SPEAKER_00: hey everybody it's monday and we have an insane news day i've been on the road i was at a south by southwest so so much news backed up about chat gpt open ai apologies big bitcoin bet the banking crisis hyperinflation so much macro stuff that i thought we would have producer rachel sit in and read the news and i would respond to it and give my insights this is uh just an action-packed

SPEAKER_01: episode that includes three j traits two that occurred this morning and one that occurred on air it's going to be a great show so please stick with us this week in startups is brought to you by

SPEAKER_02: the embroker startup insurance program helps startups secure the most important types of insurance at a lower cost and with less hassle save up to 20 off of traditional insurance today at embroker.com twist while you're there get an extra 10 off using offer code twist linkedin jobs a business is only as strong as its people and every hire matters go to linkedin.com twist to post your first job for free terms and conditions apply and cashfly is a pure play cdn and they've been doing it for over 20 years cashfly makes cdn simple effective and secure get 10 terabytes free forever if you sign up at twist dot cashfly dot com that's twist dot c-a-c-h-e-f-l-y dot com

SPEAKER_06: all right everybody let's get to the news it's monday so much news rachel has uh backed up here so uh welcome to the pod rachel be reading the news and uh what's in the news i know the banking crisis and balaji's big bitcoin bet are filling my uh twitter feed which is good that's good that it's

SPEAKER_08: not me in my twitter david sacks in my twitter feed so it's nice to see somebody else uh but what's the

SPEAKER_10: latest so there's been like you said there's just been a ton of crazy news over the weekend and into

SPEAKER_12: the morning and the latest in the banking crisis is the fed and central banks of the uk canada japan switzerland and the eu announced a coordinated coalition to improve liquidity and people on twitter are speculating that this means a quantitative ev quantitative easing and money printing will

SPEAKER_13: resume stabilizing the markets along with potential rate cuts yeah so um it's going to take coordination

SPEAKER_06: here we have fractional banking which means like we put our money in they go do stuff with it and we can't get our money back because it's not like a bank vault or a stable coin um so this is um what has to happen if there is a bank run or if people do not feel safe with their money in a fractional bank and so quantitative easing the government buying securities uh and some of this you know medium or long duration uh bonds treasuries etc and we need to get a full backstop going and i saw ubs is going to

SPEAKER_00: acquire their big rival credit suisse and each of these banks is a different story for silicon valley bank it seemed like maybe they didn't have risk defined uh properly and they weren't managing risk

SPEAKER_06: properly for credit suisse this is a bank that's had multiple problems turnover and management um and first republic bank obviously is the you know regional bank that's in play most uh they

SPEAKER_00: have i think they're trading at like 15 a share right now and as i said two week weekends ago as you probably know rachel watching my twitter hey we're seeing multiple bank runs occur in silicon valley the first one was silicon valley bank then people move their money to first republic because that's considered analogous right in the community silicon valley bank first republic these are just great banks that give great service to founders and venture capitalists and when i say great service i don't mean they're doing anything uh fugazi or corrupt they're just as i i said in a now famous trending video like if you want a mortgage they're going to come and give you that white glove service if you want venture debt they're going to come meet with you and they're going to consider your request so it's kind of that white glove service very personalized service you call you get a banker first republic actually known for that you have a personal banker there even if you're a small company because

SPEAKER_06: they know the small companies grow into big ones and so um i see there's a rescue plan for first

SPEAKER_12: republic now yes um the wall street journal reported that jp morgan chase ceo jamie diamond is leading discussions with ceos of other major banks to help stabilize first republic yeah um some quotes from the article include uh discussions while preliminary have focused on how the industry could arrange for an investment that would boost the capitals bank according to people familiar with the matter and by the way like i said this is from the wall street journal among the options on the table the people said is an investment in first republic by the banks themselves uh so we could actually see a couple of the top four u.s banks investing in first republic and by the way james diamond is like the wolf of all wolves he's involved he's super serious and that's probably good because he probably

SPEAKER_06: wants to make like a ton of money yeah uh well uh these banks are now trading the equity in them is incredibly low i mean maybe it's a three or four billion dollar company right now uh and it needs tens of billions of dollars to backstop it so i think you would basically own that bank if you were

SPEAKER_00: to start investing in it and and this is going to have a big impact on startups and what people don't understand is startups can either do you know very basic banking with a what's called a neo bank like a digital bank no storefront bank and that's fine you get your banking service you can wire money you can do payroll uh or you can wire to a payroll service uh and this is aside from the rails for doing payment right so if your startup has to do payments you're probably using stripe or adgen then you have uh banking services so you can do payroll and uh you know put your money somewhere yeah so what are you

SPEAKER_35: action yeah good no keep going keep going about startups i'm going to ask you yeah startups really

SPEAKER_00: don't need much from a bank uh they just need to have it be easy to use maybe on the margins it's nice if you can get venture debt again i don't recommend venture debt until you have a cfo because cfos know how to manage these things and debt can get you in trouble real quick because you're basically raising part of your next round then you're telling your next round investors like say the series b you got to pay for my series b and then invest in the series b it's just not a good device unless you're at your series d have predictable revenue and you're using that venture data as like a way to build i don't know if it was cafex a way to build 10 machines that you have already sold right which would be called factoring in some worlds taking your uh purchase orders and factoring in fact there

SPEAKER_16: was a company pipe.com that did that i think stripe has a product that'll pay you in advance of future earnings um but you know the ease at which you could work with first republic bank the ease of which startups can work with silicon valley bank is a true loss there was i saw um lemon.io uh also an advertiser on this program and just for full disclosure i have first republic accounts i have a silicon valley bank tiny mortgage at this office in fact uh from a couple years ago but we're selling this office uh they both have advertised on the prod on the podcast pipes advertised uh lemon.io is advertised i mean everybody basically advertises on this product it's it's the number one startup

SPEAKER_06: podcast so uh to say i'm conflicted i mean there's tiny little um conflicts but i don't have any exposure to them but lemon.io had an interesting tweet he's based in ukraine and they help um vet

SPEAKER_16: remote developers from europe for american companies in fact i want to use it to build out the software around the syndicate.com and uh the ceo uh had a series of tweets maybe you can recap those yeah so

