Welcome to the cerebral valley podcast on newcomer. I'm back with Max Child and James
Willisterman, my cerebral valley co-host and the Volly co-founders. We are very excited
about this episode. The meat of it will be a draft of the most valuable AI companies we
get. It is like, it's as close as I think in Silicon Valley you can get to a watching
sports, hopefully, or fantasy sports sports debate TV show. Yeah, exactly. Yeah, I over
complicated it to make sure I got the I won't spoil it. Anyway, you know, we took it very
seriously and even if all our listeners forget about it, I am sure James, what was the time
when it was it's five years? Five years. Yeah, in five years we'll all be checking to see
how we did. But before that, cerebral valley, the conference, is on Wednesday. And, you
know, obviously for the people going in person, you get why you're curious. It's like,
okay, but then for everybody else, the videos from the conference will be posted on YouTube
and in the newsletter newcomer.co. So this conference will be brought to you digitally. We
have not just hyped up an event. You will not get to consume. You will get to see the
talks. So look forward to that. I think usually we put some of the best ones in the podcast
feed. And we just wanted to sort of take stock going into the conference, sort of big
themes that are going to be on our mind and then what's changed from the first cerebral
valley in March. So yeah, let's get to it. All right, I said you guys could go for themes
last. I mean, go for themes first and I would play clean up. James, why don't you go?
Still thinking about my themes. Yeah, I have noticed a theme kind of gaining some heat
over the last few weeks on Twitter slash X of kind of this big debate around open source
versus close source in terms of safety, right? I think this ties in with the executive
order put out by the Biden administration and kind of a raging debate, I guess, or simmering
debate of whether these larger close source language model companies are in the business
regulatory capture through their efforts to speak around model safety and get the Biden administration
involved or whether that comes from a true concern and how that impacts companies who are
putting out open source models for the community. Okay, I'm going to jump on that open source
versus close source. I think that's probably the top in my mind. And in some ways, I'd almost
frame it as like I think the first conference ended up having more of the pro open source people,
you know, we ended with hugging face and replete both of them are very pro open source. I mean,
we're going to have Databricks back. They're pretty pro open source. Whereas this conference,
we're ending the day with Vinod Kosla who has been somewhat critical of he wants more sort of
control and regulation. We'll get the exact nuance of his open source take. And then Mustafa
Suleiman at inflection in the D mine co-founder who's going to end the day has been super supportive
of regulation and somewhat apprehensive about open source. So I think this conference in some
ways will be the response to the really pro open source of the first. Of course, I think open
source is very popular in the not open AI world. So I wouldn't be surprised still if probably the
majority of our panelists are still pro open source. Yeah, what do you guys think? I guess one
thing I wanted to ask you guys on this topic is last time at the conference, a lot of discussion
was about the letter that just came out with a lot of signatories around taking a pause on
AI development. A modesty ensigned it totally perplexingly. Yeah, which was going to be using
to a lot of people. And I guess my question is at the time it felt like there was a lot of heat
on companies even like open AI and Sam Altman for being cavalier with their development of AI
and causing an AI race and putting out chat GPT and before the bigger more cautious companies like
Google had released their AI's. And I really think in like six months that has just totally changed
where now that you have this open source community putting out models like open AI is able to look
like the responsible stewards of AI going to the White House and getting regulations in place.
Just curious if you guys agree with that assessment.
I do think a lot of the safety talk has died down. I guess the executive order is maybe a
contrary example of that, although maybe the White House timeline is like six months behind
everyone else's timeline. But it does seem like people are a lot less hung up on like
is the AI going to kill everyone in sort of the tech industry. And then I guess simultaneously we
got this massive executive order banning or not banning but regulating models over a certain size
etc etc. So it is interesting to think about if there's going to be this sort of waves and
and troughs or peaks and troughs of the hype cycle around safety. And I feel like we're in a
little bit of a trough inside the industry right now. But six months from now that could change
again. We just got to keep having conferences. And March GPT-4 was like you know a shock wave
and it had come over 3.5. And I think it felt like man if this keeps going like we're going to
have generalized AI tomorrow. And then I nobody is really caught up with them right there are
open source models and people can say oh this is better than open AI on cost. But I don't think
anyone even in specialized use cases has really had anything as smart. So then there's just a
little less fear of like oh my god it's becoming like so you know that is just not sort of
the pace hasn't been so out of control. And I think that's part of what is cool things down. But
I sort of I mean I think the binding executive order is like a huge deal and shows that there's
very active sort of government interference and like Europe has been talking about things. So I
don't know I think this question of cracking down on open source is very real but yeah but the
open source community has certainly been vibrant and and you know and the other piece of that is
that firms like Andreessen and others have been very loud about sort of positioning themselves as
super pro open source. Yeah interesting next theme all right I have a theme for you guys to think
about here. I think one of the all time maybe the number one all time Silicon Valley investing
philosophy for all categories of software is you know we like to invest in companies that sell
picks and shovels not the companies that dig for gold right we like to we like to fund the tool
makers not the tool users in many cases. And I think what's interesting just in the last six months
between the previous conference and the current conference is the the sort of what tools were
invoked six seven months ago and then what tools are invoked today. And then the sort of cherry
on top of that is open AI itself you know very very recently announced this kind of huge wave of
first-party tools improvements whether it's you know retrieval improvements context increases speed
cost releasing anyone able to make their own GPT character to kind of go after the character
eyes of the world like they really came hard on like a bunch of tools categories that I would say
six seven months ago we're being funded as like pretty compelling standalone companies and I think
it's sort of a two-way question which is one is tools investments still good idea in an AI world
like or is open AI going to eat this all alive and then I guess two like you know who is who is
still exciting in the tool space and and who do we think has like a strong foundation for the future.
