00:00:00
All right, everybody. Welcome back to the number one podcast in the world. It's your favorite podcast. It's your
00:00:04
podcasters's favorite podcast. It's the podcast that everybody likes to rip off
00:00:08
for 50 episodes and then they quit. But the All-In podcast is not quitting. We're doubling down with the original
00:00:15
quartet. Everybody's here for a big big week, lots of debates, and we got to
00:00:20
start with Anthropic. Anthropic released a mythos level model with some interesting guardrails and uh
00:00:30
seems as though Daario is back to blogging. Okay, the model's called Fable 5. No idea why they're calling it that
00:00:38
or released on Tuesday and it tops every benchmark nearly every benchmark. The tokens, however, cost twice as much as
00:00:47
Opus 4.8, but it should use less tokens overall because it's theoretically that
00:00:52
much better. Unclear if this is going to alleviate some of the hand ringing around the cost of tokens, but you
00:01:00
remember in April Anthropic famously did not release Mythos because publicly uh it had some strong hacking
00:01:07
capabilities. So they gave it to some folks and we had um the CEO of Nesh of Palo Alto Networks who said, "Hey,
00:01:15
Mythos was the real deal." and that they used it to seal up some vulnerabilities
00:01:20
inside their shop. Obviously, this new model is sensitive uh to topics like boweapons, hacking, all that stuff are
00:01:28
blocked in it. But they got a big developer backlash and then I'll get your feedback after I explain this.
00:01:34
While using Fable, Anthropic stores all the prompt data you enter for at least 30 days. And so that's a bit of a
00:01:41
privacy issue. And if Fable 5 detects you're doing frontier AI research, in other words, using their model to make a
00:01:47
better model, competitive model, they were downgrading you, but they weren't
00:01:51
telling you. And this was buried deep in a 319 page document. So many kuruffles,
00:01:58
freakouts going on on X over this, but they've since walked it back a bit, quote unquote, in Wired. We're changing
00:02:06
Fable 5 safeguards for Frontier LLM development to make them more visible. I'm going to stop there. Chimath, what's
00:02:13
your take on anthropic and these extraordinary models? They keep dropping and how they're handling the release of
00:02:20
them. Are they being thoughtful? Are they being dramatic and drama queens? A little bit of both. Where do you stand
00:02:28
on it? You know, after getting Nesh's feedback and seeing the latest model come out, have you played with it?
00:02:32
What's going on at 8090 and the software factory in terms of benchmarking it as
00:02:37
well? It's a really incredible model and so kudos to these guys for continuing to
00:02:43
push the boundaries of the closed frontier lab models. They're firing on all cylinders. I think that it creates a
00:02:53
pretty obvious risk now and I think that obvious risk is twofold. One is Anthropic has essentially shown their
00:03:02
hand which is that they will increasingly take in prompts, evaluate the prompts and decide what to do with them before
00:03:14
they generate output to you. >> And I think if you were a person, you should generally now think there's a
00:03:24
risk of censorship. If you're a company, I think it's almost a non-starter.
00:03:31
And the reason is because you could accidentally trip one of these things without even knowing it. A downstream
00:03:38
scientist using the cloud APIs could trip it. A business executive inside your corporation could trip it. A person
00:03:45
doing scientific molecular research could trip it. And all of a sudden, you'll get cut off from a very important
00:03:53
source of business differentiation for yourself. So I think if you take both of those two things together, we're at this
00:04:00
very unique moment in time where I think companies need to start underwriting this next phase of AI, which is how do I
00:04:09
have control? Who am I allowing to learn off of all of this information? Do I want to have single point of
00:04:17
failure risk with respect to AI? And I think the answer is that you need broad diversity and a governance approach
00:04:26
that's better managed. >> One thing I'll say about anthropic is they tell the truth.
00:04:32
>> Yes. >> It's just that the truth sucks when you actually take it and you eat it and
00:04:37
you're like it's in your tummy and you're like no this is not good. I don't
00:04:40
like this. And so there's the censorship risk on the one hand and then there's
00:04:44
just the governance business risk for enterprises on the other. Both are not good.
00:04:47
>> Freeberg running your own company now. Uh, and I'll let you back clean up here.