SPEAKER_12: they started on sunday march 12th where he announced that lemon.io had all the funds deposited at svp which is not great yes so his first tweet was we are a ukrainian startup that had all of our funds at svp whatever happens on monday we should be prepared to move forward with just 250 000 from fdic please share this message this will help bring new clients and keep operating to support the engineers and startups we have 400 plus vetted engineers in europe and lot am rates are 2 000 to 2.8 000 per week dm me if you're hiring i'm happy to give 15 discount for the first four weeks of work and then a day later while trying to move their money to chase he tweeted about how chase didn't really understand his marketplace business chase banker could not understand this is again this is the tweet chase banker could not understand marketplace so they filed that we do marketing

SPEAKER_49: which is like completely off um they told me that in fairness it has like seven six of the same letters market say like it has the word market in it it does yes so great and they told them basically

SPEAKER_54: they told uh alexander the ceo that they couldn't open the account unless they have a certificate of

SPEAKER_12: good standing and then he completely threw these people off at chase by saying they're remote and they don't have an office and finally on march 17th he tweeted that chase decided to close their account because they were fully remote without a physical office um and in a follow-up tweet someone questioned why alexander wouldn't just put his home address and to that he responded uh he doesn't want us to be public which is pretty reasonable and he his lease doesn't allow it and the new yc ceo and friend of the pod gary tan actually retweeted the following tweet uh from ara uh who goes in who goes in a startup founder who responded to lemon.io's situation uh this is why founders were using svb in case y'all were wondering regularly big banks have no idea how tech startups work for us they wanted two years of profit and lost statements for a five thousand dollar credit card we had 100 times that in the corporate checking account but they didn't care so yeah i guess this is why startups

SPEAKER_05: love svb so much i've been dealing with business insurance for three decades i am on the board of a bunch of companies i watch people who don't have insurance get themselves into trouble all the time switching providers has always been a nightmare it's too expensive takes too much time and often it doesn't even guarantee better coverage but now you can make switching radically simple with imbroker yes and broker is the perfect destination for industry tailored commercial insurance it's business insurance specifically for startups in broker single application helps startups get four

SPEAKER_62: quotes one two three four for four lines of coverage in just 15 minutes they connect you with one of their expert brokers for unmatched service that goes beyond your policy and listen broker is such an amazing product i use it a lot of my startups use it it's so easy to use so try in broker today with code twist and get 10 off their startup package at imbroker.com twist that's e-m-b-r-o-k-e-r dot com slash t-w-i-s-t and use the code twist so you get that 10 off it's meaningful every dollar counts right now we love you in broker thank you so much for supporting this podcast for so many

SPEAKER_00: years yeah i mean any bank that understands you whether you're a farmer there are regional banks that work with farmers they understand the season buying seeds equipment you know so you're just basically starting on second base with every conversation um and so now what founders are doing and the best practice is put your money into one of the top three or four banks until this blows over um and then you know maybe a neo bank or two and they have great services to do this but basically you want to get into as many fdic accounts as possible i'd say for seed stage and series a companies you could literally put yourself in five bank accounts a couple of money market accounts split 250k into literally five to ten accounts and have you know a million two million dollars fdic insured and if your payroll and your spending was under 250 a month i guess your uh your accounting function in your startup whether outsourced or insourced could probably manage it it's a bit challenging i think the fd ultimately what has to happen is fdic limit just needs to be one million or two million and then then you would see you know i don't know if it's 80 and i don't i haven't seen these statistics if somebody has a way to get me these statistics producers at this week in startups.com i'd be very interested in knowing how many accounts uh business accounts in the united states have under 250k in them and i think a reasonable you know multiplier is 3x or 4x so let's just use the 4x one because it's easy for math how many because you can manage rachel four bank accounts pretty easily right yeah or four money markets etc that would all be fdi and c fdic insured okay so four times that number um would be what i would want to know so how many people have one million two million three million dollars in their bank account as businesses is that 10 20 because we could probably just by raising the fdic limit make it so most people don't have to do this you know splitting their money to five or ten accounts to get some reasonable amount of protection that seems like common sense

David Friedberg: uh of course with all of this uh banking chaos one is wondering you know again for the impact for

SPEAKER_00: startups what's going to happen in the market so yeah for founders who are listening you when you have instability in the market rich people venture capitalists uh uh hnis high net worth individuals they have to stop being optimistic they have to stop being creative with their money and they have to retreat and be defensive right so they go from offensively deploying capital to defensively protecting capital that's the moment we're in what does that mean it means people are scared and people are in protect mode that means they can't write that 25k check for your uh startup right so nine out of ten angels four out of five seed funds whatever they might be in defensive mode right now trying to protect their existing portfolio trying to protect their nest egg and that's a shame now friend of the pod uh apology uh if you don't follow him on twitter he's b a l a j i he's been on this pod he's been on all in he's spoken at my events super smart but he's also a crypto maximalist i don't think he's actually a bitcoin maximalist but he is a crypto united states is in downfall the fed you know it's as some people might say catastrophist he might be a catastrophist uh and he could be right so just because you're a catastrophist does not mean you're wrong you could see a catastrophe coming and accurately be a catastrophist uh as i was a week ago when i correctly saw the bank runs happening and here we are it has been a catastrophe uh that is being

SPEAKER_72: dealt with so maybe you can fill us in on bala apologies is how you pronounce it apologies big

SPEAKER_54: bitcoin bet yeah so like you said it's an insane bet between bitcoin and the us dollar and the bet is

SPEAKER_12: that bitcoin will hit 1 million in 90 days by june 17th and the bet all started like on twitter which has been really cool to watch in real time where a user named james medlock tweeted the following on march 16th all but anyone 1 million dollars that the us does not enter hyperinflation and apology accepted the deal and it stipulated the terms over on twitter um under the proposed terms if bitcoin's price does not hit 1 million by june 17th james medlock will be winning 1 million dollars worth of usdc and one bitcoin and for reference bitcoin is trading at about 28 000 right now if bitcoin does hit 1 million in 90 days apology can keep the one bitcoin and the 1 million in usdc and jason can you kind of explain

SPEAKER_06: this bet a little bit more for us all right so he's betting that bitcoin will hit a million dollars