Yeah I mean I think that question of will open AI just eat everything applies in every part of
the game. Yeah I mean that was the first conference definitely just like how how relevant is
everybody except open AI and sort of an ongoing question. So certainly yeah I think now with the
developer conference this thread on tools and then I think you know given a chat GBT is clearly
the most popular consumer AI app. Well I'd be interested I mean you have authentic conversations with
venture capitalists regularly Eric right behind the scenes that don't all get published like yeah
do you ever get some authentic expression of concern about tools companies that they've
backed in the space that were like well maybe we jumped on this tools idea too early and it actually
isn't a standalone product or you know I was behind the scenes yeah there definitely a lot of firms
that are shitting on foundation model bets. Oh interesting. Oh yeah that's like I mean they're
expensive you know they're very speculative open AI's in the lead they feel like even with
open AI open source comes to parity very quickly so the category I hear the most dunking on is
just like foundation models and you'll see firms that haven't made a lot of bets there you know
yeah the valley loves like tools picks and shovels so sure that's not something where I feel like
I've heard a lot of regret anything I feel like I mean I've been talking like vector databases
all week and stuff like that and literally you're getting VCs talk about chips or getting companies
around chips which traditionally they've been scared about but you have sort of the fact that
Nvidia is doing so well and then you also have the fact that it's sort of cool to be like an
emerging tech investor at the moment so I think those things combined you're seeing that so no
I think VCs yeah dunking on foundation models but bullish about everything else supporting AI
yeah are there particular company types you have in mind like that you think well I mean I
think there's like obvious companies that are like maybe less hot than they were in the last
conference right I mean like Jasper Jasper would be an example right stability would be a huge example
right but I think there's company that's just like the unique feelings of that sure sure but there
are companies that we talked to last time right you know Langchain for example right that like
it's a big question as to what where Langchain fits in the universe where open AI is just
going to take care of all this crap right and then you know I think we're talking to some companies
this week that you look like they they could totally be awesome companies and I have no earthly
clue but they're providing very specialized tools like you know data cleaning or something like that
right like which again I'm excited to hear the pitch like of why that's a great idea for the company
and then probably is but like you know I would be a little scared that opening I was just going to
like kick ass in my category if I were a tools company and so I just think and I mean I think you
have to agree that character has to be pretty upset that opening I basically announced their
version of character AI tools right like so I'm just interested if there's any real feedback of
the fear factor on that front but it sounds like maybe not so yeah do you think I just jumping
diving in on the character question like there is a theme I've noticed of people kind of questioning
whether open AI for consumers as an endpoint is is really sufficient so there's kind of like the
opposite side of that question from the tooling side right like will they eat all tools and then
there's from the consumer assistance side is there a super assistant that's just chat gbt or do
you end up using all these characters or mini gpt's that they are working to enable developers to
build how do you guys feel about that debate yeah I mean I think that's that's the open question I
think it depends on the quality of the you know the foundation I don't know I mean I guess like
selfishly we build games james that are built on AI tools including in some cases large language
models so my selfish belief is is possible to build games on llms without open AI building the
exact same thing it may depend on the complexity level of the consumer app built on top of the tool
whereas characters sort of pretty vanilla re you know reprompting or you know tweaked version of
the core gpt experience but you know I think it depends on the complexity level and sort of
how much of the stack you control and and how easy it is to get to your customers I mean a phrase
I've been hearing over and over again this week you know just talking to these sees has been
retrieval augmented generation rags rags and I think there are a lot of companies right now that
think we can be really good at giving gpt the sort of information it needs and then you know open
AI or whoever can sort of do the thinking based on that but we're not we're not in the business
of making the best thinking company but we're we're really good at figuring out like what it needs
to consider and and it's possible that there are a lot of what kind of companies you think are
saying that around like I guess this is kind of pointing to data and unique data sets as kind
of a mode of some kind right like what kinds of companies you think fit that category of potentially
having a true data mode that could be valuable having a data mode yeah well um
well this isn't a data mode but I mean I'm one company as you'll hear in the second half
isn't it a data isn't it a data mode that you are saying that these companies are talking to you
saying we have well like okay there's a company exclusive clean right that let's companies run
searches right is it a data mode they're really like good at hooking companies up companies data
got up to data sources that then connect that use rag and then yeah plug into something like
open AI so it's honestly like a Bloomberg where it's like we have all the financial data in the
world it's we're really good at like getting the relevant data four companies into into this
system is is one case but at least that open AI dev day this week they seem to be going after some
similar use cases of allowing cost companies and enterprises to build their own internal
GPTs or chat GPTs based on internal company data so I think that that is still some area that
they seem to be competing in as well I mean you know eventually there's only so much one company
can do unless unless it's like the I feel like James this is your style of suggestion throughout
this series unless there's AGI and like every way we talk about this is wrong but
in any sort of traditional business sense where like you know say a moment you know has joked
in the past they're like oh we'll just ask open it we'll ask chat GPT at some point you know
like what our business model should be and we'll do that clearly that they didn't need to wait
for that they figured out some business models but you know I anyway my point is just like in