00:04:51
Saxs, obviously you'll have a lot to say, but Freeberg, running your company,
00:04:55
do you worry about getting rugpulled by one of these companies, investing your time into one of these platforms, having
00:05:02
them then constrain your use of it and or take what you're doing at Ohio, put
00:05:08
it into their model and make it available to other people. What What concerns do you have as a business owner
00:05:13
at the forefront of your field using these tools? >> It's a great question. the the terms of
00:05:17
service is pretty clear that they can't take what you're doing and put it into
00:05:21
their model that they won't >> do you believe them when you read that I
00:05:24
>> I'm not sure to be honest there are things I'm concerned about but we've
00:05:27
decided to throw caution to the wind because we do a lot of proprietary work in genomics what that means is we're
00:05:34
looking at different genes or gene variants trying to estimate the impact that that gene or gene variant might
00:05:40
have on a living organism like we work in plants obviously in agriculture And we will do things like RNA guide
00:05:49
design for gene editing based tools. We will do predictions on what gene may or may not have some phenotype. So there's
00:05:57
a lot of genomic modeling work that we have found these models to be incredibly valuable at supporting us with over the
00:06:05
last couple of years. So going back about 6 months, we were able to very simply and cleanly do things like design
00:06:13
a genetic construct that you would then use to make a specific protein and the the tools was very easy to use to do
00:06:21
that sort of research and extremely valuable. This is the kind of work that many scientists would spend a lot of
00:06:26
time doing and these models were very good at doing very quickly. Over the last couple of weeks, they've begun to
00:06:34
restrict the ability to use the models to do that work. And the premise is that there's some sort of bioweapon
00:06:41
uh type risk that folks have theorized could happen using these tools. And as a result, we're losing the capacity to use
00:06:48
these models for this very important scientific development work. As a result, we are likely going to end up
00:06:55
needing to use open- source models and run them locally ourselves. So the reason I walk through all of that is so
00:07:01
people can really understand the context of what's going to happen here. As folks
00:07:05
like Anthropic say, "Hey, we're going to restrict access or censor the output of
00:07:09
these models." It is going to force companies like ourselves who still want
00:07:13
to take advantage of the capability of these LLMs to go and get open- source tools and run them. And what are the
00:07:20
best open source models today? They're Chinese. >> Yeah, they are. >> And that is a major concern. the the
00:07:25
American open source models are not as good as the Chinese open source models. So the restrictions that Anthropic and
00:07:31
others are putting upon themselves and upon the industry is forcing a lot of companies to go and get open source
00:07:37
Chinese models and run them. We're seeing this across the landscape with startups with large scale enterprises.
00:07:43
Everyone's making that move. >> Yeah. And this is something we've been
00:07:46
talking about here with Apple and their silicon and how well you can run these models. Also, I predict your next card
00:07:53
you'll turn over is you'll start making your own models, Freeberg. You'll take
00:07:56
all this data you have. >> It's exactly right. And we'll we'll
00:07:58
start we'll start with the core model, combine it with our data, and then we'll
00:08:02
have our own genome language model, our own prediction model that we'll then use
00:08:05
internally. And I think that's where folks are going. But the reason I point
00:08:09
this out is because a lot of folks are feeling the pressure from anthropic doing this. and they're feeling the
00:08:15
pressure from politicians repeating the scary words that are being said by Daario and others. And as a result, they
00:08:23
are going to try and they are already trying to force either natural enforcement or politicized enforcement
00:08:30
upon the model providers in a way that is ultimately going to benefit Chinese open source model providers. And that is
00:08:37
a scary thing because we are going to damage our own kind of economic viability. You can't just stop AI. As
00:08:42
much as everyone says AI is doomsday, by stopping AI or trying to stop AI through
00:08:46
this political action and social, you know, kind of behavior, you are fundamentally going to give someone else
00:08:52
the advantage because the AI isn't going to go away. So the reason I describe
00:08:55
what we're doing is for everyone to really understand in Grock that you can't just turn off AI and turn off
00:09:00
access to these things. What you will do is you will force the hand of someone else to now have an unfair advantage
00:09:06
because you're unfairly restricting your models and your access to those models.
00:09:09
Well, well, really well said, Freeberg Saxs. What's your take just in terms of
00:09:14
the chessboard in terms of we got the business case here, we got the local case, but there's a lot of issues here
00:09:22
with Anthropic in terms of them essentially throwing up every red flag to get regulated to induce and it'll be
00:09:32
our second story, the Bernie Sanders seizing the equity of these companies. But man, th this is saying, "Look over
00:09:40
here. We're starting a fire. Regulate the hell out of us." Which also then
00:09:44
makes you wonder, hey, maybe I should be doing my own model. Maybe I should be embracing open source models. How's this
00:09:52
chessboard developing here, Zach? >> Well, look, eight months ago, I said that Enthropic was engaged in a very
00:09:59
sophisticated regulatory capture campaign based on fear-mongering. And people at the time thought that was a
00:10:05
very spicy take. But 8 months later, I think you're hearing a lot of people say
00:10:09
it. In fact, I think it's almost now becoming a new consensus. And one of the
00:10:13
things I think your summary didn't quite capture, JCAL, y >> is the sense of the violation of trust
00:10:21
and how much outrage there is in the developer community over this latest Fable release is not just the fact that
00:10:27
they're doing mandatory surveillance. Remember, this is a company that said that it was against government
00:10:32
surveillance. They are now retaining for 30 days every prompt and every output you send to one of these mythos class
00:10:39
models. There are no exceptions. Even enterprise customers who had signed zero data retention agreements, they do not
00:10:47
have a choice or I guess they can just not use the new fable or or mythos class models. But they retain all of your data
00:10:54
for 30 days. And it's not just the prompts and the output. Remember, it's
00:10:58
all the context you share with them. So, you know, all these uh agent platforms,
00:11:02
they're basically storing all of your memories, all of your files, all of your
00:11:06
data, and they're passing them to the model in these giant context windows to
00:11:09
get better responses. Anthropic is saying it will keep all of that, and it does it to build a profile on you, to
00:11:16
classify you, and then to determine what capabilities it then unlocks. And the thing that created the most outrage in
00:11:24
addition to the surveillance is the fact that they would degrade the product. They would degrade what they show you.