SPEAKER_00: and so uh it's at and i think at the time he did this maybe was at 25 it's at 28k right now so he's basically saying it's going to hit that by then if not uh you get a million dollars and so why would you make a bet like this well it trended right and what he's saying is there'll be hyperinflation and that everybody will run to bitcoin as a safe haven and if there's hyperinflation your dollars become worth less etc and so when you print up when you print a lot of dollars two or three times the amount then you're going to have inflation everybody saw that coming during covet remember everybody was dropping money from the sky you know here are here's payroll here's extended unemployment here's stimulus checks stimmy checks all of that money dropped onto the economy was to save the economy when the economy was shut down because everybody had to quarantine in place obviously that's over people are getting back to work but all that money still in the system has to be burnt off but that's what caused inflation many people correctly predicted that and then the fed had to fight to stop inflation they said it would be transit parade then they were wrong then they did this incredibly fast rate hikes the fastest in history to try to correct this so what biology is saying here is uh there will be more inflation and so inflation on inflation i guess in his mind equals hyperinflation because now the banks have to be bailed out so that's the bet he's making now to think that bitcoin would go from 25 to a million 40 times right four times 25 is a hundred thousand 40 times 25 k is a million for him to be making that bet you'd have to say well that's that's that's not going to happen right in all likelihood uh that would be a pretty violent run up uh and not really possible so what i think he's doing here is i would guess biology has a very large bitcoin position and there is nothing by the way illegal about this there is nothing unethical about this you can make a grandstanding bet i would put this as a you know grandstanding bet uh so he has let's say a thousand bitcoins a thousand bitcoins would be worth at twenty five thousand twenty five million dollars that means uh if bitcoin were to increase by one percent uh that would be 250k if it increased by four percent that would be a million dollars yeah if the price of bitcoin goes up four percent because of his crazy bet and this i'll call it a grandstanding bet you know like a trending bet something that takes over twitter he only needs to see bitcoin go up four percent to cover the loss from the bet anything above four percent increase

SPEAKER_54: he is profit for his bitcoin holdings does that make sense that does and it's super our producer chat

SPEAKER_12: is actually going insane because as you're saying that percentage bitcoin is actually up even more

SPEAKER_87: than that it's up 15 over the past five days okay so let's pause on that the 11 increase would all be profit for him so he has made if he in this scenario i don't know that he has this but i would not be

SPEAKER_93: surprised if he has a thousand bitcoins and 25 million dollars in bitcoin if it goes up 11 we'll just round it to 10 that's 2.5 million dollars right uh so he has made 2.5 million dollars by losing the million dollar bet this would be as if i said i will bet anybody uber would be worth a hundred

SPEAKER_24: dollars a share it's trading in like 30 bucks a share right now well i would triple a very large

SPEAKER_93: base of uber shares that i still own and if i bet somebody a million dollars that would happen i would make so many more tens of millions that i would be able to you know if it just went up to 60 or 70 or 80 i would make so much money that i could say to everybody i'm going to pay off that bet it didn't hit a

SPEAKER_24: million i lost the bet theoretically but i won because i was betting my position so um what people

SPEAKER_00: have to understand is in a free society two things uh can occur people can place bets people can talk about those bets and given social media those bets can trend and create a dialogue now when i told everybody because this relates to the claim the incorrect claim that venture capitalists called the bank run by tweeting now it's possible that venture certain venture capitalists did in fact cause a bank run on silicon valley bank by telling their portfolio companies to leave that uh certainly occurred on wednesday of thursday before um silicon valley bank was foreclosed on on friday and there was 42 million dollars was sent out and i think there was upwards of 100 million billion dollars um was asked to be reclaimed so you could say that vcs uh caused that bank run by telling all of their founders hey silicon valley bank's got problems it's been well documented for months and the ceo is saying hey there's not going to be a bank run or trust us on wednesday night again nothing illegal about doing that but some people might be able to point blame at vcs for that you would be incorrectly pointing blame

SPEAKER_01: on it you should be blaming the risk management team at silicon valley bank for not managing risk

SPEAKER_00: it's now here we are a week later and on that saturday night many of us said hey there's a bank run occurring we need to shore this up that was after silicon valley bank was in receivership so let's be clear pulling a fire alarm because you see a fire is different than starting a fire uh in this case is what we have to ask ethically morally i think rachel is is bology starting a bank run here yeah on the us dollar and here's the truth it's farcical to think that bology as well spoken as he is as popular as he is is um going to be able to take down the us dollar or america and that he is the reason for this we need people like bology pointing out when america is doing things where the fed's doing things that you know he believes are um you know possibly going to cause inflation we need that backstop obviously we need that back so because the fed didn't understand they said inflation was transitory and svb was not managing their money correctly according to you know risk management theory so he's well within his right to uh point this out he's well within his

SPEAKER_37: right to place a bet now you might not like it it might feel opportunistic the march to a billion

SPEAKER_116: users continues for linkedin they have 875 million people using the platform and with all of these recent layoffs in big tech there are an amazing array of insanely talented people out there ready to join your team the stakes are high you need to make every single position in your company count everybody has to pull their weight and there are so many great people looking to find their next opportunity right now let's call it what it is let's be candid here on this week in startups and i want you to get your first job posting for free at linkedin.com twist that's all you need to know linkedin.com twist the best place to hire somebody amazing and if you've seen those purple hiring rings around people's profile photos that is that person telling you they're hiring and that has created a massive network effect inside of linkedin go ahead and get that purple ring around your uh profile photo and then engage all over linkedin there's groups you can post you can reply to people and that will let you find better candidates faster let's just pause there if there's two things you remember from this ad read better candidates faster better faster better faster linkedin jobs better faster here's your call to action linkedin jobs helps you find qualified candidates you want to talk to faster post your job for free at linkedin.com slash twist that's linkedin.com slash twist to post your job for free terms and conditions do apply um but you know

SPEAKER_103: this apology tweet uh where he responded to me early monday morning i think is um you know i think

SPEAKER_72: one worth us uh you know just thinking about so here maybe you could read that one yeah so he responded

SPEAKER_12: to you um after you kind of asked like how many bitcoin does he have and then he did point out that um even if bitcoin raises that four percent he'll be able to pay off that sensational one million dollar bitcoin bet yeah um and he responded you still don't get it the fed is devaluing the dollar it's what dalio predicted monetization of the debt all the wars paid for in an orgy of printing dollars worth toilet paper on the other end this isn't about making money at all it's about getting innocence to bitcoin lifeboats before the fed crashes the ship and this isn't the only thing we saw from um from him this weekend we also saw him uh he gave an in-person speech which the speech wasn't circulated but his outfit was um on twitter i don't know if you were able to see that but basically

SPEAKER_00: completely irrelevant if he's wearing shorts and yeah uh what what the kids call uh flip-flops these

SPEAKER_129: days those are slides those are slides but yeah i'm cool with it if you want to wear slides uh i don't