a traditional normal world that I expect to continue you can't do everything and so whatever
they're going to have to pick their strengths and then everybody else can do other stuff
I'm going to I'm going to give a last a last theme that's sort of more in the fun sort of
philosophy sort of world you know if it any into the AGI question which is just how seriously
do people take existential risk in it you know is existential risk now seen as like this is
propaganda for people who want to make it one seem really awesome and sexy and amazing it can
do everything and then two people who want to be in this world of like shutting it down like how
much is existential risk really on people's minds anymore or is it much important that we just
make sure it's not racist and make sure you know deal with these like very practical concerns that
are obviously true now versus imagining this world where it's you know you know like our AGI
Doom episode you know figuring out how to optimize paper clips into blowing up the earth that's
kind of eluding to my earlier point about I feel like Silicon Valley has stopped taking safety
as seriously right I mean like I think they like yeah sure the executive order is not an
example of that right but in Silicon Valley I think the discussion of to your point the implied
seriousness with which people take existential risk seems pretty low to me because otherwise people
would be like blowing up GPU factories or whatever I don't know it's like if you actually believe
this had a 20% chance of killing all of society which a lot of people say like how could you
morally work in this space or at least not like you know be actively opposed to any of the
development here right and and the only people who seem to really act as if they truly believe
existential risk is real is basically you know Elisir Yudkowski and the sort of the A.I. Doomers on
Twitter right who are really out there every day saying there was there was an interview shout out
to Logan Barlett with his podcast you interviewed Dario the co-founder of Anthropic right and Dario
gave like it was like 20% chance like everything right it was like 10 to 20% chance those like really
high I think is that like in the next 100 years though again the timeframes matter like I think it
was it was reasonably soon okay but yeah I don't know I sort of think you'd have to be insane to be
working in an industry where you truly believe that progress there had a 20% chance of ending all
humanity I don't know like unless you believe in some complicated game theory about deterrence or
something and back to our earlier episode about you know the only thing that can stop a good guy
with an A.I. is or sorry a back guy with an A.I. is a good guy with an A.I.
I said I don't care but the only thing that can stop a bad guy with an A.I. is a good guy with a
better A.I. right yeah unless you can explain some pretty elaborate game theory on how progress an
A.I. by the good guys is necessary to stop extinction it just you know it doesn't compute the you
believe this stuff has 20% chance of killing us like you know and be working in the industry at
the level of Dario who literally is like the number two most powerful person in foundation models
right I mean it's just it's nonsense so I don't know I agree it doesn't seem like people take it
that seriously for real as we discussed in that episode though there's a lot of like very bad
things that could happen you know A.I. kills some of us right now yeah that was the that was the
I guess like maybe maybe people have come around to sort of a middle ground here where it's not
existential risk like extinction for humanity and more that there's a lot of bad things that could
happen over the next five to 10 years with A.I. and I think all of us would agree with that right
you know there are bad things that could occur if the you know some bad actors got access to
some really powerful sure models or something right I mean I think you the fact that Cruz is
not active in San Francisco like even if they sort of bungled their disclosures to the government
I mean Cruz is a better driver than humans in San Francisco right like they really they they
they broke the one rule which is don't lie to the government right I mean like it's just like
you can be you you can be bad you can screw up you know but don't lie to the government is like
freaking like number one on the 10 commandments of running a company right I mean like you know
that's why sbs travel whatever you know I mean like like just don't lie to the government like
you know fake it so you make it can apply in many scenarios but not with the government there
you can't do that with them so anyway can we just go back to this for a second say years
you're what are you saying Eric about um existential risk or max like you guys are saying
companies are not acting like they believe their own hype or do you agree with that or I think
I think people are now saying calling bullshit in sort of private existential risk and some of
them don't want to say it publicly because they don't want to be seen as sort of undermining
legitimate concerns but I think a lot of them are just like doesn't feel close and then how do
you feel that ties into the executive order does that feel to you like over regulation I'll
you know pre pre maturing like I've been sort of waiting until after the conference
okay to stake out make a state position partially just like get get more information also you know
I sort of shift between information gatherer and sometimes opinion havern I don't know I've
been sort of waiting to come out strong though I do have a sort of side the last theme I'm actually
interested in and this is very close to home for James and I in the last couple months is
we are increasingly discovering that the cost of running any large language model products at
consumer scale is unbelievably high yeah yeah is so high that it's it's very hard to come up with
a mental model for how you could build anything on top of a GPT4 quality model right now that would
be distributed to consumers you could maybe do it on three five especially with some recent price
cuts with you know fine tuning yada yada yada but I would say that it is shockingly high to run a very
basic application that's GPT4 level quality on LLM and I think it basically makes it impossible
to build consumer apps on these level of models today without incinerating money and so I think
it's an interesting question as to whether or not a lot of these companies that have raised $100
or $200 million dollars I have to imagine their unit economics are pretty much upside down if
they're building any sort of consumer products as in every time someone uses their product they lose
more money than they make so I wonder if the only outcome of this is that B2B companies can maybe
afford this because they can charge such staggering prices and then open AI is like the only one
that can do consumer stuff because you pretty much have to have no middleman to afford running
these applications okay so that's a great one and I'll broaden it and just say I think there's