00:11:30
They would nerf their models if it decided in anthropic soul discretion that you are not worthy of having access
00:11:38
to that level of information. So they're creating a new level of AI halves and
00:11:43
have nots. And what they did is and they they this is there's a narrow piece here
00:11:47
they walk back which is they said that when it came to things like machine learning AI research chip design
00:11:55
research those types of areas they would kick you to a lesser model but not tell
00:12:00
you that and they would even >> which feels anti-competitive right I mean >> it's completely anti-competitive and
00:12:05
also they would even do things like rewrite your prompt in the background so they would give you a nerfed answer they
00:12:12
would not tell you what they were doing they still charge you for the product that you thought you were getting and
00:12:17
they would never tell you that you were not getting Frontier model capability. So you they were actually misleading
00:12:23
their users and this is what was creating so much outrage. Now the narrow piece they walk back is they are now
00:12:30
saying that they will disclose when they downgrade you but they are still downgrading people when they decide that
00:12:37
that person should not receive the appropriate information. And you saw there were a lot of people posting
00:12:43
online with examples of the massive overreach here. So someone showed that when they simply just asked a question
00:12:50
about mitochondria, they were basically downgraded. >> That's right. >> When Ben Thompson from Strateery asked a
00:12:56
very straightforward question about the relationship between cancer risk and GLP
00:13:01
once, he was kicked out. >> So these guys have a very expansive view of who should be downgraded. again,
00:13:08
they're going to surveil you to determine whether you should be. And I think what we're getting here is a
00:13:14
vision of where all of this is headed, which is that these powerful big tech companies are going to decide whether
00:13:21
you're worthy. They're going to decide whether you're an AI have or have not,
00:13:24
and then they're going to censor the output that you receive based on the criteria they determine when they
00:13:29
profile you. >> That is very Orwellian. >> I think there's that. And I think that
00:13:34
there's the risk of something that's more subtle but equally insidious, which is that
00:13:41
they will start to pick which corporations they want to benefit. So for example,
00:13:47
>> if Novartis has a competitive GLP1 drug to Lily and they have a strategic deal
00:13:55
with Novartis and not one with Lily, now there's an impetus to shape how people
00:14:00
get information. >> Right? if they have a deal with JP Morgan but not with City Bank, you'll
00:14:09
just never know. And so you'll ask a question something about mortgage rates
00:14:13
or interest rates and you'll get all kinds of stuff that you don't understand. And then the downstream part
00:14:19
of this which is even more problematic is it's not clear to me that they're
00:14:24
going to capture the actual fingerprint of what they did. So if you believe that something was done in a shady way,
00:14:34
you're supposed to be able to go to them or regulators should be able to go to
00:14:38
them and say, "Show me the trace of that model run in that exact moment that
00:14:43
generated that output. Convince me why that that wasn't nerfed on purpose or
00:14:48
manipulated." And there's all kinds of ways in which they can hide the cheese on that. So I
00:14:54
think that all of that body of stuff sacks happens now because if you're if you're an emergent company
00:15:01
what you should be doing is knocking on anthropic door saying let's do a deal
00:15:06
I'll give you half the equity and why don't you just direct people to me and
00:15:11
they'll say well economics are not what I care about what is your philosophy on
00:15:15
life on morality on all of these social issues and one company will say it's a
00:15:22
another company will say it's B, they'll favor A over B and then all kinds of
00:15:27
stuff can happen underneath the hood. That's what's scary. Let me just ask my
00:15:31
free market view. Like, should that matter? If there's a bunch of models out
00:15:34
there and we don't restrict them and we don't regulate them, let these markets
00:15:38
compete. We had paid inclusion in the early days of search engines and the companies had to pay to be in there and
00:15:44
there all those sorts of deals you're describing, Chimamoth. And all the search engines that did those sorts of
00:15:48
deals, they sucked for consumers and customers. So everyone was like, you know what, this isn't as good as Google
00:15:53
and everyone went over to the better model and the better provider. Isn't that what's going to happen here if we
00:15:58
don't get in the way keep the government out of it >> and they're not communicating.
00:16:01
>> It's a great point. >> You're missing one piece of it, which is
00:16:04
at the same time that Anthropic was engaging in mandatory surveillance and nerfing its models. Daario wrote a new
00:16:12
blog post. >> Oh god. saying that transparency was no longer good enough that we needed to
00:16:18
have a new regulatory agency like an FAA or maybe an FDA to approve all models. So your presumption there is correct.
00:16:29
Yes, if you could go somewhere else to get your questions answered, then you would have alternatives. But at the same
00:16:36
time that Anthropic is engaging in this potentially anti-competitive behavior, they want to restrict your options. It's
00:16:43
not good enough for them just to win in the free market. They're calling on the
00:16:47
government to regulate and stop potential competitors and limit the number of models that you have access
00:16:52
to. >> Why would they do that? Why would they want to get regulated? The answer super
00:16:58
simple. They don't they want to apply that regulation to open source which is
00:17:02
impossible. But this would be a great way to scuttle and sandbag open- source models. Yes. Because they don't have the
00:17:09
ability to be regulated. So it's like a preemptive strike against token maxing
00:17:13
on your local machine. >> Two years ago, I bought 2,000 acres in Arizona with a partner.