SPEAKER_00: think he's any less intelligent for wearing slides i know some brilliant people who wear slides and dress like they're going to the gym or like going to the pool at the four seasons i'm here for it it's totally fine smart people can dress casual that's just a dumb uh twitter meme to be like hey trust a guy who's wearing slides and shorts but the truth is i mean obviously i do get it i understand that the fed's printing a lot of money yeah and i understand his bet i think i don't think it's an unreasonable bet but and i actually do believe that he is sincere so i'll just say all those things i believe it's it's a fine bet to make it's a reasonable bet to make on bitcoin in a time of depletion i don't know if i would make it but he certainly is allowed to i believe he's allowed to pump the bet um and then i believe he's sincere that he thinks he's helping people all of those things i take him at his word i literally take apology at his word he's trying to help people but i also think he should disclose how many bitcoin he has which is a little bit of a safety risk but he should say i am over 20 of my networks in bitcoin so that would be a way for him to let people know disclaimer or transparency i have 25 of my net worth in bitcoin now there is a little bit of a security thing for him to say his net worth or to say that he's got a lot of bitcoin because then you know some russian you know chinese whoever you know like crazy uh hackers could then try to hack him to get his bitcoin so i do think there is a little bit of security risk for him but there's no security risk to say i have a double digit percentage of my net worth in bitcoin just so you know i have made my bitcoin bet i made my bitcoin bets at ten thousand twenty thousand any dollars above that'd be a way to say it i bought a lot of bitcoin at an average price of ten thousand i'm in the black on bitcoin that would be really great i think if he did that but i don't have any problem with it and we actually need people placing bets we need people discussing these issues because i actually think if i'm being honest that the fed doesn't understand what's going on here or they're not doing a great job and i think some of the banks are taking too much risk and i think we have a broken system where they're making decisions on outdated data and they're over steering one way or the other so they oversteered by printing too much money then they ignored uh increasing the rates and they thought inflation was transitory and now i think they're raising rates too fast and they should go at a slower pace so i

SPEAKER_24: think like and i'm not an expert on this but it's pretty obvious that those three mistakes were made too much money was printed they waited too long to raise rates and they got it wrong on inflation and now they're going too fast and going too fast breaks things and what did it break it broke the banking system and now we have to play cleanup so i think going pausing or doing you know maybe less rates uh let's raises and just saying hey for three months we're going to just let things settle and assess you know what's happening in the ecos in the in the economy but what i've learned over this year and a half you know as being an adult living through my third boom bust cycle and maybe having a lot more experience and a lot more time to consider these things and a lot more money to work in the system is that macro is very hard i think there might only be micro in the

SPEAKER_00: world which is to say people make transactions uh based on their best interests and then there are

SPEAKER_24: people who try to interpret those the cumulative nature of all of those the cumulative output of all of those micro transactions to make macro predictions and it's incredibly hard to make those macro predictions some things are obvious if you raise rates you know it's going to cause um

SPEAKER_01: inflation to go down it's going to cause consumption go down but you know people are still trying to figure out what's going on with employment as well one thing and they can't seem to figure it out they're trying to figure out the real estate market they're having a hard time figuring that out so

SPEAKER_12: i think macro is hard is what i've come to the conclusion anthology kind of i think things similar to you in the way that like banks are are not doing so hot right now and they seem to be broken because although we weren't able to like hear anything that he was talking about during his discussion where he uh were the slides um some twitter users posted some of his key topics which included banks are insolvent in the us bitcoin is going to go to 1 million in 90 days he converted 99 of his net worth to bitcoin oh there it is uh yeah which was tweeted we can't confirm that though we weren't there this is by crypto underscore mckenna um and her profile picture or their profile picture is a statue um he also predicted that a civil u.s civil war is going to happen okay is this dramatic

SPEAKER_01: or are we having yeah i would say if you were looking at this and you wanted to

SPEAKER_00: say that there are venture capitalists who are catastrophists if you were to use that word catastrophist this would be truly catastrophizing most banks are insolvent and the civil u.s civil war is about to happen now if you were to look at reality and say well what is bology's point here what if you were to steal man his argument january 6 certainly looked like a civil war to me you know a microcosm of one trump being arrested on tuesday tomorrow reportedly um for the hush money payments uh in that case like could that trigger some sort of uh civil war i don't know like civil war the entire country goes to war but it's certainly there's division so going from division to civil war that's a big jump but you could see that the country is more divided now than it feels like it's been in our lifetime btc going to 1 million that's him talking his book because he's 99 in btc hyperinflation imminent i don't know if hyperinflation is imminent some people believe deflation is going to happen because people are going to have no money to spend so if people have run through this is what smart people are telling me people have run through their savings uh consumers in the united states are in debt and people are kicking on austerity measures as i've talked about like 2022 and 2023 being years of austerity people are reducing their spend you know doing staycations you know not yoloing nfts etc so you know he could be wrong on all of these counts btc might be trading for less than 90 days most banks will be bailed out uh deflation could happen and a us civil war will be averted because

SPEAKER_148: some new candidate comes out and says let's stop fighting with each other let's balance the budget

SPEAKER_01: and solve these problems i actually think i'll take the other side of this i think most banks will be bailed out i think btc will be trading plus or minus 10 of where it is now and let's just put a pin in this so that we can actually check back in 90 days uh producers put this on your calendar i believe most banks will be bailed out two or three will be insolvent uh btc will be trading plus or minus 10 percent in 90 days uh hyperflation will not happen if anything we'll see some deflation uh and i believe uh biology will convert from 99 of his net worth in btc to a lower percentage and i

SPEAKER_137: guarantee that a civil war does not happen this year all right well we're gonna have to come back

SPEAKER_155: in 90 days and fact check put that on your calendar we're gonna make like a graphic to make sure with like a little check mark to make sure uh i mean people think i'm a catastrophist because i use

SPEAKER_157: cap blocks and i was like hey guys there's three bank roads happening i see them all right here

SPEAKER_08: i didn't predict civil war i mean come on yeah but you know what he cashed you also he cashed the

SPEAKER_24: ships out right like he left for singapore or somewhere uh i think is the report so i think i

SPEAKER_148: don't know if he gave up his u.s citizenship but i could see apology giving and i think he's a u.s citizen so um somebody can correct me if i'm wrong there but i think he like is a full-on get out of

SPEAKER_08: the u.s the u.s is the u.s empire is going to die just like uh freeberg's been talking about a lot of

David Friedberg: people this dalio book keeps telling people like hey empires have a certain amount of time and here's

SPEAKER_103: what empires look at like at the end and for those of you who haven't read the book uh the changing world order why nations succeed or fail it's pretty sensationalistic uh but it's well researched and it