an overall metatheme of that is the rubber meets the road right there's this do people really care
about existential risk anymore because it's like oh we're just in the practical problems like
it's really expensive to run consumer apps and I've heard people in enterprise ask like
are our big companies customers sticking around or do they try the demo and say we didn't like
it anymore and I mean you know Chachi BT I think we're going to see in a slide during the
presentation you know activity sinks and then they have new features and goes up like what's the
resiliency so just this yeah the rubber mean the road all right we've gotten all the hype like
how are people delivering and now it's sort of the moment where it's turning into sort of more
regular sort of business style questions I mean this stuff is so expensive to run that it's
hard to imagine you would not have to run into regular business style questions pretty quickly
now right I was not in the internet industry in 1993 when you had to go buy your own servers or
on a website you know so that this may be analogous to that time you know where you had to be a garage
band server farm to afford to run these kind of companies but even that would have been that would
have been fixed cost right now talking about right we're talking about variable cost more customers
you get more expensive yeah I raised erasing software margins like so like I'm pretty interested
to see if anyone can run a consumer product without just incinerating money besides open AI so
stay tuned or or use I would say I would say the last caveat would just be using smaller models or
yeah using smaller or smaller or fine tuning more or something or gbt4 gets 20 times cheaper in the
next year and a half and then this is a dumb irrelevant you know retrospective discussion
but it really has to go down by like a 10x to be interesting so they did 3x so we could go we need
two more of those to really be in a good category of cost yeah we should I mean that is a good
reminder that we should ask money like how much does this cost yeah where do you make money if like
you know prices change one way or the other yeah all right next stop I really think you're going
to want to listen to this the draft pick please give us your feedback I want insane like memos
about like what the order strategy should have been and just feel free to weigh in we did you know
I think we explained all the particulars in in James right you talked yes yeah we're about to
we don't need jump into the rules and there were there was probably more time spend behind the
scenes debating the rules then I'm preparing for the draft I was insistent about a particular rule
change that you know I mean you'll hear so we had a lot of fun sent you Eric at newcomer.co you
can always send me your opinions about who's an idiot I hopefully I think I'll try and post the
list so please forgive us that it was a set list there are companies I asked if I could pick
companies off list is sort of an insider with Intel but I was I was not allowed to pick off list so
I you're restricted to a list of companies and we didn't pick you please hire new marketing officer
I don't know sorry yeah all right find out why your your valuation was also not you know
listed in pitch book right because of the list was selected from companies who are
register who have have a recent over one hundred million dollar rounds in pitch there's go listen
all right we are going to move into the draft we this probably the thing we're most excited
about at least informed to do taking a page out of sports podcast political podcast playbook
and we are going to do a draft pick of the top AI companies James do you want to explain
the metrics for who wins this and how we know and all that here's what we're doing Max Eric and I
are drafting teams of AI startups the startups must have risen over one hundred million dollars
I wish that was in the case partially because I've Intel on like companies that are promising
but James isn't letting me include random I feel like there's an you know it's a risk they are
not even valued at that yet but I yeah I think we have we have 43 companies in the list
and we're only picking five each so many of these will not make it on the board so I'm excited
to see what happens but this really shows I think the amount of companies who have raised ginormous
rounds in AI is this a hundred million plus list you have to have raised raised over a hundred
million not valuation raised it's probably known I because I do think also there probably
ones that aren't so some other criteria I used to kind of narrow down this list you had to be
using generative AI or the company had to be have a generative AI or generative AI
infer a kind of core to it I kind of excluded some industries like bio and health care defense
silicon chips robotics driverless or AVs and China I just wanted to filter down to companies we
might have a chance of knowing something about or being predictive about rather than well just
yeah predicting whether some biotech company is going to use AI to discover a new drug or something
and I also wanted companies founded before excluded companies founded before 2013
so most of these companies have been actually founded in the last three years or something
but there are a few who have been founded maybe earlier so that's the game and we are going to
determine who goes first by creating a handicap auction so I was adamant that we do this James
and Max were like oh you know that's part of the remnants but to me like a huge part of this
is how how much would you pay to get to pick open AI because I honestly think we're all going to
pick open AI we should talk about what the what the grading criteria is yeah here so the way we are
drafting these teams we are we are trying to create a team with the highest total valuation on
November 1st 2028 and final valuations would be determined by pitch book either kind of a
reported M&A price or a current stock price if if the company is public or the last valuation
if the company is still going but is private and if the company is shut down I think that would
count as a zero for that team if like two years does it have to have raised at the price within
the last two years or something because there's just whatever the last fundraiser was yeah just
but like it might not be worth that anymore one question I have is let's let's say there's a company
on this list who has raised who has raised a huge round maybe in the zarp era but never raises
again but is a zombie company in five years are recounting that at their I think the last
valuation no it needs to raise another round or do something in the next five years because otherwise
there's a strategy of just like picking the most highly valued companies and just betting that
like nothing happens that's the efficient market hype out there I don't think I think it's
theory something should have to have current valuation other is the best part of future value needs
another either of us it needs another fundraiser reporting yeah I think so Max are you okay with that