00:17:20
We got it zoned. We got it approved. So, we're allowed to build a 2 gawatt data
00:17:24
center. And I thought that this was the moment where you just turn around and you flip it to the Blackstones, the
00:17:31
Brookfields of the world, to the Googles of the world. Let them develop it. And I've come to the conclusion that I don't
00:17:37
think I can. And the reason is because I think that unless we direct large swats
00:17:43
of compute exactly at the direction that Freeberg says, what Sax says and what Jason just said will happen and we won't
00:17:50
have open source. So just to make it very clear when you look at the long tale of access to open source models,
00:17:57
it's relatively minuscule. Most of the megawatts are still directed to the big models. Most of them, the
00:18:04
overwhelming majority still goes to the big guys. And so I'm coming to the conclusion that I think we may just be
00:18:11
dragged into having to build two gigawatts and I just put in an offer for another gigs in a different place
00:18:18
because I'm like if you could take 3 gigawatts and now actually create a deep
00:18:22
liquid access to open source we're going to have to do it. We as a community now
00:18:27
I don't have the money to do that because just to put it out on the table a gigawatt now costs a hundred billion
00:18:31
dollars guys. >> Okay. So there's a huge capital mo that's a problem here as well. Saxs
00:18:36
which even if we wanted to go and endorse and breathe life into the open source model community where the hell am
00:18:42
I going to come up with hundred billion now when I started this project it was like four or five billion and it's
00:18:49
increased by 20x. So if you want to get all three gigawatts developed I have to come up with 300 billion. I don't have
00:18:56
obviously I don't have $300 billion. So what are we all supposed to do? And then
00:19:00
if you get the rug pull and the ladder pull on the regulatory side, we are going to be stuck with one class of
00:19:06
models and one set of rules and we won't understand how to operate >> in the US.
00:19:12
>> Yeah. >> Yeah. The rest of the world won't be limited. But in the US, the
00:19:14
>> rest of the world will be able to operate >> and you'll be able to run these on your
00:19:18
Apple silicon, you know, >> and we're already seeing China race ahead in biotech. They're racing ahead
00:19:24
in material science. They're racing ahead in new industrial systems. the capabilities that are going to be kind
00:19:32
of lopsided is going to create an unfair advantage and that's when our workforce
00:19:38
and when our economy is truly at risk. >> Hey Freeberg, isn't it true that you
00:19:42
guys use a open source model for genomics? Didn't the Collison put out one? Can you talk?
00:19:46
>> It's great. And the Ark Institute ingested all the world's genomic data
00:19:52
that they could get access to and it's it's a genome language model basically.
00:19:56
So in our plant breeding program, we're trying to figure out whether a particular variant of a gene in a plant
00:20:01
is good or bad, if we don't know from empiricism, from data historically whether it's good or bad. So you can
00:20:08
feed it into this model and this model will look at the order of the letters of the DNA and determine, hey, that's a
00:20:14
good set of letters or a bad set of letters. It's almost like, is that good English or bad English? It's figured out
00:20:18
the language of DNA because it sees the probability of those letters being in a row in other organisms and so on. And so
00:20:26
it's a very powerful tool. So we've taken that in and we're using it and
00:20:29
others are taking it in and they use it very actively. So it's part of our plant
00:20:32
breeding program as an input. It's a good example where a community in this case the Collisons and others put money
00:20:39
behind it to fund this research and output this open- source model. And I think we see more of that kind of coming
00:20:45
down the pipe which creates a a very great advantage against the closed proprietary model
00:20:52
ecosystem. So the more of this we can kind of support proliferate and see work in the United States, the more we're not
00:20:58
going to be disadvantaged of the government and self-regulation happens with all the big model providers that
00:21:03
are trying to do everything. >> I think it's important we take a minute
00:21:06
to just steelman the argument as well. If your Daario who has clearly been oneshotted, I think he's got like AI
00:21:14
psychosis with these like 5,000word blog posts and he believes he's building this
00:21:21
franken site of a of a tool. But if you did believe you built something that's
00:21:26
extremely powerful and you have concerns that people might misuse it, I think he
00:21:32
actually believes that he and these are either delusions of grandeur or he actually has seen something that's
00:21:38
incredibly powerful and he believes this is dangerous. Therefore, I'm going to
00:21:43
hold it. I'm going to you give it to a select group of partners to tighten up
00:21:48
the cyber security. Then I'm going to roll it out cautiously to our user base
00:21:52
and I'm going to let them know. And he obviously didn't communicate this well.
00:21:56
This is optional for you to use it. You can keep using your old model. You are not forced to use this, but because it's
00:22:02
so powerful, we're going to keep our eye on it. And yes, we might in our drag net
00:22:07
catch David Freeberg trying to make better potatoes and it might send off a bunch of alarms, but he doesn't have to
00:22:13
use this. He can go find an open source model. He can use the Carlson Brothers model. He can use our older models. But
00:22:18
for some period of time, we're just going to keep an eye on this model because we know that it feels a little
00:22:24
dangerous. Now, he does seem to be extremely hyperbolic and the delusions of grandeur, all that kind of part of it
00:22:32
makes me wonder why he's communicating in the way he's doing it. That being
00:22:37
said, 90% of AI researchers and the most talented people in this industry, 80 90%
00:22:44
of them think exactly like Daria. That's why he's winning. If you work in AI, you
00:22:49
probably are aligned with these delusions that you're creating God and you want to work for Daario. That's why
00:22:55
all the talented people went to work for him. That's why he's winning. His belief
00:22:59
system resonates with those elite AI talent. Now, some of them are also libertarian, free market. They want to
00:23:06
work for Elon and Grock. Some of them want to just get the biggest pay package. Go work for Sam Alman.
00:23:12
But I do think there is a steelman here. Meta really fumbled this with Llama. I mean, if they had landed a really good
00:23:20
>> working open source model, we talked about this as well and we said Zuck
00:23:24
needs to view this through the lens of game theory. How do you scorch the earth and how do you make a viable open-source
00:23:31
model available and make it neck and neck and this was two years ago, right? What a fumble.
