It's not really "introspection" or "Cogito, ergo sum", but I found that over tens of millions of tokens of iterations the system was able to create something novel. I am showing off one of the dashboards it created based on the deep research and Python modelling conducted:
https://ai-capex-sens-wxc4b.ondigitalocean.app/
---
TLDR; AI hype is one thing, but how else would you actually go pedal to the metal spending the GDP of a middle income country on new AI data centers? Basically, theres hundreds of billions of dollars in cloud business from enterprise contracts, boring IT stuff but its where a lot of the money in tech comes from and its real. Beyond the pontification about superintelligence is a similar reality for AI in the eyes of the hyperscalers that scaled the cloud business in the past, hundreds of billions to trillions in business from enterprise contracts (cloud replaced on-prem servers, AI replaces or augments knowledge workers).
The model estimates AI revenue from automatable knowledge-work spend, stress it with adoption/pricing assumptions, and run it against capex + operating + financing costs under delays/shocks. In the base case, it doesn\u2019t break even by 2032, but of course there's other sources of revenue so this isn't all that doomer or surprising that its risky to spend so much. However, its nice to have numbers and sensitivity to back that conclusion up instead of hand waving it.
---
There's been a lot of AI mania over vibe coding, but the main conclusion I got from running this harness is we can in fact use AI harnesses to simply think and make better decisions too. In this case, the AI is reducing the burden of injecting rigor into writing and decision making."},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: AI models AI capex"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://ai-capex-sens-wxc4b.ondigitalocean.app/"}},"_tags":["story","author_liujjw","story_47484145","show_hn"],"author":"liujjw","created_at":"2026-03-23T00:54:08Z","created_at_i":1774227248,"num_comments":0,"objectID":"47484145","points":2,"story_id":47484145,"story_text":"Bernie interviewed Claude recently (https://www.youtube.com/watch?v=h3AtWdeu_G0). The one-shotted responses were predictably a little sycophantic, but some responses were a little surprising. Made me think I'm kinda doing something similar in a more permissive timeframe, I have a harness over Claude that "persistently" studies AI spending, it does deep research, responds to my inputs, persists notes, writes code, etc.
It's not really "introspection" or "Cogito, ergo sum", but I found that over tens of millions of tokens of iterations the system was able to create something novel. I am showing off one of the dashboards it created based on the deep research and Python modelling conducted:
https://ai-capex-sens-wxc4b.ondigitalocean.app/
---
TLDR; AI hype is one thing, but how else would you actually go pedal to the metal spending the GDP of a middle income country on new AI data centers? Basically, theres hundreds of billions of dollars in cloud business from enterprise contracts, boring IT stuff but its where a lot of the money in tech comes from and its real. Beyond the pontification about superintelligence is a similar reality for AI in the eyes of the hyperscalers that scaled the cloud business in the past, hundreds of billions to trillions in business from enterprise contracts (cloud replaced on-prem servers, AI replaces or augments knowledge workers).
The model estimates AI revenue from automatable knowledge-work spend, stress it with adoption/pricing assumptions, and run it against capex + operating + financing costs under delays/shocks. In the base case, it doesn\u2019t break even by 2032, but of course there's other sources of revenue so this isn't all that doomer or surprising that its risky to spend so much. However, its nice to have numbers and sensitivity to back that conclusion up instead of hand waving it.
---
There's been a lot of AI mania over vibe coding, but the main conclusion I got from running this harness is we can in fact use AI harnesses to simply think and make better decisions too. In this case, the AI is reducing the burden of injecting rigor into writing and decision making.","title":"Show HN: AI models AI capex","updated_at":"2026-03-24T00:39:55Z","url":"https://ai-capex-sens-wxc4b.ondigitalocean.app/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"lukasego"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["deep","cogito"],"value":"Hi HN,
I built a Google Docs Add-on that analyzes your writing to reveal your thinking patterns.
I like writing. For me, its purpose has always been clarifying ideas. Writing not only sorts your thinking - it improves thinking.
In the age of LLMs, however, people use them for generating texts. They got it the wrong way around.
Writing is for solidifying your narratives. It\u2019s for yourself.
Many have discussed essay writing. Most recently and famously, in this HN thread on the MIT paper on cognitive debt (https://arxiv.org/abs/2506.08872). PG had a nice essay about good writing (https://paulgraham.com/goodwriting.html).