or I mean I don't care I don't think it's super relevant so whatever yeah it needs something whatever
right are we agreed on that yeah we're agreed so if the company does not raise any round between
now and November 1st 2028 yeah according to pitch book a new round they will count as a zero as well
fair okay yeah brutal um so how are we doing the bid for first and second place
well I think the the idea here is if you you know we're gonna just go down the line here and
name a price that you are willing to subtract from your total score and then in the end and then
determine you know the order in that who in that order of those three handicaps and we will
deduct those scores at the all right let's just do a live auction like whoever gets the highest
price and then we stop it okay so then it's a coin flip okay fine coin flip for a second yes
okay let me pull up my digital coin flipper okay all right and then it's a snake draft yes all right
let's let's start I bid negative 45 billion for open AI for first to go first for first take okay
I'll go I'll go to 50 sure I'll go I'll go 55 all go 60 60 billion negative 60 billion
how is opening I gonna have an exit I just some confused by this it doesn't they don't need to have
access they just need to have a fundraise okay I'll go I'll go to 70 75
okay that's fine max you're done I'm done I'm out you got it at 75 all right okay first pick
you have a lunatic open AI okay bad pick bad pick now I just I just think the number of
companies that are worth like hundreds of billions of dollars on the secondary market is just
extremely low so you're basically discounting the net equity value to like five or 15 billion now
or whatever it is so sorry yeah it's currently valued at 80 or 90 right on secondary markets and
you took 70 75 off that so right you're you're basically saying I'm you know 10 10 is where I
it's my starting point and I have to get another hundred billion or whatever on this valuation if
it's it's a trillion dollar company like yeah that's yeah then I crushed it like but are they
gonna be a trillion dollar company as a subsidiary of Microsoft I just don't believe that but
that's fine this is why it's a exciting contentious pay yeah great okay all right
Eric's gonna go first and Max and I are flipping for second place Max name your header tails
here I'll take heads heads you're up second okay Eric we know what you're picking do you want to
make it official open AI congratulations new what he was doing smart guy smart guy anything you
want to anything you want to add there or what I do also think the fact I think Microsoft
be an investor helps drive up the valuation because they're getting spend back on their system
there's all this revenue round trip and going on I think it'll definitely raise another round or two
so I feel like it's super safe bet I think I'm getting it if we know it's reported to be 90 billion
I'm getting it like yeah with but if Microsoft buy if Microsoft buys an additional 2% to go
from 49% ownership to 51% ownership that's that's like an exit in my opinion like oh that's an
exit I agree yeah yeah what do you know no majority ownership of open AI I mean it's it's a
scenario Microsoft at that point that no it's just over then like yeah I think no these are deals
on the way I value it as more it's worth more definitely if other people if people are getting shares
at a higher price there there will be no shares of open AI though in this context I don't know we
can we can see what actually plays out in the valuation goes up and there are people who own a lot
of it then yeah it has no valuation there's no way that you're saying well well I guess the shares
have to dissolve is what you're saying for it to be an acquisition that's right yeah once they're
in Microsoft I accept that like I don't get to write up the Microsoft benefit okay all right
even though I do think that's a possible outcome but I just don't think they ever make it to
whatever a trillion dollar valuation I don't know if Sam Altman would sell oh you know
away control to Microsoft I don't think so either and like there's anti-trust and like I think
this is the most Microsoft is probably gonna own but I I'm not willing to accept like if they own
60 percent I'm sure no that's fair that's fair okay yeah all right all right all right max
I will take Databricks at number two yeah smart pick smart pick like I like Databricks
I I like to think that I mean the goal here is obviously to really get a couple huge winners
or one really huge winner right and I think that Databricks is obviously already a very large
company you know tens of billions of dollars and I think there's probably many tens more or
potentially hundreds in headroom there and so yeah I don't know I think it's a good call option
on a really massive company so that's when grabbing Databricks and also my favorite speaker from
cv1 and Ali go see the CEO and he's coming back for cv2 so I'm just a little homer pick here
smart yeah this is giving me now my stomach's turning I do Databricks like I've been super bullish
about them for the long times who not to have them in my list as it hurts all right James
all right pick three all right I've got two that I'm debuting you get two in a row snake draft
you get two in a row I got two okay I just do both of them yes all right I am going to go
with hugging face as my first pick wow that's that's high that's high I think you could have gotten
the value bet later in the later in the draft there no I might have sounded like Eric was going
was I'm hugging face as well yeah I think I just I just really think hugging face is going to exit it
under 20 to Microsoft basically so that's good for me though that's like a pretty good lock-in
of 20 million you need to stack those hundreds to win this draft James we're going up against
the trillion dollar open AI here so then we should only pick foundation model companies you should
pick whatever you think can go to a trillion Eric you know I think anything hugging face has a
lot of room for I think potentially getting to that hundred billion level if they kind of maintain
with their position in in essential you know position in the open open open source community and
they potentially even build tooling around this that helps other companies use open source in
their in their enterprises and yeah I just think there's a pretty high ceiling there get hope for
seven point five billion yeah I'm just saying get up is is everything hugging faces and more
in my opinion right they decided to exit at the top at seven and a half so I don't know I mean
I love hugging face for the record we're you know big fans but I don't know not sure about the
pick on the upside and there are four of them at the end of the day like yeah all right James second
pick hosting me I mean we love but yeah for my second pick I am going with anthropic huge second
runner in the foundation model wars I would argue just recently struck a pretty major partnership
with Amazon and has a lot of the same people who are originally working at open AI so I think