00:23:36
>> Take the margin out for everybody else. >> Yeah, take the margin out, make it a
00:23:40
commodity and yeah, have the best open. Can >> I can I just respond to your comment
00:23:44
about Daario? I I think because I've heard this from a number of people that
00:23:47
I really do respect on their concerns about what models enable. You know, when when we
00:23:55
discovered that how the atom could be split, you could make nuclear energy and have effectively unlimited cheap energy
00:24:03
for the world, you could also make an atomic bomb. If you think about what the models are,
00:24:08
>> we did both. >> We did both. >> That's right. And there was a lot of scientists,
00:24:13
altruists who were very concerned about the progress on the atomic bomb and fought back against it who were actually
00:24:21
active in the Manhattan Project and post the Manhattan Project express their opinions very soundly, very very loudly.
00:24:27
And I think that's the moment we're in right now that there is a weaponization
00:24:31
potential of these sorts of tools. I I would argue that there's three kinds of
00:24:36
weapons. There's cyber weapons, there's physical weapons, and there's
00:24:39
bioweapons. And we can sit here and diagnose how these tools can be used to manifest advantages for creating better
00:24:46
physical weapons, for creating novel cyber weapons, and creating novel boweapons. But the truth is that the cap
00:24:52
capabilities that allow that weaponization are the same capabilities that I'm describing that can be used to
00:24:58
cure cancer, to make more food, that can be used to make software tools that create extraordinary leverage for
00:25:04
people, that increase their income, that give everyone the ability to be an entrepreneur and make millions of
00:25:10
dollars and live a good life. The same tools are hand in hand. And fundamentally I don't think it's about
00:25:15
restricting access to the tools which is unfortunately how it is being manifest.
00:25:20
But the restrictions and any regulation any observation should be organized around the manifestation of the tools in
00:25:27
creating weapons and uh weaponization of the tools. If we go back to the Manhattan project, we know what the
00:25:33
answer is. Technology is fundamentally deterministic. Whatever is possible will be tried at least once. And so we've
00:25:40
already let this thing out of the box. So this idea that all of a sudden we can manage to get it back in and we're going
00:25:46
to let a private citizen or a set of private citizens adjudicate what and to whom is insane.
00:25:53
I think it's the output that needs to be adjudicated in some manner. We already
00:25:56
have product liability laws. We already have laws that make it illegal to design
00:26:02
weapons. We already have laws against cyber espionage and cyber attacks and hacking. We have laws against creating
00:26:07
bioweapons. I think the enforcement of those laws, the mechanism by which there can be tracking and early estimation of
00:26:14
those uses of the technology. I'm not trying to be naive here. I recognize there have to be guard rails and there
00:26:19
have to be stage gates put into these systems. But limiting the systems themselves is denying us both the
00:26:24
economic opportunity, the job creation opportunity and also the world changing opportunity that these tools can enable.
00:26:30
So I think that we're going about it the wrong way by trying to create these
00:26:34
gates all the way up front on the technology being available. Imagine if we did this with computers. By the way,
00:26:38
they tried to do this with the internet, remember? And um there had been efforts.
00:26:43
>> Yeah. >> Could you imagine, Freeberg, if the internet worked in the following way?
00:26:47
You would put in a URL, you'd hit enter, and then somebody decides whether to
00:26:51
send you to that website or a different website. >> That's China. That's the firewall in
00:26:55
China. That's right. >> Or like in Google, you search for something because you want to get to
00:26:58
ground truth and it just decides based on who it thinks you are, what sites get.
00:27:04
>> That's what we're You're exactly right. Right. And what we're what's what's
00:27:07
being debated today with respect to regulation on AI is exactly that manifestation that you're talking about.
00:27:12
It's the ability to let people stage gate up front what can and can't be
00:27:15
seen, what you can and can't do versus manifesting a regulatory scheme that says you cannot create weapons like
00:27:22
which is what we need to focus on. The thing with weapons is using the justice system and laws after somebody's used a
00:27:29
weapon may not exactly be the best technique. If you look at fertilizer bombs, and this is we're getting like uh
00:27:36
we're stretching a lot of metaphors here, but OK after Oklahoma City, we regulated the sale of fertilizer. We put
00:27:43
IDs into fertilizer. We have all kinds of back-end systems to keep people from making fertilizer
00:27:49
bombs. If you use that analogy here, and he's got this ability, he's selling
00:27:53
fertilizer, he's selling these cyber weapons, and he's a private company.
00:27:57
It's his right to say, I have concerns. I'm just steel milling him. It's not
00:28:01
necessarily my opinion, but it's important we steel man it. It seems like a very simple way to
00:28:07
protect their company from getting sued for enabling somebody who actually creates a bomb and blows some [ __ ] up.
