the open hardware and software landscape is amazing.
Thanks for reading! Let me know if anyone's interested in more posts about this bowling nonsense."},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: I replaced a $120k bowling center system with $1,600 in ESP32s"}},"_tags":["story","author_section33","story_48968606","show_hn"],"author":"section33","children":[48968835,48968993,48969421,48970367,48970410,48970474,48970522,48970570,48970584,48970586,48970593,48970621,48970688,48970750,48970833,48970842,48970852,48970889,48970896,48970902,48970977,48971008,48971117,48971134,48971208,48971244,48971255,48971258,48971264,48971269,48971271,48971294,48971295,48971297,48971389,48971447,48971487,48971503,48971504,48971576,48971638,48971673,48971690,48971752,48971770,48971822,48971848,48972008,48972079,48972112,48972118,48972160,48972185,48972194,48972205,48972328,48972353,48972359,48972493,48972498,48972617,48972677,48972706,48972737,48972743,48972765,48972779,48972791,48972835,48972862,48972901,48973099,48973125,48973236,48973270,48973328,48973345,48973385,48973409,48973450,48973558,48973784,48974089,48974119,48974121,48974251,48974322,48974443,48974613,48974670,48974763,48974797,48974816,48974951,48975046,48975118,48975439,48975569,48975758,48975798,48975989,48976022,48976034,48976037,48976115,48976193,48976253,48976937,48977045,48977163,48977359,48977820,48978132,48978189,48978219,48978393,48978529,48978583,48979244,48979369,48979460,48979547,48979730,48979865,48980020,48982507,48983896,48983994,48985518,48992390,48997484,49000593,49038230,49068598,49113562],"created_at":"2026-07-19T14:41:16Z","created_at_i":1784472076,"num_comments":359,"objectID":"48968606","points":2935,"story_id":48968606,"story_text":"I might be the only SRE on Earth with his own bowling center. It's a more in-depth gig than you'd think.
My family and I bought an abandoned 8-lane bowling center in the rural mid-west. In our small town there weren't many recreation options for families. You've heard of a food desert? This is an R&R desert.
It had been abandoned for a good reason. The roof leaks, the electrical system was constantly surging, and my 70-year-old bowling equipment (still) doesn't work perfectly. The system that keeps your score is particularly interesting to me. It's the thing you watch during your game, but it fades into the background beyond that. Turns out these things are really cool, but absurdly expensive.
Ours was installed in 2008 and cost six figures. It's calculating ball speed and trajectory, camera-based pin detection (object detection and trig, on ICs!), runs the fouling, the animations, the pinsetting machine and ball return. Very cool stuff for its age.
From the business perspective, my facility only cost me $105k. To forklift-replace the score keeping system runs anywhere between $80-$120k, depending on features, vendor, and unit age. No upgrades or service contracts, mind you, and every feature and customization is a new line item. That's for a 1:1 replacement on a system installed in 2008. Incredible, given how fast the tech world moves.
Replacement parts cost a shocking $4000 per pair of lanes. But wait, the bowling machines themselves are 70 years old, so what's this "advanced" system actually doing back there? Actuating a single relay to trigger that big old machine. Everything else is strictly mechanical. Meanwhile I've got a six-figure invoice in my hand. I'm upset.
Given the state of open hardware, computer vision, real-time event streaming, and open source running megascale products worldwide, there had to be a way to do this myself.
So far I've built an equivalent prototype for about $200 per lane-pair, $400 if you're fancy. ESP32 and ESPNow with an RS485 fallback, reporting to a raspberry pi lane computer that's really just redis and a state machine bolted to an ESP32 gateway for the mesh.
Since it's all ESP32, I've got a fistful of spare controllers in a drawer, pre-flashed or waiting to be. All common off-the-shelf hardware: microcontrollers wired to relays, optocouplers, and IR-break-beam sensors, each running slightly different firmware. Writing the firmware and protocol is the actual hard part.
It's an ESPNow star-topology mesh: each node emits events from its sensors and accepts commands for its controls, reporting to a gateway node connected to the raspi over UART. From there it's event streaming: RX packets get translated and tossed into redis, commands relay back out to the mesh as needed. RS485 sits underneath as a wired fallback for noisy RF environments.
Once the data's in redis, it's familiar middleware/React/websocket/pub-sub stuff. Any React dev can build their own UI and bowling animations. Since it all runs on commodity hardware, I can do legit anything I want as the proprietor, and I own all my data. Repairs take five minutes; I can swap the rig on a lane pair in under 10. I'd bet a house like mine could go from zero to running in an hour or two.
We're calling it OpenLaneLink, and I plan to open source the hardware, firmware, and software stack when it's ready. Bowling is fun, and I want to help keep it affordable for alleys like mine.\nI hate vendor lock-in. I'm not a fan of closed systems, calling support for every hiccup, or paying to "white label" my own equipment. Want to go Tron-themed for a night? Good luck finding a neon neumorphic theme in something bought at the turn of the century.
All that bugged me. Sure, bowling equipment is niche, but the open hardware and software landscape is amazing.
Thanks for reading! Let me know if anyone's interested in more posts about this bowling nonsense.","title":"Show HN: I replaced a $120k bowling center system with $1,600 in ESP32s","updated_at":"2026-09-04T18:26:22Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"AntonioLi"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["thunder","compute"],"value":"Hey HN, I\u2019m Antonio from Nori Robotics (https://norirobotics.com). We build a $1,688 bimanual mobile robot in San Francisco for robotics developers and researchers.
I started working on Nori while doing robotics research at Columbia. I was teaching robots through human demonstrations, but getting my hands on affordable hardware was difficult. Most labs have one or two expensive robots, which makes it hard to collect large datasets, run long experiments, or test across several robots.
So I built my own. After seven iterations the latest Nori has:
* 19 degrees of freedom
* Two 7+1 DOF arms with a 1.5 kg payload per arm
* A 55 kg telescoping lift
* A differential wheeled base
* Four 720p, 30 fps RGB cameras
* 2D lidar
* A dual microphone array with full-duplex voice communication
* A 432 Wh battery
* A Raspberry Pi 5 with 4 GB RAM (SLAM and safeties are run on board, heavier ACT and VLAs must be run from a computer via LAN or a server via WAN)
Getting this under $2,000 was the main engineering challenge. Nori has more than 100 moving and structural parts, so costs add up quickly across actuators, bearings, wiring, power delivery, and assembly. Some main choices we made to get the cost low was using high-ratio servos instead of QDD motors, and using a wheel base instead of legs.
We assemble each robot in San Francisco and have designed it to be easy to manufacture and repair (we offer 3D files to print repairs).