SPEAKER_00: i think maybe and i think dalio points this out it's not as catash it's not as much of a catastrophe

SPEAKER_72: when one of these things one of these societies moves to the next um as uh producer nick pointed out because you usually have like an orderly transition like the uk and the united kingdom used to be the kingdom and then it was america and maybe america you know china an authoritarian country becomes the dominant country but that doesn't mean america goes away just like the uk hasn't gone away it just means it's not number one anymore so okay so we're gonna have to come back

SPEAKER_154: in 90 days everybody has 90 days to read the changing world order book club book it's a good

SPEAKER_08: read we'll catch you up catch you up on what everybody else is thinking and talking about

SPEAKER_116: okay everybody when it comes to the blocking and tackling of running a tech startup you don't need to reinvent the wheel we all know that cdns are a place where startups can really over complicate things content delivery networks you've heard this before if you're in the tech business you need to understand you don't need custom authentications or custom code if you're a startup you just need to check out cashfly cashfly is a pure play cdn and cdns are literally all they do and they've been doing it for 20 years they basically created the category cashfly makes cdn simple effective and secure that's what i want you to take from this simple and secure that's cashfly listen if you're going to go with a cdn from one of the major providers you're eventually going to outgrow them so don't burn your startup credits using a cdn at one of the larger players nope just let cashfly handle it for you they're going to help you deliver your content faster than your competitors which means user engagement goes up it doesn't matter if it's videos mobile apps gaming they're going to help you scale faster because that's all they do that's all they focus on so you want the pure play cashfly will save the day twist listeners get 10 terabytes free forever if you sign up at twist.cashfly.com that's t-w-i-s-t dot c-a-c-h-e-f-l-y dot com to get 10 terabytes for free forever twist.cashfly.com

SPEAKER_12: let's talk about chat gpt so much going on open ai launched gpt4 last week and it's definitely scaring some people including the ceo sam altman and if you want to use gpt4 right now you can only access it by signing up for chat gpt plus which costs 20 a month i uh i am paying for it i told

SPEAKER_103: everybody on my team they can go pay for it i've literally had the sales team uh i had producer nick

SPEAKER_00: do a um explainer to a little demo of chat gpt4 because we're doing this launch angel summit again this is where we get 100 angel investors to come together we're going to do a live taping of all in at it um should be great uh and we play poker and you know a bunch of angels and seed funds get together trade notes um and i was like who would you get to sponsor uh and a summit of angel investors and seed funds and it was like i would go for banking including svb uh you know comerica whatever i would get law firms like wilson cincini i would get insurance providers like in broker i would get cloud providers like aws and nasher and literally it gave me an answer of who to target as sponsors for the angel summit that included people who had previously sponsored the angel summit i was like well that is the sdr function and in uh sales we call that uh um sales development ref i think is what it stands for and those are people who warm up leads right so now you can use gpt4 and ask it hey who should sponsor this week in startups or i want to do an ai conference who would be great

SPEAKER_38: speakers and then you can uh think about what the topics would be at your ai conference and all that stuff and that brainstorming that would have occurred in a meeting rachel now could occur in gpt4 and

SPEAKER_00: you'll probably get 70 80 of what would happen in the meeting with five people yeah and if you had hired a producer for your conference you know probably will do 70 or 80 of what the producer would have done or producers on this very podcast so what does that mean it means you could have instead of having three producers you could have one maybe or instead of maybe instead of three you could have two right now which doesn't take a genius to realize that what we are writing notes for this very podcast or sales executives are looking to find leads you'll probably that function will go away and you could have one producer producing three times as much content or one salesperson reaching out to three times as many contacts for sales in other words human efficiency is about to 3x and you can see that that with github's pilot or other um folks who are writing code the code demos for gpt4 are extraordinary why is it extraordinary well that corpus of programming is a very tight corpus of very repetitive activities what does it mean it means if you were writing a modern app you would have a login screen you'd have a delete my account screen you would have a change my password screen you might have a feed like an instagram or twitter feed you might have a gp turn on my gps and tell me my location feed or function you might have a messaging function like you know dms on twitter or on instagram

SPEAKER_186: well those that code's been written it's open source and it's been indexed and it's been indexed at a

SPEAKER_00: specific location github and it's got a lot of participants over there who have edited and their open source projects therefore something like a language model can really process that easily just like images have already been processed for so many uh years perhaps decades that it's very easy to say hey make me a jedi we know what a jedi is and then have them instead of using a lightsaber have them use i don't know a crossbow or a bow and arrow like it it could make that pretty easily because it's a corpus with images or code that is narrow and easy to iterate on and then do reinforcement learning on

SPEAKER_24: so everybody will be three times more productive yeah don't worry producers are like their jobs are

SPEAKER_12: safe because i hear it's not like up to date with data past 2021 so until it gets that done where everyone's in the clear um some other jobs though that might be getting taken away or might be um and able to do something else with their time um were released by a research study with upenn and open research in the paper they predicted what us jobs will be impacted by llm's uh language learning models like gpt um yeah it was really really cool and a twitter user actually broke down the paper's major takeaways um okay well accounting for other generative models and complementary technologies our human estimates indicate that up to 49 percent of workers could have half or more of their tasks exposed to llms so research suggests that half of all of us jobs will be cut in half or made twice as

David Friedberg: efficient by llm that's crazy uh this makes total sense because i just explained it with developers sdrs

SPEAKER_00: and producers of podcasts right so what's going to happen is something like angel list or crunchbase or linkedin like will have their own built-in llms and you'll be able to say instead of using advanced search on crunchbase or linkedin like people are currently doing to find startups that have raised their seed funding that are based in new york city and that have founders who went to stanford or an ivy league school right let's say that was your search you could say i want founders who are building startups at the seed stage in the northeast that went that have at least one founder who went to an ivy league school uh and that have had their and have an app in the app store right okay really that would be something you would put an associate on or a researcher really at a venture firm well that research will probably take a week maybe put 50 hours into it that same research will be able to do this in one hour or two hours right and they're gonna have to do prompts they're gonna have to clean up the data and then over time they may not have to clean up the data so maybe instead of a researcher doing it an associate would do it maybe instead of an associate doing it a principal a managing director or even the general partner at a firm could do it but that's something we've seen already over time and people are still thriving and we have record low unemployment so i'll take the other side of the bet which is people will just be twice as efficient we'll have twice as many products in the world solving twice as many problems and doing it for less money just like cloud computing meant and fractional uh cto's outsourced hr outsourced accounting outsourced cloud computing outsourced office space like we work all of that didn't lower the number of startups it increased the number of startups and the amount of money it took to get a product to market went from being three million dollars 20 years ago 25 years ago take maybe two three million dollars to get an internet product to market then it went to 250 and then