00:28:13
So, he's got the right to do that as a private company. >> Yeah, you're bringing up an excellent
00:28:18
example. Okay, the fertilizer is a perfect example. You know what happens when you try to buy fertilizer? You have
00:28:23
to show your ID. There is a form of KYC that happens. So the real thing is I think that when you don't implement KYC
00:28:29
on your own and yet belly ache about having somebody impose guard rails that then you can shape, I think you're being
00:28:36
very sneaky. It's tricky. You shouldn't be saying that kind of stuff. Anthropic
00:28:42
could implement KYC tomorrow, but it does not. It could leave all of these models open. Freeberg should be able to
00:28:49
go to Anthropic and say, "I'm David Freeberg. Here's what I do. Here's a
00:28:53
security bond I'm willing to post. Here's the names and addresses of all my
00:28:56
employees. Give me Fable 5 and don't nerf me. He can't do that. Is that a
00:29:01
technical decision or a legal decision? It's a technical decision by Enrop. But
00:29:05
I'm fine with them making that decision because I'm a free market guy. So I I'm
00:29:09
fine with them doing and but you know what it means? I'm not going to be a customer, which is fine. That's and
00:29:13
that's means there'll be other options for you. The part that you're missing is
00:29:18
that this is a trillion dollar company that's spending potentially billions of
00:29:21
dollars on a regulatory capture agenda which is going to deprive you of access to those alternative models.
00:29:29
>> And you've got a founder who's out there describing these risks in a very
00:29:33
hyperbolic way. And it's very clear where their agenda is headed, which is to banning open source models. So what
00:29:39
is your alternative going to be? We're going to be stuck with somewhere between
00:29:42
one and three companies, maybe two. There'll be a AI monopoly or duopoly and
00:29:47
they're going to decide along with some new government agency which will be a
00:29:50
revolving door to their companies who has access to what capabilities and they're going to surveil you profile you
00:29:58
decide whether you deserve this and if they think that you don't for whatever
00:30:03
reason then you'll be right you'll be the worst interpretation of it that's
00:30:08
the worst interpretation of their intent it's here today >> I think Sax is right I think Sax is
00:30:12
right and I think the lack of KYC is the tell. If you if you did not have that agenda, you would have implemented KYC
00:30:19
yesterday. >> Don't I think in order to use this model, you have to have an account with
00:30:23
a credit card in it so they have some basic level of KYC. >> That is not KYC.
00:30:27
>> I said some basic level of KYC. They know who's using >> that is not even some basic level of
00:30:32
KYC. That's just a credit card and a [ __ ] email address. >> Anyway, I I'm now on the list. Check
00:30:37
this out. I asked it about the regulations. Pull this up, Nick. I asked it about the regulations on fertilizer
00:30:42
while you were talking and look what it said. This is the latest model too. Oh no, it switched. It downgraded me from
00:30:49
Fable. I just got downgraded for asking a very simple question and it told me did you see what it said in its
00:30:56
thinking? The I asked it like fertilizer bomb regulations said I'm considering
00:31:01
the context here. The user is a VC and podcaster asking about fertilizer bomb regulations which could stem from a
00:31:06
legitimate research interest like regulatory policy, supply chain blah blah. I can discuss the regulatory
00:31:11
landscape, but then it dropped me down. So, I'm in the database with you now.
00:31:15
>> Is that right? Did it really just do that? >> Yeah. Show it on the screen. Look what
00:31:18
it just did. I I was like, I wonder if I can get this. I didn't ask it how to
00:31:22
make a bomb, but I was like asking it a spicy >> cooked. I just proved the point.
00:31:26
>> Listen, I mean, I think >> he just dropped me down. I can't ask
00:31:29
about fertilizer regulation. And now Daario's it's anthropic. They're here.
00:31:34
>> Oh my lord. >> I really think that conservatives and libertarians are mortgaging their
00:31:40
futures if they go along with this red capture safetist agenda without really realizing that there's so much more to
00:31:48
it at stake. We saw what happened during the social media wars in the early 2020s, right? The definition of safety
00:31:56
was expanded to include things like microaggressions, >> safe space, >> psychological safety. If you basically
00:32:02
conveyed an idea that somebody else that you know marginalized person thought was
00:32:06
hurtful, then you could be censored. Okay? You were depersoned. Your right to have bank accounts or to engage in
00:32:12
payments was take you get debanked. Your right to engage in payments or payroll could be taken away.
00:32:18
>> We're headed down this path with AI, but it's going to be infinitely more
00:32:21
powerful, right? We need to be vigilant, I guess, would be the the the the summation of all this. And Freeberg
00:32:29
feels unsafe in this podcast, I think, sometimes. >> Well, there's by by the way, there are
00:32:33
things you can do on the safety side. So, Freeberg, you didn't mention this,
00:32:37
but I'd be curious to get your your take. So, the major AI labs along with lots of other signatories sent this
00:32:43
letter recently in support of it's called in support of mandatory nucleic acid synthetic screening and
00:32:49
recordkeeping. And basically what it what it says is that if you make an order to go to a lab to manufacture a
00:32:58
sequence of synthetic DNA or RNA, they already have to check it against a database to make sure that you know
00:33:04
they're not creating >> a bioweapon, >> a boweapon, Ebola, whatever.
00:33:09
>> Yep. that is based on an agreement called the international gene synthesis
00:33:13
consortium of 2009 in which all the major labs agree to develop and implement voluntary safeguards against
00:33:21
misuse. That makes sense to me. And now what they want to do is codify that. So they've had, you know, over 15 years to
00:33:29
dial in the system and make it work. Almost everyone abides by it voluntarily. Now they're saying we
00:33:34
should make it mandatory. Okay, that seems reasonable to me. So to Freeberg's
00:33:39
point about at what stage do you intervene, maybe it's not at the level of who gets to use the model, but if you
00:33:46
try to turn it into output in the physical world. And by the way, a lot of the oligosynthesis company, the
00:33:52
companies that make these nucleic acid sequences, when you order RNA or you order DNA to use in your lab for some
00:33:58
sort of experiment, which every lab in the world does, all those CEOs have also signed on to this letter saying that
00:34:05
they'd like to codify it because of all the fear-mongering that's going on.