Our open SDK includes teleoperation and demonstration tools: https://github.com/Nori-Robotics/nori-sdk-py
We also built a browser-based simulator so you can try it out: https://lab.norirobotics.com/nori/model
We\u2019ve shipped our first robot and are building the next batch. Eventually, we want people without robotics experience to teach Nori tasks and share them with other owners.
Currently the hardware is already capable of basic cleaning tasks, opening drawers, restocking shelves and pouring beers. Here is a video of Nori doing things: https://youtube.com/shorts/VRfVXHfQvD8
We make money by selling the hardware for $1,688, with optional paid software on top. Parts of hardware are open source. More details are in our hardware paper: https://doi.org/10.48550/arXiv.2605.16537
If you work in robotics, what would you build with a robot at this price? What would you change about the hardware?"},"title":{"matchLevel":"none","matchedWords":[],"value":"Launch HN: Nori Robotics (YC S26) \u2013 A low-cost humanoid robot for development"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://www.norirobotics.com/"}},"_tags":["story","author_AntonioLi","story_49525153","launch_hn"],"author":"AntonioLi","children":[49525480,49525494,49525592,49525758,49525794,49526048,49526136,49526386,49526474,49526771,49526793,49526819,49526980,49527044,49527108,49527169,49527196,49527372,49527792,49527804,49527835,49527852,49528104,49528345,49528862,49529412,49530122,49531144,49533351,49533464,49533470,49533688,49534364,49534509,49538649,49547831],"created_at":"2026-09-01T17:35:10Z","created_at_i":1788284110,"num_comments":66,"objectID":"49525153","points":200,"story_id":49525153,"story_text":"Hey HN, I\u2019m Antonio from Nori Robotics (https://norirobotics.com). We build a $1,688 bimanual mobile robot in San Francisco for robotics developers and researchers.
I started working on Nori while doing robotics research at Columbia. I was teaching robots through human demonstrations, but getting my hands on affordable hardware was difficult. Most labs have one or two expensive robots, which makes it hard to collect large datasets, run long experiments, or test across several robots.
So I built my own. After seven iterations the latest Nori has:
* 19 degrees of freedom
* Two 7+1 DOF arms with a 1.5 kg payload per arm
* A 55 kg telescoping lift
* A differential wheeled base
* Four 720p, 30 fps RGB cameras
* 2D lidar
* A dual microphone array with full-duplex voice communication
* A 432 Wh battery
* A Raspberry Pi 5 with 4 GB RAM (SLAM and safeties are run on board, heavier ACT and VLAs must be run from a computer via LAN or a server via WAN)
Getting this under $2,000 was the main engineering challenge. Nori has more than 100 moving and structural parts, so costs add up quickly across actuators, bearings, wiring, power delivery, and assembly. Some main choices we made to get the cost low was using high-ratio servos instead of QDD motors, and using a wheel base instead of legs.
We assemble each robot in San Francisco and have designed it to be easy to manufacture and repair (we offer 3D files to print repairs).
Our open SDK includes teleoperation and demonstration tools: https://github.com/Nori-Robotics/nori-sdk-py
We also built a browser-based simulator so you can try it out: https://lab.norirobotics.com/nori/model
We\u2019ve shipped our first robot and are building the next batch. Eventually, we want people without robotics experience to teach Nori tasks and share them with other owners.
Currently the hardware is already capable of basic cleaning tasks, opening drawers, restocking shelves and pouring beers. Here is a video of Nori doing things: https://youtube.com/shorts/VRfVXHfQvD8
We make money by selling the hardware for $1,688, with optional paid software on top. Parts of hardware are open source. More details are in our hardware paper: https://doi.org/10.48550/arXiv.2605.16537
If you work in robotics, what would you build with a robot at this price? What would you change about the hardware?","title":"Launch HN: Nori Robotics (YC S26) \u2013 A low-cost humanoid robot for development","updated_at":"2026-09-04T18:13:20Z","url":"https://www.norirobotics.com/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"anshchokshi"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["thunder","compute"],"value":"Hi HN, I'm Ansh, founder of Mireye (https://www.mireye.com). I'm building the infrastructure AI agents use to make decisions about physical places: data, enrichment, tools, and signals for any US location, behind one API and MCP server.
Here's a demo video: https://www.youtube.com/watch?v=haqO6UbUqU0
To try it, paste https://www.mireye.com/skills.md into your agent, and grab a free key at mireye.com (5,000 credits, no card). Docs are at https://docs.mireye.ai. The fastest way in is to ask the API any question about any US location. Test it on a place you know, cold and grade it against what you know.
Before Mireye I was building construction agents and hit this wall myself: my agent could reason about anything online but knew nothing about the ground under it. Then a Fortune 500 insurer told me their engineers had given up on underwriting agents for the same reason. Frontier models keep hallucinating when asked a specific question about a specific place.
My first product was a niche site-screening app. Customers tested it on places they knew and the answers held up, but nobody cared about the app. They wanted the engine underneath. Usage agreed: 311 of the 317 fields in the catalog get queried and no use case dominates. So I killed the app and started building the infrastructure instead.
Mireye is not a dataset with an API on top, because facts alone are not a decision. An agent runs the whole job through it: cited facts, a bare address enriched into owner, acreage, structures, and nearby power, tools for the operations models get wrong, and signals when something changes, like a rezoning filing. Data, enrichment, tools, signals.
The tools came from watching agents fail. We build agents on our own infra, and the same things kept breaking: an agent would eyeball a distance instead of computing it, grab the wrong parcel for an address, or burn its whole budget halfway through a batch. Each failure became something an agent can call: deterministic geometry and drive-time tools, parcel resolution, a quote endpoint that prices a job before it runs, and skills that package whole workflows like screening a site or underwriting an address.
The hard part surprised me. Every source has to be gathered (sometimes county by county, in whatever format each county publishes), normalized into one schema, contracted (where it comes from, what each value means, how often it refreshes), and then kept fresh forever. We run that loop for 366 fields today, served multi-tenant from one index, and every source fought back differently. Maryland publishes a dataset literally titled "Hidden Property Owner Names." I filed public records requests in North Carolina because nobody indexes sewer mains.
The deeper problem is meaning. Two counties publish a field with the same name and it means different things. And the most dangerous value is null. Does it mean "no flood zone here" or "this county never mapped floods"? Put a model in front of that gap and it fills the silence with a plausible number. So we type absence. Every field returns ok, absent, or failed. Customers have told me the refusals are why they trust it.
The newest piece is on-demand indexing. Ask for a field we don't have and a long-running agent researches sources, collects the data, tests it against ground truth, and indexes it, usually within a day. I'll take a few field requests from this thread and report back with what it built.