SPEAKER_24: it went to 25k and then it went to essentially free just sweat equity the time it takes for the two founders to produce something and so all of this is amazing which is why i'll be candid here and put a little plug in it's why we started founder dot university where we have 300 founders come and for 12 weeks we have them take all these courses learn how to build companies and at the end we invest in 20 other companies 25k to be their friends and family around there will be many more people

SPEAKER_00: trying to build software that solves a problem in the world and that means if there was going to be a niche piece of software for tennis players or pickleball players those might have not been venture backable previously but now they will be because they're going to cost a fraction to make and

SPEAKER_24: you could make a really robust app instead of for three million dollars a year you know 20 people working in an app company you're going to be able to do it instead of 20 people working for a modern app company working on all platforms you'll be able to do with five people or 10 people right

SPEAKER_188: and then more flowers will grow so i think nothing to worry about for now for now there's another

SPEAKER_12: takeaway that high paying jobs are also having more exposure uh to llms and could be heavily impacted so basically the most exposed high paying jobs involve programming and writing um and those are really really easy for llms to replicate and the least exposed high paying jobs involve science and critical thinking and these are harder for llms to replicate so if you're a high paying vc that invests in like life science companies and you need to use critical thinking your job is safe so shout out to dave friedberg there but in the context of the paper uh it defines exposure as quote a measure of whether access to gpt or gpt powered systems would reduce the time required for human to perform a specific dwa detailed work activity or complete a task by at least 50 percent and then it included like a chart um that the researchers measured how much exposure different job types would likely have to llms and there you can see a bunch of different jobs scroll down on the bottom um you'll even see like search marketing strategists which um would be super helpful for like our team so really cool chart to

SPEAKER_00: look at just so people understand this chart when it says 100 that means 50 of the job can be done by chat gpt or another language model i think what this doesn't understand here where it says like you know oh this one is really exposed accountants auditors news analysts reporters that doesn't mean the job goes away it just means the job is going to be massively more efficient and let me tell you something there's things that journalists should be doing that they're not doing right now because they don't have enough time to report on stories like calling the sources and checking the facts like uh you know maybe doing some analysis so what this is going to do is it's going to free up time for people to be better at their jobs so right now you have a large number of journalists who just rewrite stories by other people they don't do any primary research they don't call anybody on the phone they don't ask people for their opinions they don't have time for analysis they don't have time to fact check well if the if gpt4 prepares a lot of information for them and then they just go right into fact checking of what gpt4 told it that seems like a pretty good use of time and we have an anal we have an analogy for this rachel um people said when the wikipedia came out well then you didn't need journalists or you didn't need you know college students because they could just go to or college students wouldn't be doing anything with their papers because they go to the page on china they go to the page on semiconductors they go to the page on taiwan and get all the relevant facts well all chat gpt is doing is doing that faster it's going to all the relevant facts and giving it back to you and letting you hit a button that says make this uh easier to read make this like it's written for a five-year-old make this like it's written for a graduate student you know rewrite this say it in uh less words say it more concisely be more verbose right you've seen those tools where you can in notion or other uh pieces of software change the copy that's written well a human used to do that and they used to collect the facts cut and paste them from wikipedia put them into a document and rewrite them and i had to like tell you know inside.com folks i don't want you cutting and pasting rewrite it you can get the facts from other places of course cite the facts and then rewrite it yourself never use the cut and paste function all right well that's over so what can you do you can call people on the phone you can check facts and you can spend more time doing higher level functions so let's say you were writing something about uh taiwan semiconductor and the chance of china invasion invading taiwan and what the impact of that would be but gpt4 you can't trust for that output right now but you could trust it to give you the basic copy to set up the story well now you can call 10 people because you have the time let's say you had 10 hours to write the story you can go call 10 people for an you know a half hour each and then spend five hours doing original reporting so that's what this is going to do i don't think people are going to just start laying people off at their company i think they're going to do better work so the sales team will be able to do three times as many leads which means they should be able to sell three times as much yeah that's what's going

SPEAKER_87: to happen is wild efficiency so this study i think is a little bit to use a word catastrophizing

SPEAKER_12: well we know that buzzfeed is definitely going to be doing exactly what you said uh with their journalists going to be writing things that maybe are a little bit more in depth because they announced that they were going to start doing their listicles which i don't know if you've ever seen one of those where it's like the top five like best pool floaties of the year they're super random um but those are going to be done by chat gpt uh excuse me just gpt technology so that's going to be interesting seeing like this one company um pivot this early on and last week uh on the day gpt4 was launched openai also did a live stream and it featured the co-founder and president greg brockman and it tested some of the features um there's like a 50 second clip that is pretty cool

SPEAKER_218: where it takes like a photo yeah let's play this it's pretty neat i have here a nice hand-drawn mock-up of a joke website so i'm just going to take out my phone literally take a photo of this mock-up all right going to send it to our discord and now we wait and so we can actually take now this output so literally we just said to output the html from that picture and here we go actual working javascript filled in the jokes for comparison this was the original of our mock-up and so there you go going from hand-drawn beautiful art if i do say so myself to working website and this is all just potential right you can see lots of different applications we ourselves are still figuring out new ways to use this so we're going to work with our partner we're going to scale up from there but please be patient because it's going to take us some time to really make this available for everyone

SPEAKER_103: yeah i mean is it impressive or is it not impressive i remember when html wizzywig editors came out

SPEAKER_00: what you see is what you get and people were blown away when apple came out with a word processor in the you know let's call it late 80s 87 80 89 and you could then type something in and change the font and then print it on a apple printer and have different fonts because up to that point you had dot matrix you only had one font and you were like wait a second you would have to hire a printing company to use different fonts now the fonts are built into your computer whoa this is just an extension of that he's using ocr optical character recognition to whizzy wig what you see is what you get build a website uh and then publish it that's not and write the code it's not all that uh groundbreaking when you think about it it's a collection of technologies building on top of it what would

SPEAKER_08: be super impressive here is if it said and then um make 10 categories of jokes that are the most the 10 most popular categories of jokes go figure that out yourself dad jokes you know inappropriate jokes uh you know mom jokes dad jokes whatever then publish those jokes allow people to uh add their own jokes and then have a commenting system and a rating system and make it like reddit now that would