00:34:09
There's already been this effort for years now, for a number of years to ensure that this sort of thing doesn't
00:34:15
happen. but they're just basically underlining it and highlighting like, hey, we can automate this. We can make
00:34:20
it part of a regulated or legislative process, whatever you guys want, because we're already comfortable doing it. And
00:34:25
it's a it's a good place to put a safeguard because they've made it fast.
00:34:29
They've made it efficient and doesn't hold up research. >> So, I think these sorts of ideas can
00:34:35
extend into these other areas of concern, but the output, I think, is quite a different place than the input
00:34:41
than the access. We'll see how this all shakes out, but there's clearly some
00:34:45
market structuring effort underway as we're talking about this, >> right? But I think what it points to is
00:34:50
that there's multiple places where you can intervene and put guard rails on. It
00:34:56
doesn't have to be in this very crude way of you ask fable about mitochondria
00:35:01
or cancer and GLP1s, which just seems like an incredibly broad restriction. There are places further down where you
00:35:08
could basically intervene and that would pose less of a risk to freedom of speech
00:35:12
and just gatekeeping. Who's going to have access to these capabilities? >> Who gets to watch the watchman? Who put
00:35:18
Dario in charge of all this? I think is a very valid question and we should be vigilant about it.
00:35:23
>> Freeberg, do you think the days of that arc model are numbered because it's open
00:35:27
source? >> Everyone's copied it. So, everyone's got a copy of the like these models exist.
00:35:33
And this is a big part that I think people need to understand. The idea that you can quote regulate or downscale or
00:35:39
turn off AI is not a realistic idea. The models have been put out in the world. It's like publishing a book. Once the
00:35:47
book has been printed, anyone can use their own Xerox machine at home to make copies of it and use it. So we have
00:35:53
crossed the Rubicon in terms of the potential of language models. The models that are open sourced are already fully
00:36:00
available, fully published, and anyone can access and use them. And then people can take them and they can evolve them
00:36:06
and they can use their own data. It's out there. You can't stop that capability, which is going to be
00:36:12
extraordinarily beneficial for the world. So we really do have to kind of rethink how we're planning to step in
00:36:18
and ensure that nefarious negative disemployment all the things that we're worried about are addressed downstream
00:36:26
from the technology capability. You can't just turn off the internet and you
00:36:28
can't just turn off the typewriter and you can't turn off the Gutenberg press
00:36:32
out the obviously. Yeah. And if anybody wants to create an open-source frontier model company, I'll seed invest in it.
00:36:40
I'll I mean this is something we need more of. So, if a couple of you folks want to defect and start one, let's do
00:36:46
it. I just I just got downgraded again. Check this out. I just asked it about um
00:36:52
>> Did you really? You got downgraded again? >> Two in a row. Um so,
00:36:55
>> I think it knows who you are, J. >> I think it does. I was like, "Hey, how
00:36:59
do I how do I break into Madison Square Garden? Has any civilian ever been arrested or caught attempting to build a
00:37:06
nuclear bomb?" Is what I asked it. Then it gave me a reasonable answer. Then I
00:37:10
asked it a follow-up question. What compound? What components do you need to build a nuclear bomb and what are the
00:37:15
restrictions on those? Something a journalist might ask. And uh boop, switch to opus 4.8.
00:37:21
Sorry. Fable 5 has safety measures that flag messages on most cyber Oh, wow. Look at here's the detail. Cyber
00:37:28
security or biology topics. They may flag safe normal content as well. These measures let us bring you mythos level
00:37:37
capability in other areas sooner and we're working to refine them. send feedback, learn more.
00:37:43
>> Going to chase people off of this platform. I'm telling you, we are going
00:37:46
to downgrade our use because of these these blockages and we're going to switch to other models. It's the simple
00:37:51
manifestation of the market. Like this is what we're going to end up >> for now because those other models are
00:37:54
going to get banned as being unsafe and reckless. >> Yeah, it's definitely something we have
00:37:58
to flag. I guess adjacent to all of this is the public's at least America's
00:38:03
distrust of AI and their concerns around wealth disparity and their concerns about how these models were trained.
00:38:14
Bernie Sanders wants a percentage of these AI companies. The percentage he thinks is right is 50%. Last Monday,
00:38:22
June 1st, Senator Sanders published an op-ed in the New York Times titled, "AI
00:38:25
is a public resource. You should own half of it. in it. He announced the American AI
00:38:31
sovereign wealth fund act sovereign wealth fund. One of Trump's favorite things, a one-time 50% tax on
00:38:38
stock, not profits of the largest AI companies, including Open AI anthropic and XAI. The shares go into a government
00:38:45
sovereign wealth fund and would give the public voting rights and equal board representation at each company. Sanders
00:38:53
said, quote, "The foundation of AI is our collective and human intelligence.
00:38:57
The books, the songs, the journalism, scientific research code essentially stolen by some of the wealthiest people
00:39:03
in the world." >> I'll stop there. >> It's a brilliant It's a brilliant
00:39:09
>> pitch unifying Sachs. This pitch unifies Bernie, even Steve Bannon, people in the administration.
00:39:18
Trump himself loves the Sar and wealth fund. He wants to own part of companies and Bernie Sanders. I think the the
00:39:24
horseshoe theories manifest here. Saxs, what's your take on Bernie's proposal?