People have built things I didn't plan for: insurance teams screening portfolios for flood, wind, and wildfire, a proptech cleaning messy listing addresses through enrichment, a wellness brand scoring street corners for poster spots, a robotics company sourcing warehouses, data center site selection, drone deployment planning, school bus routes for a city, signals for human trafficking investigations. The mission is to index every inch of the earth and make it as queryable as the web.
Pricing is public: the free tier is 5,000 credits a month, $19/month for 25,000, $99/month for 120,000, custom for enterprise.
I'd love to hear where the agents you're building touch the physical world. And I'd love it more if you run Mireye on a place you know and tell me what we got wrong."},"title":{"matchLevel":"none","matchedWords":[],"value":"Launch HN: Mireye (YC S26) \u2013 Infrastructure for Physical World AI Agents"}},"_tags":["story","author_anshchokshi","story_49552616","launch_hn"],"author":"anshchokshi","children":[49556083,49557110,49557575,49561935,49565464,49569834],"created_at":"2026-09-03T16:24:13Z","created_at_i":1788452653,"num_comments":8,"objectID":"49552616","points":28,"story_id":49552616,"story_text":"Hi HN, I'm Ansh, founder of Mireye (https://www.mireye.com). I'm building the infrastructure AI agents use to make decisions about physical places: data, enrichment, tools, and signals for any US location, behind one API and MCP server.
Here's a demo video: https://www.youtube.com/watch?v=haqO6UbUqU0
To try it, paste https://www.mireye.com/skills.md into your agent, and grab a free key at mireye.com (5,000 credits, no card). Docs are at https://docs.mireye.ai. The fastest way in is to ask the API any question about any US location. Test it on a place you know, cold and grade it against what you know.
Before Mireye I was building construction agents and hit this wall myself: my agent could reason about anything online but knew nothing about the ground under it. Then a Fortune 500 insurer told me their engineers had given up on underwriting agents for the same reason. Frontier models keep hallucinating when asked a specific question about a specific place.
My first product was a niche site-screening app. Customers tested it on places they knew and the answers held up, but nobody cared about the app. They wanted the engine underneath. Usage agreed: 311 of the 317 fields in the catalog get queried and no use case dominates. So I killed the app and started building the infrastructure instead.
Mireye is not a dataset with an API on top, because facts alone are not a decision. An agent runs the whole job through it: cited facts, a bare address enriched into owner, acreage, structures, and nearby power, tools for the operations models get wrong, and signals when something changes, like a rezoning filing. Data, enrichment, tools, signals.
The tools came from watching agents fail. We build agents on our own infra, and the same things kept breaking: an agent would eyeball a distance instead of computing it, grab the wrong parcel for an address, or burn its whole budget halfway through a batch. Each failure became something an agent can call: deterministic geometry and drive-time tools, parcel resolution, a quote endpoint that prices a job before it runs, and skills that package whole workflows like screening a site or underwriting an address.
The hard part surprised me. Every source has to be gathered (sometimes county by county, in whatever format each county publishes), normalized into one schema, contracted (where it comes from, what each value means, how often it refreshes), and then kept fresh forever. We run that loop for 366 fields today, served multi-tenant from one index, and every source fought back differently. Maryland publishes a dataset literally titled "Hidden Property Owner Names." I filed public records requests in North Carolina because nobody indexes sewer mains.
The deeper problem is meaning. Two counties publish a field with the same name and it means different things. And the most dangerous value is null. Does it mean "no flood zone here" or "this county never mapped floods"? Put a model in front of that gap and it fills the silence with a plausible number. So we type absence. Every field returns ok, absent, or failed. Customers have told me the refusals are why they trust it.
The newest piece is on-demand indexing. Ask for a field we don't have and a long-running agent researches sources, collects the data, tests it against ground truth, and indexes it, usually within a day. I'll take a few field requests from this thread and report back with what it built.
People have built things I didn't plan for: insurance teams screening portfolios for flood, wind, and wildfire, a proptech cleaning messy listing addresses through enrichment, a wellness brand scoring street corners for poster spots, a robotics company sourcing warehouses, data center site selection, drone deployment planning, school bus routes for a city, signals for human trafficking investigations. The mission is to index every inch of the earth and make it as queryable as the web.
Pricing is public: the free tier is 5,000 credits a month, $19/month for 25,000, $99/month for 120,000, custom for enterprise.
I'd love to hear where the agents you're building touch the physical world. And I'd love it more if you run Mireye on a place you know and tell me what we got wrong.","title":"Launch HN: Mireye (YC S26) \u2013 Infrastructure for Physical World AI Agents","updated_at":"2026-09-04T21:01:41Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"pain_perdu"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["thunder","compute"],"value":"Today I received a match through YC\u2019s Co-Founder Matching Platform. The user claimed to be an experienced British technologist named \u201cDave J.\u201d
Immediately after matching, they sent me a WhatsApp link and asked to connect there. They then said they were \u201chaving a procedure\u201d and were unable to voice or video call.
A reverse image search of their YC profile photo led to the website and GitHub of someone named \u201cDavid W.\u201d When I asked the person I was speaking with for their GitHub, however, they provided a completely different account filled with infosec-related projects.
There were several other red flags:
\u2022 Their command of English was poor despite claiming to be British.\n\u2022 They claimed not to have LinkedIn, while the real David W., whose photo appeared to have been used, has an active LinkedIn profile.\n\u2022 They were unwilling or unable to verify their identity by voice or video.
I didn\u2019t have the time or interest to let the interaction play out beyond a few minutes, but please be careful on the Co-Founder Matching platform. There appear to be people using fabricated identities and potentially stolen photos to initiate conversations and quickly move them off-platform.
This also isn\u2019t the first suspicious interaction I\u2019ve encountered there. On a previous occasion, someone wanted me to apply for jobs on their behalf using my identity or post to HN for them.
I\u2019ve reported this one to YC."},"title":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["thunder"],"value":"Beware: Malicious Profiles on YC Co-Founder Matching"}},"_tags":["story","author_pain_perdu","story_49552237","ask_hn"],"author":"pain_perdu","children":[49552517],"created_at":"2026-09-03T16:02:14Z","created_at_i":1788451334,"num_comments":0,"objectID":"49552237","points":5,"story_id":49552237,"story_text":"Today I received a match through YC\u2019s Co-Founder Matching Platform. The user claimed to be an experienced British technologist named \u201cDave J.\u201d
Immediately after matching, they sent me a WhatsApp link and asked to connect there. They then said they were \u201chaving a procedure\u201d and were unable to voice or video call.