yeah and I've heard really you know positive things from developers about cloud in general so yeah
I'm really excited about anthropic okay this is a tough pick I'm gonna I'm gonna skip through the
really high fundraisers here for my second pick go a little more deviant I'm going to take
pine cone the database company I think that what I'm hearing from my little birdies on the ground
the little birdies in the in the AI communities pine cone is the the database of choice for AI
developers of all stripes and I think if you look at the history of Silicon Valley you see really
strong exits from a lot of different database companies so yeah I mean I like pine cone as I don't
know I mean this would be a really good pick if I could get you know discounted for the fact that
their evaluation is way lower than most of the companies on this list but regardless I'll take
the upside I you know as a reporter does went went to a VC for advice before this oh yeah nice
I asked at the last minute but he said pine cone not there on his list he only gave one
of these five because I think I heard number three enterprise are wasting tons of time building toy
prototypes that will ultimately be replaced by apps so all I all I care about is just the crazy
upside case and I think that's that's what I'm going for here so all right I'm gonna do I get to
now I don't know you get to know you get to yeah you get to now you get to swing back I'm like
nervous I'm going to take inflection please explain you got to defend your picks here yeah
with the idea one sort of top town obviously like deep-mind co-founder Mustafa Surleimon I'm not
picking in music speaking but you know I I do think super legit company I mean obviously
you're betting a lot on the future I do think you know if it's go for big swings huge foundation
model potential is key smart technology I think I think also just like give I feel like there's a
floor and that it's also like a potential like acquisition to just like get some big tech company
capable worst case so yeah I'm going inflection for number two okay and oh this is tough one
all right I'm taking character AI I just oh man I feel like that I know bad take bad take
you would have taken it right yeah bad use you would have taken it's the take bad bad
no no bad take is just a way joke about every pick being bad sorry would you have taken character
like I never would have but it sounds like James would have so I think I would have taken it yes
I'm going on a foundation model forward strategy and I think character you know has perhaps
a clear path to a product than most I think I mean they have a very talented CEO
I mean Theresa Horowitz is going big on them so they're going to get lots of follow on funding
rounds yeah we'll see I don't know yeah okay my friends I you know something I I still think like
the you know right now the bot is a relationship bot thing is a little overhype but you know
there's time what do you think about character as it is today what is your it do you agree with
the take that it's all sex chat or or not I don't I don't know I'm not like an expert on the
product but to me there there's sort of you get a dual bet in that they're super hardcore building
their own foundation model deep on technology and they're actually have a product that's out there
that people are finding fun ways to use and like I mean you know marks like we're with hot or not
so like anything that they have the sort of instinct to say like let's get users that that to
me is like a good good muscle and like even if this first iteration isn't like the one you want
the New York Times to be writing about I think I think it's a good sign I think it's a good sort of
out of the money you guys were not a collection though no I think it flexions are reasonable but I
like your picks I like your yeah I don't I honestly don't really know these foundation models
which is why I'm gonna pick a foundation model next baby yeah I mean I literally think
I could be choosing foundation models past open AI and anthropic as a a monkey throwing darts
at the proverbial board would do better than me here but I'll just go with co here I'll just
take the market deciding it's worth a lot it seems like it's a number three number four best
foundation model who the F knows is my defense of this pick but I got everybody's
ready I got to grab a foundation my everything wants to be in business so I feel like they're
gonna face a ton of competition they were early sure um yeah I'm not I'm I'm freely willing to
give them I'm I'm freely willing that does my lowest conviction pick here but I just feel like
I got to have a foundation model and that's me is the best pick on the board so so be it
onward and upward maybe I'll get lucky okay I'm ready with my two picks all right please far away
all right first pick go going with another foundation model of course gotta have one I will say
you can see why VCs are investing big in foundation models because if the goal is to go for the
fences yep you're like you know foundation models are are you know big swings and so like
I'm nobody here is really so far trying to accumulate like about like I could name some sure
things but then you know you want like a swing anyway James I'm going with AI 21 labs which is a
Israeli foundation model company solid just just announced a $155 million raise in August at a
$1.4 billion valuation and the participants included Google and Nvidia they create foundation
models particularly targeted at writing and they are known for their cutting edge Jurassic models
so yeah I'm impressed with the founding team and the product yeah and they've got some I think
they have a partnership maybe with Amazon like I've seen them come up but it feels like they're
under the radar a little bit I think it's a little bit of an under the radar yeah pick but a solid
team building foundation models yep I mean I would say that is you know if if they are Israeli and
they're valued at $1.4 probably in the US they'd be valued at $14 billion so I think yeah exactly
like you know I that that seems like a good good good cost of living adjustment right there
all right and with my next pick I am taking replete oh I'm John yeah another CV CV AI alumni
alumnus if you will so replete you have ugly face and uh replete they're on the same panel
together uh you're just a you're just a big CV AI chapter here I love it I like to meet the founders
and get to know them practicing for uh in the end it's really it's really just a bet on the
founders it's really about the people it's really about the founders what do you think about this
stuff yeah founder market fit if you will all right what what you're actually as founder twitter
are our replete logging phase like in the like trillion dollar case they're like fighting with
each other right they're both like in the coding world I don't I don't agree with that I think that
replete the trillion dollar case here is that software engineering changes dramatically due
to AI and we enter this world of everyone's an engineer you you you're using a cloud-based IDE