SPEAKER_00: be like whoa or if it just told you hey we built the website for you here are 10 more things you can add to this website would you like to add a voting system would you like to add comments would you like to add a share button would you like to have it produce a daily joke every day um which by the way i bought the website aday.com to do this exact business which i was going to do subdomain joke dot a day recipe dot a day and now that actually ai exists i could probably just build that with chat jpt and have it just come up with garbage and pollute the internet and i think that's going to be the next thing here is the pollution of the internet if you thought social media polluted the internet and spam and content websites and people in manila writing just horrible articles and doing keyword stuffing now imagine like people just take chat gpt and tell it buy 10 domains a day build 100 websites a day produce a thousand articles a day and then interlink beyond interlink between them

SPEAKER_01: the pollution on the internet it's going to be absolutely like a fog of just garbage out there it's going to be like how do we know this is real and this is where we're going to need to have um uh credentials matter so every article knowing a human wrote it versus a computer is going to become

SPEAKER_00: more and more important and who the author so i think authorship now in a world of pollution becomes more valuable you can find a gazillion stories about tech or recipes what's happening now i think rachel is those stories um are going to pollute the internet so much that somebody who's authoritative like gordon ramsey uh or let's say cp from nixfan tv or just somebody myself on uh seed investing you on millennials and gen z culture that authority is and people who do sub stacks or do podcasts they become more valuable because now the the average customer is going to be like wait who can i trust who's a real human who actually is thinking about this and who actually has authority or has wisdom or experience so wisdom in a world of infinite garbage being produced for free wisdom becomes the difference wisdom becomes a difference so maybe you think biology as we talked about in the first half of the show is really smart well now his intelligence or his catastrophizing whichever however you want to frame it is uh or or both it could be both uh becomes more valuable so all of this makes the actual elite

SPEAKER_24: people in the world more valuable because average is now infinite the average production of content the average production of opinions is just a commodity

SPEAKER_12: sam altman the ceo and co-founder was interviewed by abc news reporter rebecca jervis on friday and he didn't talk about um journalists very much but he did talk about his biggest fears with open ai and it's

SPEAKER_238: a pretty wild 60 second uh 60 second clip we we've got to be cautious here and and also i i think it doesn't work to do all this in a lab you've got to get these products out into the world and and make contact with reality make our mistakes while the stakes are low but all of that said i think people should be happy that we're a little bit scared of this i think people should be a little bit scared a little bit you personally i think if i said i were not you should either not trust me or be very unhappy i'm in

SPEAKER_239: this job so what is the worst possible outcome there's like a set of very bad outcomes one thing i'm

SPEAKER_238: particularly worried about is that these models could be used for large-scale disinformation i am worried that these systems now that they're getting better at writing computer code could be used for offensive cyber attacks and we're trying to talk about this i think society needs time to adapt and

SPEAKER_240: how confident are you that what you've built won't lead to those outcomes well we'll adapt it also i think you'll adapt it as negative things occur for sure for sure yeah i think that is a

SPEAKER_24: reasonable thing to say we're putting these tools out here they can't really do too much harm right now play with them everybody has access to them and we have a dialogue about it i think what sam is saying is actually true and i think they need to cite sources say where they got the data from and i think they need to get permission for the data they put into these algorithms i think they're going to get sued um if they haven't already had they probably have a gazillion lawsuits like queued up i think there has to be a massive amount of lawsuits here as well to say did you use

SPEAKER_00: the wikipedia without permission and without credit because wikipedia requires even though you can use the wikipedia days that you need to cite it i think so they're fine with you using it commercially but they want a citation i believe that's their creative commons license and then i think the same is true for open source if you use open source there's a you know you're allowed to use open source but you have to reciprocate by giving back your changes and i think you have to uh do a citation and then there's images and stuff like that that they're just whole sale stealing or people's articles that they're stealing websites without permission and i think all

SPEAKER_24: of that has to be worked out so i think they are in the ask for forgiveness not permission uh which is a tradition in silicon valley whether it's uber or lyft going into a market or doordash or you know name the company that kind of puts product out there and then says let's see what happens and uh we'll monitor it i i don't disagree with him on that i do think they need

SPEAKER_72: to get ahead of citations um and other people are doing citations to me that matters a lot as

SPEAKER_00: a content creator um and i think as a society we're going to get to the point rachel where we actually say um that if you produce content on the internet and it's produced by in part by a language model you're going to have to say this was done by a language model i think disclosure of this is computer generated versus what's real just like in a documentary um they tell you hey this isn't actually people's voices this is a simulation right uh this is a reenactment rather so the reason they do the reenactment is so they don't get sued on the back end so if something's a reenactment and they're having characters do voiceover based on conversations that you know may or may not actually be accurately transcribed they say this is a reenactment based partially on facts this is based on a true story but not actually the true story that analogy should come to articles so buzzfeed is going to do their listicles it should say right at the top this is produced with chat gpt4 and this human this actual here's the byline of the human um this is the byline of the human who uh edited what the chat gpt4 made and you should link to who you should link to the original chat gpt4 citation here's what chat gp4 gave me and here's what chat here's where chat gpt4 got it from that would be a nice um regulate yourself so you can either be regulated by the government and the law or you can kind of

SPEAKER_01: preemptively regulate yourself that would be a nice way to regulate yourself i think and avoid that

SPEAKER_54: i made a bunch of trades j trading's back let's see it okay i see it some of the screenshots were shared in our producer chat and they're pretty you you kind of went off i went off uh yesterday

SPEAKER_01: and just while we're on the show i made a trade so as everybody knows i've been j trading just trying to see if i can become a world-class uh public market investor so i just took a little bit

SPEAKER_00: of money put it into and you go to jtrading.com to see these trades but i bought 350 shares of uh microsoft based on chat gpt i think that they're doing an incredible job executing and putting this into products and they have product velocity and i think these products matter are going to make people more efficient which is great for society so i bought some shares of microsoft about ninety four thousand dollars worth i also bought uh a hundred thousand dollars a thousand shares of google alphabet because i think that they're going to come out with a series of product that will be better than chat gpt4 i'll predict that that their models will be better i think that they have been conservative because they don't want to get sued back to move fast break things microsoft is moving fast and breaking things because they're behind google's conservative because they maybe are winning more

SPEAKER_01: but now they have to change their position so microsoft and google i believe will be the two biggest winners in ai i also think tesla will be as well but i don't trade tesla shares because i'm friends with the owner um and that could lead to down everybody's asked me that question why don't you trade tesla it's like well if you're good friends with the owner that could cause problems