A reverse image search of their YC profile photo led to the website and GitHub of someone named \u201cDavid W.\u201d When I asked the person I was speaking with for their GitHub, however, they provided a completely different account filled with infosec-related projects.
There were several other red flags:
\u2022 Their command of English was poor despite claiming to be British.\n\u2022 They claimed not to have LinkedIn, while the real David W., whose photo appeared to have been used, has an active LinkedIn profile.\n\u2022 They were unwilling or unable to verify their identity by voice or video.
I didn\u2019t have the time or interest to let the interaction play out beyond a few minutes, but please be careful on the Co-Founder Matching platform. There appear to be people using fabricated identities and potentially stolen photos to initiate conversations and quickly move them off-platform.
This also isn\u2019t the first suspicious interaction I\u2019ve encountered there. On a previous occasion, someone wanted me to apply for jobs on their behalf using my identity or post to HN for them.
I\u2019ve reported this one to YC.","title":"Beware: Malicious Profiles on YC Co-Founder Matching","updated_at":"2026-09-03T23:38:48Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"Thrashed"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["thunder","compute"],"value":"I\u2019ve been a NameCheap customer for 13 years. I\u2019ve also helped out an old college club paying for a .com they use (that is registered to me under my name, address, and phone number). During a recent leadership transition, the incoming club lead wanted to make changes to the DNS and didn\u2019t know to contact me. They figured out the domain name was parked at NameCheap, so they initiated a password reset using the domain name. I got a password reset email and immediately filed a NameCheap support ticket saying \u201cI did not initiate this\u201d. They called me to verify I was the one who filed the ticket, and then followed up with a canned email with tips like check your anti-virus.
The incoming club leader was persistent though, and called NameCheap support. He convinced them the domain registered in my name and address really belonged to his club, and with no verification or validation whatsoever, NameCheap changed my password, and changed the email address associated with my account. All because someone simply asked nicely on a phone call.
Meanwhile in the background, someone advised the new club leader who I was and we were able to connect and get things transferred over. Ultimately I was happy to give them access or even ownership if they wanted (student club turnover being what it is, it\u2019s likely a domain doesn\u2019t get renewed and gets gobbled up by a squatter, which is why I was keeping it current for them).
But NameCheap had no way of knowing any of this. As far as NameCheap was aware, this was a personal account of mine. They demonstrated they were perfectly able to pick up a phone and call me (to verify my initial support ticket) but when someone calls them and says \u201cbut I really want access to that account\u201d they don\u2019t bother?
I\u2019d hesitate to even call this social engineering. It\u2019s clearly a massive vulnerability. I\u2019ve already moved a dozen of my most critical domains out of NameCheap after seeing just how easy it is for a third party to completely take over a NameCheap account: just ask nicely."},"title":{"matchLevel":"none","matchedWords":[],"value":"Tell HN: Namecheap gave my account to an unverified third party"}},"_tags":["story","author_Thrashed","story_49028037","ask_hn"],"author":"Thrashed","children":[49028360,49028427,49028511,49028523,49028570,49028639,49028692,49028703,49028754,49028778,49028872,49028881,49028894,49028910,49028935,49028969,49029006,49029024,49029026,49029048,49029076,49029124,49029174,49029194,49029333,49029556,49029643,49029692,49029749,49029880,49029903,49030180,49031010,49031692,49037478,49037872,49061666,49062102,49067182,49077890,49082194],"created_at":"2026-07-23T21:05:29Z","created_at_i":1784840729,"num_comments":177,"objectID":"49028037","points":498,"story_id":49028037,"story_text":"I\u2019ve been a NameCheap customer for 13 years. I\u2019ve also helped out an old college club paying for a .com they use (that is registered to me under my name, address, and phone number). During a recent leadership transition, the incoming club lead wanted to make changes to the DNS and didn\u2019t know to contact me. They figured out the domain name was parked at NameCheap, so they initiated a password reset using the domain name. I got a password reset email and immediately filed a NameCheap support ticket saying \u201cI did not initiate this\u201d. They called me to verify I was the one who filed the ticket, and then followed up with a canned email with tips like check your anti-virus.
The incoming club leader was persistent though, and called NameCheap support. He convinced them the domain registered in my name and address really belonged to his club, and with no verification or validation whatsoever, NameCheap changed my password, and changed the email address associated with my account. All because someone simply asked nicely on a phone call.
Meanwhile in the background, someone advised the new club leader who I was and we were able to connect and get things transferred over. Ultimately I was happy to give them access or even ownership if they wanted (student club turnover being what it is, it\u2019s likely a domain doesn\u2019t get renewed and gets gobbled up by a squatter, which is why I was keeping it current for them).
But NameCheap had no way of knowing any of this. As far as NameCheap was aware, this was a personal account of mine. They demonstrated they were perfectly able to pick up a phone and call me (to verify my initial support ticket) but when someone calls them and says \u201cbut I really want access to that account\u201d they don\u2019t bother?
I\u2019d hesitate to even call this social engineering. It\u2019s clearly a massive vulnerability. I\u2019ve already moved a dozen of my most critical domains out of NameCheap after seeing just how easy it is for a third party to completely take over a NameCheap account: just ask nicely.","title":"Tell HN: Namecheap gave my account to an unverified third party","updated_at":"2026-09-04T18:26:22Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"SilenN"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["thunder","compute"],"value":"Hi HN, we built an open source model gateway. It's a single place to manage our own self hosted, frontier, and open source models in one place.
It\u2019s is rust native, built for concurrency, and implements all the config quirks across models and providers (streaming formats, tool calls, model parameters, rate limits, and different error behavior).
The gateway adds under 1 ms for BYOK requests and under 2 ms when Experiential supplies the provider key. It has every major inference provider, and 1000+ models refreshed daily via a codex agent that opens a PR.
Compared to other similar projects we\u2019re open source, take no markup, allow you to mix local models with a marketplace, and use your traffic to (opt in) train you a model. Simple routing doesn\u2019t warrant a 10% token markup.
The way we do this is given standardized OTel traces, we mine representative real tasks, use text world models to simulate rollouts for various models, apply an LLM judge, and fit a nearest neighbor classifier on top of an embedding of a prompt to decide the optimal model for each request. Usually this can map out a better pareto curve on cost/quality than just calling single models but it\u2019s not perfect.
Using these simulations we can also do things like suggesting cache hit optimizations, new model suggestions, and training models.