you can use AI really aggressively as your coding and you don't even need to really learn how to
code and I think they're the ones who are furthest along it kind of building that vision that I
do believe is possible with AI that was that was not on my draft board uh all right uh yeah
I don't think I would pick it yeah um I'm gonna go with another homer pick uh I will be taking
modular uh another founder I will be interviewing on stage at cva i2 for the for the back store here
yeah for the back store here Chris Ladner who co-founded modular literally invented the swift
programming language that is on every iphone every mac every ipad in the world probably one of the
top five to ten most used programming languages ever invented uh so he's obviously a beast he has
shown tremendous facility in creating really low level technical products that are used by literally
hundreds of millions of people if not billions before that he created a tool called low level virtual
machines that is also basically everywhere so I'm just saying Chris Ladner gets w's and he is
trying to reinvent the programming language for a i and replace python with a new language called
the mojo that modular is invented and so look I frankly don't know what the business model is here or
where we're going with this part but you know Chris Ladner puts points on the scoreboard so that's
what I'm hoping he'll do for my team here so modular AI I am breathing like a huge sigh of relief
because I took character instead of this company thinking the character was going to be a more
competitive pick and I think we're right proved true and this company did not get scooped up
I am taking what I would have actually probably put as my third pick but strategically dropped to four
clean which is sort of doing corporate search secoy about company light speed rovy the key guy
at lightspeed is on it and memoon homit a cliner perkins is an investor in this company so I just feel
like it's one of these like I don't know what do they all know uh so you were you were really like a
venture capital fundamentalist here like I like I like I was like you know Eric's like I've ever heard
of that fallacy of appealing to authority well I go the other way you know like you should always
have field authority I mean what we don't have the financials what are you doing you're looking for
signal like I mean what is better signal than like secoyamamoon and rovy like I you know I don't
know what it is I think they're good vc's and like I I don't have metrics so what am I going
and honestly my vc guru that I went to here's his case for glean semantic search over business data
is the biggest use case this opportunity over shadows all the in house builds people are using
link chain etc for so instead of instead of cobbling it together yourself use glean like I buy
that wasn't wasn't this titanium pitch which also like you know it's sort of oh yeah
around you know you need to look into the eyes of the covenors are like are you going to actually
help your investors make money out of this company um anyway um okay well there you go okay so
now I have a last one this one I'm debating between like product that's solid protect my sort of
minimum mm-hmm versus like upside potential man this is like I I have empathy from vc's it's like
should I go crazy like hoping a dream you know more should I go sort of like product that people
actually use today I'd like to think you would do more than 10 minutes of research which is what
I don't know did do vc's do that I don't know though they get they get the financials we don't get
any I guess some of these companies have a lot you're you're you already did what most vc's do
which is ask their vc friends they shouldn't invest in just go with that it works uh lots of
people say what's he doing that um all right I I'm going for the the hope and dream of the entire
European continent mistral AI oh my gosh I thought you were gonna say something else but yeah yeah
Vladimir Zelensky okay um all right mistral is you know super hype I think light speed is an investor
European um you know they they're very early super super hypey but uh you know smart smart founders
I'm really enjoying this draft because I really would not have picked almost any of the companies
you guys picked and I feel interesting that must apply to me too so yeah we can review each
other's teams at the end here um I know all right six would be so I'll tell you I'm gonna I'm gonna
I'm I am also gonna go hope in a dream here a little bit more um so I obviously read you know
read and researched all the founders of these companies and I was very taken by this founders
vision of self improving foundation models so the idea of recursive improvement of a foundation
and so that was the pitch of imbu our friend can you know yeah uh who on a open a dream and I
assume some really solid demos has raised $234 million so so but I think if you're going for
true takeoff trillion dollar companies as it were the first person to figure out how to make AI
recursively self improving will has a good shadow being a trillion dollar company so I'll I'll
I'll put some some chips on Kengeon and see what happens yeah I think they're very bullish on
like agents right I mean right and self improving agents as well which I think was exciting um so I'll
throw I'll throw five dollars into that 234 million dollar pot and help them get there right James
final pick final pick all right with my final pick I am taking adept
mmm interesting very highly valued cerebral value one yeah obviously super high valued
and highly funded as far as I can tell they've raised yeah close to half a billion dollars right
woo woo woo woo essentially they are they are trying to create a new way to interact have
interact with computers essentially allow creating agents that can control computers um I don't
even know I put this as a copilot I guess but I don't even know if this is like this seems like a
brand new type of concept you know creating agents AI agents that can interact with computers uh
on your behalf it's a pretty science fictiony idea uh so you know could be a huge flop or it could
change the way you know the world works founders have left it droves I feel like lots of
internal turmoil uh I feel like it was the founders the founders earlier other found you know
there was a founder break up at the company um which which is certainly a negative sign but
little scary but but you never know still from from the from the ashes rises the scenics right I mean
the CEO is you know uh david luan uh former I think VP at open AI certainly is there sort of
early days a super smart guy um but yeah can't can't keep the original team together interesting
I just think I think working on AI agents in some capacity that can like tear earlier point about
self-free and self-improving AI and uh you know reinforcement or maybe maxi were talking about that um
I think there's something there that could unlock a lot of value so I'm excited that they're