Hence, I built a simple Google Docs Add-on, Cogilo (https://cogilo.me), that puts the writer\u2019s thoughts at the center.
Cogilo analyzes text sections for four cognitive markers. It reflects how much you suffer from analysis paralysis, information overload, and uncertainty in your essay. And it shows where you are in a flow state.
Here\u2019s a demo of Cogilo: https://www.youtube.com/watch?v=QOUZ2c0Rtjw
People can try it out on the Marketplace (https://workspace.google.com/marketplace/app/cogilo/31975274...), no sign-up.
You get reflected if your text sounds incoherent or circles too often around established concepts, with Cogilo marking passages where your thinking can be improved the most.\nOpus 4 then powers rewrite suggestions in your own writing style, incorporating feedback, to convey the same meaning, but more nuanced. They get inserted, and you can compare new and old, and modify.
Ultimately, the goal is that people automatically reflect their thinking processes and enhance them over time through Cogilo.
My team uses it for deep writing sessions, so it might be useful for you. It has helped us reason out concepts - and become better at reasoning itself.
Cogilo follows some personal principles on Writing & LLMs, maybe you\u2019d like to critique them: https://cogilo.me/blog/cogilo\nHappy to hear about views on LLMs as an analytical tool first!"},"title":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["cogito"],"value":"Show HN: Cogilo \u2013 Cognitive Mirror in Google Docs"},"url":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["cogito"],"value":"https://workspace.google.com/marketplace/app/cogilo/319752742380"}},"_tags":["story","author_lukasego","story_44338112","show_hn"],"author":"lukasego","created_at":"2025-06-21T15:00:14Z","created_at_i":1750518014,"num_comments":0,"objectID":"44338112","points":4,"story_id":44338112,"story_text":"Hi HN,
I built a Google Docs Add-on that analyzes your writing to reveal your thinking patterns.
I like writing. For me, its purpose has always been clarifying ideas. Writing not only sorts your thinking - it improves thinking.
In the age of LLMs, however, people use them for generating texts. They got it the wrong way around.
Writing is for solidifying your narratives. It\u2019s for yourself.
Many have discussed essay writing. Most recently and famously, in this HN thread on the MIT paper on cognitive debt (https://arxiv.org/abs/2506.08872). PG had a nice essay about good writing (https://paulgraham.com/goodwriting.html).
Hence, I built a simple Google Docs Add-on, Cogilo (https://cogilo.me), that puts the writer\u2019s thoughts at the center.
Cogilo analyzes text sections for four cognitive markers. It reflects how much you suffer from analysis paralysis, information overload, and uncertainty in your essay. And it shows where you are in a flow state.
Here\u2019s a demo of Cogilo: https://www.youtube.com/watch?v=QOUZ2c0Rtjw
People can try it out on the Marketplace (https://workspace.google.com/marketplace/app/cogilo/31975274...), no sign-up.
You get reflected if your text sounds incoherent or circles too often around established concepts, with Cogilo marking passages where your thinking can be improved the most.\nOpus 4 then powers rewrite suggestions in your own writing style, incorporating feedback, to convey the same meaning, but more nuanced. They get inserted, and you can compare new and old, and modify.
Ultimately, the goal is that people automatically reflect their thinking processes and enhance them over time through Cogilo.
My team uses it for deep writing sessions, so it might be useful for you. It has helped us reason out concepts - and become better at reasoning itself.
Cogilo follows some personal principles on Writing & LLMs, maybe you\u2019d like to critique them: https://cogilo.me/blog/cogilo\nHappy to hear about views on LLMs as an analytical tool first!","title":"Show HN: Cogilo \u2013 Cognitive Mirror in Google Docs","updated_at":"2025-06-22T05:51:14Z","url":"https://workspace.google.com/marketplace/app/cogilo/319752742380"}],"hitsPerPage":50,"nbHits":4,"nbPages":1,"page":0,"params":"query=Deep+Cogito&tags=story&hitsPerPage=50&advancedSyntax=true&analyticsTags=backend","processingTimeMS":14,"processingTimingsMS":{"_request":{"queue":25,"roundTrip":27},"fetch":{"query":11,"scanning":1,"total":13},"total":14},"query":"Deep Cogito","serverTimeMS":40}