It\u2019s open source, so you can deploy it on your own infrastructure, use our hosted version with 0 markup, or read how we design for maximum availability on our website."},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: We built open OpenRouter that turns usage into a better model"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://github.com/experientiallabs/experiential"}},"_tags":["story","author_SilenN","story_49471407","show_hn"],"author":"SilenN","children":[49471977,49471998,49472268,49472738,49473165,49473195,49473371,49473410,49473874,49474132,49474138,49474604,49475767,49476057,49476194,49476304,49476654,49477102,49478018,49478084,49482736,49484242,49485872,49503144,49527839,49540041],"created_at":"2026-08-27T21:18:35Z","created_at_i":1787865515,"num_comments":47,"objectID":"49471407","points":222,"story_id":49471407,"story_text":"Hi HN, we built an open source model gateway. It's a single place to manage our own self hosted, frontier, and open source models in one place.
It\u2019s is rust native, built for concurrency, and implements all the config quirks across models and providers (streaming formats, tool calls, model parameters, rate limits, and different error behavior).
The gateway adds under 1 ms for BYOK requests and under 2 ms when Experiential supplies the provider key. It has every major inference provider, and 1000+ models refreshed daily via a codex agent that opens a PR.
Compared to other similar projects we\u2019re open source, take no markup, allow you to mix local models with a marketplace, and use your traffic to (opt in) train you a model. Simple routing doesn\u2019t warrant a 10% token markup.
The way we do this is given standardized OTel traces, we mine representative real tasks, use text world models to simulate rollouts for various models, apply an LLM judge, and fit a nearest neighbor classifier on top of an embedding of a prompt to decide the optimal model for each request. Usually this can map out a better pareto curve on cost/quality than just calling single models but it\u2019s not perfect.
Using these simulations we can also do things like suggesting cache hit optimizations, new model suggestions, and training models.
It\u2019s open source, so you can deploy it on your own infrastructure, use our hosted version with 0 markup, or read how we design for maximum availability on our website.","title":"Show HN: We built open OpenRouter that turns usage into a better model","updated_at":"2026-09-04T20:55:05Z","url":"https://github.com/experientiallabs/experiential"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"Daniel-Pan"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["thunder","compute"],"value":"Duration May - Aug 2026
My Background:\nI am a founder of multiple SaaS company. Most recently i am building a AI native ticketing system for support platforms, simply put it takes your issues coming from Github, Discord, Forum or email and have a AI agent run through all the sources and generate a reply that is posted back.\nPricing is tier based with a freemium model, direct signup, without demo.
Marketing Strategy for X:\nWe started marketing using text and image based post, with our website link put as CTA or the blog (medium) link.
Total spending:\nWe spend Around $3,509 for the entire 4 months.
Audience:\nWe narrowed down our audience , starting from targeting globally to focusing on Tier 1 markets especially as they have normally have higher rate of subscription buying.
Results according to X own numbers:\n46M impressions and 26K+ link clicks.
Compare with our own website tracker (Matomo):\nTwitter \n613 visits\n885 actions\n1.4 actions per visit \n50s avg. time on website\n74% bounce rate
Actual Signups:\nConsider it near zero, no actual follow ups, or growth
My observation:\nFor each post there are crypto bots, and random bots that engage with the post, resulting in fake engagement and clicks. \nThis platform strategy bring little or no value for our business."},"title":{"matchLevel":"none","matchedWords":[],"value":"We ran X (Twitter) ads for 4 months and here are the results"}},"_tags":["story","author_Daniel-Pan","story_49475011","ask_hn"],"author":"Daniel-Pan","children":[49475290,49505218,49532902],"created_at":"2026-08-28T06:23:28Z","created_at_i":1787898208,"num_comments":4,"objectID":"49475011","points":10,"story_id":49475011,"story_text":"Duration May - Aug 2026
My Background:\nI am a founder of multiple SaaS company. Most recently i am building a AI native ticketing system for support platforms, simply put it takes your issues coming from Github, Discord, Forum or email and have a AI agent run through all the sources and generate a reply that is posted back.\nPricing is tier based with a freemium model, direct signup, without demo.
Marketing Strategy for X:\nWe started marketing using text and image based post, with our website link put as CTA or the blog (medium) link.
Total spending:\nWe spend Around $3,509 for the entire 4 months.
Audience:\nWe narrowed down our audience , starting from targeting globally to focusing on Tier 1 markets especially as they have normally have higher rate of subscription buying.
Results according to X own numbers:\n46M impressions and 26K+ link clicks.
Compare with our own website tracker (Matomo):\nTwitter \n613 visits\n885 actions\n1.4 actions per visit \n50s avg. time on website\n74% bounce rate
Actual Signups:\nConsider it near zero, no actual follow ups, or growth
My observation:\nFor each post there are crypto bots, and random bots that engage with the post, resulting in fake engagement and clicks. \nThis platform strategy bring little or no value for our business.","title":"We ran X (Twitter) ads for 4 months and here are the results","updated_at":"2026-09-04T05:01:48Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"nkohari"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["thunder","compute"],"value":"I\u2019m Nate, the founder of Ardent. We just shipped our public beta, and we\u2019d love your thoughts!
Ardent is an agent running in a desktop (Electron) app built to help with knowledge work, designed for less-technical people outside engineering. Think Codex or Claude Cowork, but built around collaboration and customization.
I know, I know, it\u2019s yet another agent harness! Ardent is a little different \u2013 it leans heavily on codegen to solve problems. Most agents are basically just a bag of tools and a while loop, but Ardent writes code to complete tasks, making it much faster and more token-efficient. [0]
Lots of companies are trying to build their own agent harness right now, because token budgets make software appear cheap, and it seems like a competitive advantage to own a custom agent.
The thing is, agents are deceptively complex. They aren\u2019t light wrappers around the model anymore; the harness makes a big difference in how capable the agent is. They also come with lots of sticky problems: context management, sandboxing and permissions, authz/authn to remote services, etc.
Instead of building your own harness from scratch, you can customize Ardent using abilities: TypeScript libraries that can expose tools and prompts which can be used by the agent. They\u2019re like agent skills, but closer to regular code libraries: type-safe, importable, and with well-defined permissions. Ardent can help you build abilities, but you can also create them by hand if you prefer.
Anyone can share abilities with everyone in their organization with a single click, and if Ardent notices that the user is doing repeatable work, it offers to transform the adhoc solution into a reusable ability. Say you prepare the same project update every Friday, pulling issues from Linear and docs from Google Drive. Ardent can turn that workflow into a reusable ability that you can share with everyone in your company.