working in that area I'm sure got it my zoom out I'm most stressed that I should have just
hope for a second and gotten data bricks for cheap because I do think that you got a great deal
got a cheap one there yeah and data bricks is like a pretty sure thing almost more sure than
opening I I feel like the upside um I mean our producer says he's picking the script to be
we're editing the podcast with that and also that was one I was seriously considering
interesting for five working product but I didn't know how much technical under like what
their level of text sophistication was so I didn't see what are what are other ones on the board
left on the board anything you guys are surprised to see I mean Jasper obviously is like everybody is
afraid of yeah uh rapper snorkel has gotten a lot of hype with like internal data cleaning
companies that's what I've heard yeah yeah I they're they're always bouncing around when I'm hearing
writer raised a bunch of money at five hundred nine so I'm interested to see sort of there's
sort of a good price I mean lots of money back lots of money good price for investors it's interesting
humane which makes that little pin that's supposed to replace your iPhone that you wear on your jacket
or whatever that you talk to nobody put that on the board that seems like a tough tough sell
hard hardware is hard I just hardware is very hard yeah yeah they just launched it Paris Fashion Week
which is odd I don't know I mean scale you know was making bank off of throwing humans itself driving
car problems and now they're trying to make bank at a human reinforcement learning for foundation
models mm-hmm but it's just hard to believe you know it's humans like they just don't seem like that
you want you don't need we don't need them anymore we don't need it right I know I want to be like
on all of this little grand thing I mean stability not being in the discussion you know compared to
last six months ago is quite interesting right honestly there's been well first of all there's
been a ton of competition and you know mid-journey has gotten super good and all that stuff and then
secondly yeah obviously there was a story about the founder making a bunch of stuff up which was
interesting I think a journey is or was left off the board here because they haven't taken VC
funding by the way yeah yeah could be a super valuable yeah I know hard to maybe never get
evaluation it's a cash flow business but mid-journey man I would definitely have taken that before
mistral I think the I mean I think weights and biases really good company but I'm just like
what sort of is it really gonna gobble the world runway love the founder like he's
so charming but you know I feel like maybe a do you know the narrative there is like does
Adobe wish they could have bought runway for like way less than it paid for Figma you know when
runway is sort of the relevant company now yeah I mean lots of great companies I wouldn't
write off most of them I don't even know what goes to is I didn't look at that that's the most
high fund rate did anybody research goes student I researched it a bit so my understanding is
there a European company who kind of got started in this in this tutoring space and raised a lot
of money and now is building LMS to do tutoring but I was a little skeptical of like whether that was
a you know current product or not or prototype yeah so that's why I kind of rolled them out my
last point would just be I was really struck in doing the research how unbelievably difficult it
is to even tell what the products of a lot of these companies are or how they're differentiated and
I mean you you could really just like throw buzzwords like large model learning like reinforcements
something something internal data external data you know whatever I just spit out a new
tech article with like a fake company name every day and I would find them very similar to a lot
of the research I did for this so it's yeah I mean I will side you know like I've written about
it's for code I mean partially it's an earlier it's like earlier on the foundation model journey
but yeah there are a lot where it's like it is remarkable how many similar ish like either data
or foundation model or something about enterprise plus LLMS companies that are that are in this
category it's it's amazing to me oh there's my my zoom on my view is that like
max I mean a lot will fall on data bricks of course and that anthropic could be the people
plausibly think anthropic could beat open AI in which case like James got a great
pick discounted open and traffic and traffic is the only one I like from James yeah really
I think I think inflection is also quite strong the rest not so sure yeah care man and where is your
your teams just rock solid my team is immagulate no I would say other than data breaks I have no sure
things but I really took I took call options on hopefully giant upside like I could see my team
having I could see my team having three zeros on it but as long as the two are big I'm I'm going
for the win that way yeah we we really had to do a lightning a speed round of learning Vc's
portfolio strategy here we do not have the date you know any we don't get to talk to anybody where
it is severe this is not real Vc we were playing basically a blindfolds on but it was very
enjoyable conversation piece it's kind of funny that we had to like pick companies that could
return the fund you know they all have to like yeah yeah yeah all right very excited to see in
five years hopefully we'll follow up and like some of us will look like maybe all of us will look
like idiots in this space will be probably wildly overhyped and maybe my you know negative 75
billion will like swamp the whole value and then we can just I don't think that's likely but
as a closer if you just want to read off everyone's team oh yeah yeah yeah for people not
watching the teams are team Eric with a negative 75 billion dollar handicap that I paid to go first
so I could scoop up open AI number one inflection number two character AI character AI number three
clean number four and mistral AI number five max you want to read yours yeah my team is
Databricks pine cone co here modular and imbu also known as the best team and James
I went with hugging face andthropic AI 21 labs replet and adept that's our episode
and that's the cerebral valley AI series watch for youtube videos I'm sure I'll have max and James
once I'm like not fatigued of talking back to sort of reflect on it but I'm not promising it
immediately but this has been a great run thank you for sticking with us and you missed some of
the earlier episodes go back and listen to them thank you so much to Scott Brody who's been the
producer through all this and help help us figure out the show so shout out to him thanks to Riley
Cancela Gabi Caliendo both key and sort of getting the conference and everything together thank
you for everyone who's going to speak and yeah stick around please like comment subscribe
on youtube subscribe to newcomer.co and go play song quiz on your Alexa all right yeah thanks so
much thanks guys thank you all right bye