SPEAKER_148: like people would assume that i'm trading on some information that i'm not so i just rather not trade it although my heart would say that they're probably going to be in the top three ai companies in the world especially with their robot so i'm kind of bummed that i don't trade it but uh you know uh that's why and then here's my third trade uh amazon i just bought another 500 shares of

SPEAKER_01: amazon amazon's been like one of my worst trades to date but i saw andy jassy is laying off 9 000 people so this is my standard if the founder or i'm sorry if the ceo or founder is laying people off

SPEAKER_148: that means they want the stock price to go up therefore it's a good trade and this is a cynical cynical trade i'm not proud of it but i uh i bought facebook at 94 when zuckerberg announced his

SPEAKER_24: first layoff and i think that one worked out really well it might be my best j trade i think

SPEAKER_261: and if people want to check these out they can go to just jtrading.com i bought the domain name j

SPEAKER_24: trading yeah oh that's awesome and i don't know if these latest trades have been put in there yet i don't think that they have not but they'll be in later today and um right now according to the website i don't know if this formula is correct right now i am up 12 basis points and the s p 500 is down 2.27 so if i put it in an index fund i would have lost two percent and i'm just over 10 basis points yeah right now my facebook trade is at 107 percent uh netflix i'm up 12 percent adobe i'm up 22 percent apple four percent uber 3.6 shopify 19 percent taiwan semicon 6 percent worst trade in the portfolio stitch fix down 21 amazon down 16 disney down 15 warner brothers down 12 robin hood down 14 and twilio down 15. but i stand by those trades those trades being off 15 is fine with me i i'm doing this for it with a 10 year outlook but i will rebalance this uh nick i want to do like a j trading quarterly show where we look at the trades and then we decide to take out the two weakest maybe and double down on the one or

SPEAKER_01: two best as like a kind of device but do not this is not investment advice rachel i have never actively traded stocks it's the first time in my life i'm actively trading stocks and to say it's active just

SPEAKER_00: educational it's just for educational purposes this is an active active would be every day i look at it and make adjustments i'm making no adjustments i'm only long i don't even have short positions brad garson was like you gotta add some short positions here i'm like i've never shorted a stock i don't know how to do it i don't know why i would do it he's like well if you're long one and then you think this is the loser you put on a pair trade i'm like you know what i just want to be long i don't want to be short stuff i just want to pick things that i think 10 years from now will be better but maybe i need to add some pair trading and some options here um which means i

SPEAKER_72: think i need to really uh spend more time doing it anyway there's been a great show great job rachel uh helping out with the news reading uh for everybody who's in the live chat rachel you got

SPEAKER_01: to jump yeah yes all right okay i'll let you go you're dismissed uh salute to rachel everybody uh give five stars to rachel for her uh coming in and reading the news great job uh good to have

SPEAKER_72: rachel in the fold how do you think she did everybody oh i like her uh great rachel was fantastic 12 of 10 for rachel rachel gets six stars wow people love rachel i you know i had a feeling

SPEAKER_148: um yeah and i'm not going to compare her to molly or whatever but i think they're both great and uh

SPEAKER_01: for folks wondering what happened with molly uh we tried to have molly sort of do three days a week here because i'm moving the show from six days a week to four because i need to recapture some of my time i'm going to be working on a reality tv show it looks like and doing seven days a week of podcasting

SPEAKER_72: was just too exhausting for me um and uh yeah yeah i i uh i won't get too much into the molly details other than to say still remain a great fan of hers wanted her to continue uh made her what i thought was an incredibly generous offer to be co-hosts two three days a week but she declined and um if she and hopefully she could be on the show in the future uh or maybe even sit in for me when i'm not around but uh yeah she's awesome and i hope she finds a great i think she'll probably announce some great gig shortly so uh looking forward to that and she has my hundreds percent support for life okay

SPEAKER_286: so i told everybody i'd take a question from tobias jason do you think more startups will be destroyed by chat gpt then create and and products like it and created by it great question i do think chat i think

SPEAKER_00: um language learning models advanced ones um are on the cusp of eliminating the bottom third of information work in the world maybe the bottom half so i think within the next five years the bottom half of information work will be done by language learning models just like 80 of driving right now can be done very safely highway driving by autopilot yeah you know and full self-driving and then it just depends on us getting comfortable with deploying that so how many people have cars that have full self-driving

SPEAKER_08: that can do adaptive cruise control and lane control on the highway i would much rather right now drive on a highway if you told me i could pay you know uh ten dollars uh per hundred miles to drive on a highway that only had cars with full self-driving you know which is adaptive cruise control and lane

SPEAKER_24: control let's say i would pay that i would pay a premium so i would pay 30 bucks every time i went to

SPEAKER_08: tahoe or did a 300 mile drive to la i would pay that 30 bucks to go the 300 miles to la uh in a in a separate highway only with those cars and then that would subsidize everybody having one of those

SPEAKER_24: cars or retrofitting cars to have it why would i do that because the safety would be unbelievable now would i do that on the streets of san francisco would i do it on the back roads of lake tahoe no i don't think the technology is perfected there i think there's a lot of random things that could occur like snow ice uh lines those highways are just not maintained properly where those roads are not maintained to the way highways are so let's take that same analogy to uh language models would i want to just read an article right now on an important issue like trading a stock on uh quarterly reports

SPEAKER_00: or would i want a human to do it i would want a human to do the quarterly report thing now would i take an article right now on how to um ideas for a 10 year old's birthday party sure like what what's the downside give me 10 ideas for a 10 year old's party give me 10 jokes there's no harm if i read the jokes and they're bad dad jokes and two are good and eight are terrible and don't make sense i'll just take the two if i get 10 ideas on what to do for a 10 year old's birthday fantastic i take the 10 ideas two of them were great oh do a pirate ship thing oh go to the beach and hire somebody to collect shells and put nets out and identify fish oh it sounds like a great idea for a party uh a birthday party great i can then evaluate it from there but i wouldn't take stock advice

SPEAKER_01: on should i j trade amazon at this point because as we've seen probably not a great idea uh so

SPEAKER_72: they have it folks uh that is what i believe will happen and then i think yeah if this i think the bottom

SPEAKER_00: half of startups or bottom third of startups won't be necessary because language models would do that but i think those same founders will find something else to do the next higher level thing so if in the example of the birthday parties well okay here's the 10 birthday party ideas two of them are good there'll be a startup that will be a marketplace of people who could throw birthday parties event planners etc and they will be even better parties because they'll have more creative ideas and they'll

SPEAKER_01: come together together quicker for a lower price just as one example