Your company\u2019s ability catalog is like a standard library. To solve a problem, Ardent searches for existing abilities, and then writes glue code to compose them. Because the code is ephemeral, it doesn\u2019t need to be maintainable, which lets us use cheaper, less-capable open weight models like Kimi and GLM. [1]
Code executes on the user\u2019s machine in a Deno sandbox with zero permissions by default. Abilities declare required permissions up front, and any code that exceeds the current scope gets terminated and prompts the user for approval.
Ardent can also connect to a bunch of tools which you\u2019re already using, like Google, Linear, and so on. We support MCP servers, but in our experience, MCP is slow, token-expensive, and provides a pretty bad experience for both agents and users. [2]
Instead, Ardent connectors handle auth, but instead of using MCP, we generate client shims which call the service\u2019s primary API. This is a much better fit for Ardent\u2019s codegen approach, because the agent can read the client and treat it like any other library.
There\u2019s a bunch of other interesting stuff in Ardent: memory, conversational recall, scheduled tasks, prompt suggestions, and so on. During our beta period, we\u2019re selling inference at-cost, and we\u2019re giving each organization some free credits to experiment.
Please take a look and tell us what you think! We\u2019d love your feedback, whether you think what we\u2019re doing is good, bad, or meh. I\u2019m also happy to answer any questions that you might have.
---
[0] Cloudflare has taken to calling this \u201ccode mode\u201d: https://blog.cloudflare.com/code-mode/
[1] In our benchmarks, Ardent consumed 63% fewer tokens than Opus running in Claude Cowork, with half the model requests.
[2] MCP servers rarely define input/output contracts, and it\u2019s hard to write code that calls a function when you don\u2019t know its arguments, or even its arity! Here's a good read on the subject: https://www.blacksmith.sh/blog/code-smith-code-mode"},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: Ardent, a code-first agent for non-engineering work"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://ardent.ai/"}},"_tags":["story","author_nkohari","story_49550931","show_hn"],"author":"nkohari","children":[49551403],"created_at":"2026-09-03T15:01:24Z","created_at_i":1788447684,"num_comments":2,"objectID":"49550931","points":10,"story_id":49550931,"story_text":"I\u2019m Nate, the founder of Ardent. We just shipped our public beta, and we\u2019d love your thoughts!
Ardent is an agent running in a desktop (Electron) app built to help with knowledge work, designed for less-technical people outside engineering. Think Codex or Claude Cowork, but built around collaboration and customization.
I know, I know, it\u2019s yet another agent harness! Ardent is a little different \u2013 it leans heavily on codegen to solve problems. Most agents are basically just a bag of tools and a while loop, but Ardent writes code to complete tasks, making it much faster and more token-efficient. [0]
Lots of companies are trying to build their own agent harness right now, because token budgets make software appear cheap, and it seems like a competitive advantage to own a custom agent.
The thing is, agents are deceptively complex. They aren\u2019t light wrappers around the model anymore; the harness makes a big difference in how capable the agent is. They also come with lots of sticky problems: context management, sandboxing and permissions, authz/authn to remote services, etc.
Instead of building your own harness from scratch, you can customize Ardent using abilities: TypeScript libraries that can expose tools and prompts which can be used by the agent. They\u2019re like agent skills, but closer to regular code libraries: type-safe, importable, and with well-defined permissions. Ardent can help you build abilities, but you can also create them by hand if you prefer.
Anyone can share abilities with everyone in their organization with a single click, and if Ardent notices that the user is doing repeatable work, it offers to transform the adhoc solution into a reusable ability. Say you prepare the same project update every Friday, pulling issues from Linear and docs from Google Drive. Ardent can turn that workflow into a reusable ability that you can share with everyone in your company.
Your company\u2019s ability catalog is like a standard library. To solve a problem, Ardent searches for existing abilities, and then writes glue code to compose them. Because the code is ephemeral, it doesn\u2019t need to be maintainable, which lets us use cheaper, less-capable open weight models like Kimi and GLM. [1]
Code executes on the user\u2019s machine in a Deno sandbox with zero permissions by default. Abilities declare required permissions up front, and any code that exceeds the current scope gets terminated and prompts the user for approval.
Ardent can also connect to a bunch of tools which you\u2019re already using, like Google, Linear, and so on. We support MCP servers, but in our experience, MCP is slow, token-expensive, and provides a pretty bad experience for both agents and users. [2]
Instead, Ardent connectors handle auth, but instead of using MCP, we generate client shims which call the service\u2019s primary API. This is a much better fit for Ardent\u2019s codegen approach, because the agent can read the client and treat it like any other library.
There\u2019s a bunch of other interesting stuff in Ardent: memory, conversational recall, scheduled tasks, prompt suggestions, and so on. During our beta period, we\u2019re selling inference at-cost, and we\u2019re giving each organization some free credits to experiment.
Please take a look and tell us what you think! We\u2019d love your feedback, whether you think what we\u2019re doing is good, bad, or meh. I\u2019m also happy to answer any questions that you might have.
---
[0] Cloudflare has taken to calling this \u201ccode mode\u201d: https://blog.cloudflare.com/code-mode/
[1] In our benchmarks, Ardent consumed 63% fewer tokens than Opus running in Claude Cowork, with half the model requests.
[2] MCP servers rarely define input/output contracts, and it\u2019s hard to write code that calls a function when you don\u2019t know its arguments, or even its arity! Here's a good read on the subject: https://www.blacksmith.sh/blog/code-smith-code-mode","title":"Show HN: Ardent, a code-first agent for non-engineering work","updated_at":"2026-09-04T04:11:18Z","url":"https://ardent.ai/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"dr_blueberry"},"story_text":{"matchLevel":"none","matchedWords":[],"value":"Hey HN. We built an openai-compatible API for running open-weight VLMs, OCR VLMs and ViT-based vision models.
The motivation was mostly frustration when running these models in production and discovering all the details around serving VLMs, especially around visual accuracy.
A few footguns we kept running into:
- quantized models served under the same name (this one still drives me nuts): providers often serve models with different quants, environments, vLLM/SGLang serving params with the same model-id. Vision is especially sensitive to this; some quants that look fine on text benchmarks noticeably hurt OCR/small-text/spatial accuracy.
- video performance is varied: when we tested with popular routers on video-native VLMs, more than 80% of providers didn't support video inputs, and even fewer let you control FPS. If you care for time-resolution in videos, none of these providers work even if the models themselves are capable of it.
- document inference is all pipelining: rasterizing PDFs, parallelizing page workers, retrying when pages fail inference, dealing with rate-limits, etc can get tricky quickly and takes substantial developer time.
- standardization making abstractions leaky: this is less about vision per-se, but generally for serving models with high-quality output assurances. context-limits, max resolution, FPS sampling, quants, GPU SM architecture, can all add variability to (vision) quality even if the model-id claims to be the same.
We wanted one place to run OCR models, VLMs and ViTs that we could confidently use for our own internal agents and evals. The gateway was born from this need internally, and now we're opening it up to the public - you can swap the model name to compare GLM-OCR, dots.mocr, PaddleOCR VL, Qwen3.8-27B, Gemma4-26B-A4B etc. We handle the serving/runtime/pipelining underneath, with the goal of giving high-quality visual intelligence.
Are there any other vision "footguns" people have run into? especially cases where "same model" across two providers gave materially different outputs.
Try different models on Gateway simply by updating the model name:
uvx vlmrun gw chat <doc>.pdf -m glm-ocr
uvx vlmrun gw chat <doc>.pdf -m deepseek-ocr-2
uvx vlmrun gw chat <doc>.pdf -m pp-ocrv6
uvx vlmrun gw chat <video>.mp4 -m qwen/qwen3.5-0.8b -p "describe the video""},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: Run open-weight OCR, VLM and vision models behind one API"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://www.vlmrun.com/gateway"}},"_tags":["story","author_dr_blueberry","story_49568379","show_hn"],"author":"dr_blueberry","children":[49568614],"created_at":"2026-09-04T18:31:25Z","created_at_i":1788546685,"num_comments":0,"objectID":"49568379","points":5,"story_id":49568379,"story_text":"Hey HN. We built an openai-compatible API for running open-weight VLMs, OCR VLMs and ViT-based vision models.
The motivation was mostly frustration when running these models in production and discovering all the details around serving VLMs, especially around visual accuracy.
A few footguns we kept running into:
- quantized models served under the same name (this one still drives me nuts): providers often serve models with different quants, environments, vLLM/SGLang serving params with the same model-id. Vision is especially sensitive to this; some quants that look fine on text benchmarks noticeably hurt OCR/small-text/spatial accuracy.
- video performance is varied: when we tested with popular routers on video-native VLMs, more than 80% of providers didn't support video inputs, and even fewer let you control FPS. If you care for time-resolution in videos, none of these providers work even if the models themselves are capable of it.
- document inference is all pipelining: rasterizing PDFs, parallelizing page workers, retrying when pages fail inference, dealing with rate-limits, etc can get tricky quickly and takes substantial developer time.
- standardization making abstractions leaky: this is less about vision per-se, but generally for serving models with high-quality output assurances. context-limits, max resolution, FPS sampling, quants, GPU SM architecture, can all add variability to (vision) quality even if the model-id claims to be the same.
We wanted one place to run OCR models, VLMs and ViTs that we could confidently use for our own internal agents and evals. The gateway was born from this need internally, and now we're opening it up to the public - you can swap the model name to compare GLM-OCR, dots.mocr, PaddleOCR VL, Qwen3.8-27B, Gemma4-26B-A4B etc. We handle the serving/runtime/pipelining underneath, with the goal of giving high-quality visual intelligence.
Are there any other vision "footguns" people have run into? especially cases where "same model" across two providers gave materially different outputs.
Try different models on Gateway simply by updating the model name:
uvx vlmrun gw chat <doc>.pdf -m glm-ocr
uvx vlmrun gw chat <doc>.pdf -m deepseek-ocr-2
uvx vlmrun gw chat <doc>.pdf -m pp-ocrv6
uvx vlmrun gw chat <video>.mp4 -m qwen/qwen3.5-0.8b -p "describe the video"","title":"Show HN: Run open-weight OCR, VLM and vision models behind one API","updated_at":"2026-09-04T18:53:50Z","url":"https://www.vlmrun.com/gateway"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"rezvovmobile"},"story_text":{"matchLevel":"none","matchedWords":[],"value":"Tim O'Reilly just published an essay arguing that AI is a medium, like photography or painting. His key line: "I judge the output, not the means used to produce it."
He describes his own process: a 300-word prompt becomes a 660-word draft, which he rewrites into a 3,000-word essay. Maybe 40-50 words are Claude's. He's proud of how he uses AI. He says we're still in the "filmed stage plays" era of AI writing, and the masters will emerge.
This resonates with me. I'm an engineer, not a writer. Over the past few years, I've been using AI extensively \u2014 for drafting articles, structuring complex technical ideas. The AI helps me articulate thoughts I couldn't express alone. But I always rewrite, edit, and make sure every word is mine.
Here's my dilemma: should I disclose that I used AI as a tool?
When I don't disclose, I feel dishonest \u2014 like I'm claiming a skill I don't fully have. When I do disclose, people dismiss the work as "AI slop" before reading it. I've had journalists ignore emails, investors go silent, and peers assume the ideas aren't really mine.
Tim argues against Ted Chiang's position that AI writing lacks "choices." Tim says a prompt can contain a lifetime of experience, like Leonard Cohen's "Hallelujah" taking five years or Dylan's "Blowing in the Wind" taking ten minutes \u2014 both are masterpieces.
But here's the practical question for this community:
If you're an engineer, founder, or technical person using AI to communicate ideas you couldn't express otherwise:
1. Do you disclose AI use? When and how?\n2. Have you faced backlash or dismissal when you did disclose?\n3. Do you think "judge the output, not the means" is realistic, or will AI-assisted work always carry a stigma?
I'm genuinely trying to figure out the right approach. Not to cheat, but to be honest without being penalized for using a tool that helps me think and communicate better.
Link to Tim's essay: https://substack.com/app-link/post?publication_id=7730464&post_id=213041071"},"title":{"matchLevel":"none","matchedWords":[],"value":"Ask HN: Should we disclose AI use in our work?"}},"_tags":["story","author_rezvovmobile","story_49469914","ask_hn"],"author":"rezvovmobile","children":[49470254,49568866],"created_at":"2026-08-27T19:21:26Z","created_at_i":1787858486,"num_comments":1,"objectID":"49469914","points":2,"story_id":49469914,"story_text":"Tim O'Reilly just published an essay arguing that AI is a medium, like photography or painting. His key line: "I judge the output, not the means used to produce it."
He describes his own process: a 300-word prompt becomes a 660-word draft, which he rewrites into a 3,000-word essay. Maybe 40-50 words are Claude's. He's proud of how he uses AI. He says we're still in the "filmed stage plays" era of AI writing, and the masters will emerge.
This resonates with me. I'm an engineer, not a writer. Over the past few years, I've been using AI extensively \u2014 for drafting articles, structuring complex technical ideas. The AI helps me articulate thoughts I couldn't express alone. But I always rewrite, edit, and make sure every word is mine.
Here's my dilemma: should I disclose that I used AI as a tool?
When I don't disclose, I feel dishonest \u2014 like I'm claiming a skill I don't fully have. When I do disclose, people dismiss the work as "AI slop" before reading it. I've had journalists ignore emails, investors go silent, and peers assume the ideas aren't really mine.
Tim argues against Ted Chiang's position that AI writing lacks "choices." Tim says a prompt can contain a lifetime of experience, like Leonard Cohen's "Hallelujah" taking five years or Dylan's "Blowing in the Wind" taking ten minutes \u2014 both are masterpieces.
But here's the practical question for this community:
If you're an engineer, founder, or technical person using AI to communicate ideas you couldn't express otherwise:
1. Do you disclose AI use? When and how?\n2. Have you faced backlash or dismissal when you did disclose?\n3. Do you think "judge the output, not the means" is realistic, or will AI-assisted work always carry a stigma?
I'm genuinely trying to figure out the right approach. Not to cheat, but to be honest without being penalized for using a tool that helps me think and communicate better.
Link to Tim's essay: https://substack.com/app-link/post?publication_id=7730464&post_id=213041071","title":"Ask HN: Should we disclose AI use in our work?","updated_at":"2026-09-04T19:10:20Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"rhettjull"},"story_text":{"matchLevel":"none","matchedWords":[],"value":"I run a small agency and we\u2019ve been building client websites for years. The work is hands-on, repetitive, and time consuming. We tried platforms like Replit and Lovable, but the output didn\u2019t hold up for production use. Things were missing, editing was limited, and reliability was an issue.\nOut of that frustration we built Zylo, an AI platform that generates production-ready web apps. It is not instant like Lovable (our builds take roughly 15 minutes on average depending on complexity) but we focused on completeness and reliability instead of speed.\nWhat Zylo does\nGenerates full stack Next.js projects including frontend, backend, and database setup
Built-in domain management so you can bring your own or purchase directly inside Zylo
E-commerce system that feels like a lightweight Shopify with product management and categories
Stripe integration through API connection for payments
How it works under the hood\nYou interact with an AI chatbot that coordinates several agents following the same processes we used manually when building sites
Agents generate code, proofread, and check for missing assets or design issues
Any build, runtime, or TypeScript errors are automatically caught and repaired before deployment
A final agent handles production deployment and gets the project hosted and ready for a domain connection
Editing\nWe put a lot of effort into the editing side. There is a live Monaco editor that renders the site. You can click directly on any component, section, or entire page and pass that as context back to the agents for regeneration. This was something we found lacking in other tools and wanted to solve.\nWhat\u2019s next\nWe are currently working on a visual workflow builder. Think of something similar to GoHighLevel\u2019s UI, but instead of manually wiring things, the AI fills in the code and functionality behind the scenes. The idea is to let people map out flows visually and have them actually run in production.\nWe would love feedback from the HN crowd. The big trade-off we made is slower build times in exchange for more complete and reliable projects. Do you think that trade-off makes sense, or would speed always win out for you?"},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: AI that trades speed for reliability in site generation"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://www.myzylo.app"}},"_tags":["story","author_rhettjull","story_45418719","show_hn"],"author":"rhettjull","created_at":"2025-09-29T21:03:41Z","created_at_i":1759179821,"num_comments":0,"objectID":"45418719","points":2,"story_id":45418719,"story_text":"I run a small agency and we\u2019ve been building client websites for years. The work is hands-on, repetitive, and time consuming. We tried platforms like Replit and Lovable, but the output didn\u2019t hold up for production use. Things were missing, editing was limited, and reliability was an issue.\nOut of that frustration we built Zylo, an AI platform that generates production-ready web apps. It is not instant like Lovable (our builds take roughly 15 minutes on average depending on complexity) but we focused on completeness and reliability instead of speed.\nWhat Zylo does\nGenerates full stack Next.js projects including frontend, backend, and database setup
Built-in domain management so you can bring your own or purchase directly inside Zylo
E-commerce system that feels like a lightweight Shopify with product management and categories
Stripe integration through API connection for payments
How it works under the hood\nYou interact with an AI chatbot that coordinates several agents following the same processes we used manually when building sites
Agents generate code, proofread, and check for missing assets or design issues
Any build, runtime, or TypeScript errors are automatically caught and repaired before deployment
A final agent handles production deployment and gets the project hosted and ready for a domain connection
Editing\nWe put a lot of effort into the editing side. There is a live Monaco editor that renders the site. You can click directly on any component, section, or entire page and pass that as context back to the agents for regeneration. This was something we found lacking in other tools and wanted to solve.\nWhat\u2019s next\nWe are currently working on a visual workflow builder. Think of something similar to GoHighLevel\u2019s UI, but instead of manually wiring things, the AI fills in the code and functionality behind the scenes. The idea is to let people map out flows visually and have them actually run in production.\nWe would love feedback from the HN crowd. The big trade-off we made is slower build times in exchange for more complete and reliable projects. Do you think that trade-off makes sense, or would speed always win out for you?","title":"Show HN: AI that trades speed for reliability in site generation","updated_at":"2026-09-04T07:51:04Z","url":"https://www.myzylo.app"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"paedubucher"},"story_text":{"matchLevel":"none","matchedWords":[],"value":"Ted Unangst used to have a nice blog called "flak". It was available under flak.tedunangst.com
Now the page seems to be gone completely. Does anybody know what happened to him or that page? I really miss that blog."},"title":{"matchLevel":"none","matchedWords":[],"value":"Where did Ted Unangst's website go?"}},"_tags":["story","author_paedubucher","story_49565127","ask_hn"],"author":"paedubucher","children":[49565136],"created_at":"2026-09-04T14:24:27Z","created_at_i":1788531867,"num_comments":2,"objectID":"49565127","points":1,"story_id":49565127,"story_text":"Ted Unangst used to have a nice blog called "flak". It was available under flak.tedunangst.com
Now the page seems to be gone completely. Does anybody know what happened to him or that page? I really miss that blog.","title":"Where did Ted Unangst's website go?","updated_at":"2026-09-04T15:14:05Z"}],"hitsPerPage":50,"nbHits":39,"nbPages":1,"page":0,"params":"query=Thunder+Compute&tags=story&hitsPerPage=50&advancedSyntax=true&analyticsTags=backend","processingTimeMS":15,"processingTimingsMS":{"_request":{"roundTrip":16},"afterFetch":{"format":{"highlighting":4,"total":4},"merge":{"mergeLoop":{"prepareNextHit":1,"total":2},"total":2},"total":2},"fetch":{"query":10,"scanning":1,"total":12},"total":15},"query":"Thunder Compute","serverTimeMS":20}