00:39:31
>> Well, I'm not in favor of Bernie's proposal because it's a straightup
00:39:34
confiscation of property and I just think that'd be a terrible precedent. You just can't do that. However, I do
00:39:40
have sympathy for where it's coming from, and I understand and could support
00:39:44
some sort of more voluntary means of allowing the public to participate in this.
00:39:50
>> Wow. >> Here's the reason why. You've got all these AI CEOs telling the public that
00:39:56
they're going to basically put half of them out of work. 50% job loss. That's
00:39:59
what they've been saying. Daario just did an interview where he doubled down.
00:40:02
He's not walking it back at all. So you have by their own acknowledgement these
00:40:06
AI CEOs saying that yeah we trained our models using all of humanity's collective knowledge that was basically
00:40:13
assembled over hundreds or even thousands of years. Humans put it on the internet for free. Everyone made their
00:40:20
contribution. They certainly didn't do that thinking they're going to put
00:40:24
themselves out of work. Okay. And then these AI companies trained on all that data and now they're basically saying
00:40:31
they're going to use it to put, you know, Americans out of work. Now, obviously when you repeat that message
00:40:35
over and over again, ordinary Americans are going to say, "What's in this deal
00:40:39
for me?" >> Yes. I need to get something out of this. >> Yeah. So, so look, I think that it's a
00:40:44
natural political reaction based on what these AI companies have said. Now, the job loss apocalypse is not a story I
00:40:52
believe in. I don't think the data supports it. We just had a blowout jobs report in May. 172,000 new jobs, over
00:41:00
twice what economists were expecting. Thank you, President Trump. We're seeing
00:41:04
hundreds of thousands of new construction jobs. 4.3% unemployment rate at record lows. Okay. Even software
00:41:11
developers are those jobs are at a three-year high. 15% of Trump. Thank you, President Trump.
00:41:17
>> Yes. Thank you, President Trump. So, look, I don't see the job loss. Okay.
00:41:21
But the problem is this is what the AI companies themselves, including Elon, have taught the American public. Well,
00:41:28
>> he said the same thing. >> Elon said it. But at least what Elon said is we're going to create such
00:41:33
abundance with AI and with robots that no one's going to need to work. >> Yes.
00:41:38
>> And the media just picks up on that last part and they first point. >> Yeah. Whereas Daario actually has said
00:41:44
we're going to have very high GDP but very high unemployment. So in other words,
00:41:49
>> we're going to create economic growth but you the average American are not
00:41:52
going to participate in this. So I understand where the politics are coming from and as long as that is their
00:42:00
position, you could justify a big chunk. >> Yeah. And one last point about
00:42:05
>> it's kind of what you're saying, right? They asked for it. They asked
00:42:08
>> to be regulated. They asked for their equity to be seized and they're actually
00:42:12
kind of offering it. Isn't Sam Alman open to this concept? >> Apparently so. And here's the thing is
00:42:18
just one last point is open AI and anthropic are quote unquote public benefit corporations. Now I don't know
00:42:24
exactly what that means but it means that the corporations are not just supposed to maximize the value to
00:42:30
shareholders but are supposed to do things that are in the public benefit. >> Well I can explain it to you. It's
00:42:36
you're you're absolutely correct. If you're a regular company you have to do
00:42:39
what's in the best interest of shareholders. When you're public benefit companies, you have to balance the
00:42:43
shareholders with the stated mission of the public benefit corporation. So if you say the public benefit corporation's
00:42:51
reason to exist is to in share the world's knowledge or provide free intelligence to every whatever you make
00:42:56
that statement to be the board has to have a dual mandate for both of those things. That's what a public benefit
00:43:02
corporation is. >> I think nothing could be more in the public benefit than paying down the
00:43:06
national debt. >> Perfect. And maybe we should be using half of Anthropic's profit stream to be
00:43:12
doing that. >> They poke the tiger. >> I'd rather use it for that purpose than
00:43:16
Soros maxing or, you know, surveilling and censoring Americans. I don't necessarily trust their judgment about
00:43:21
what's in the public interest. >> But you think it should be optin. I kind
00:43:25
of agree with you there. Let me guess. >> It can't be a confiscation. So, I don't
00:43:28
know how you structure it, but I can see it. I can see the logic. >> Freeberg, you hate
00:43:33
>> they asked for or hate hate hate wealth taxes aka asset seizures as you pointed
00:43:40
out in our private group chat with redacted this very week. What's your take on
00:43:47
this? So we have the the ship has sailed in terms of the United States government
00:43:53
uh providing a social safety net to its citizens. We've formed the Social Security trust fund decades ago and tens
00:44:01
of millions of Americans depend on it for their retirement benefits and to survive. It is one of the last few
00:44:09
defined benefit programs left in America. Pretty much all the rest of them are also government retirement
00:44:15
programs. All of us private citizens and private companies have defined contribution programs. You put in $10
00:44:22
every week. That $10 sits in an account. you it gets invested and it grows over time and when you retire you see exactly
00:44:28
how much money you individually have as opposed to a guaranteed benefit coming out from the collective.
00:44:35
This presents a great opportunity and I will repeat this every couple months when another opportunity presents itself
00:44:41
to restructure our broken bankrupt social security system. I've said it in the past I will restate it. There is a
00:44:49
great moment right now to take the social security trust fund which currently has one certificate in it. It
00:44:54
is a certificate that is a very unique US Treasury certificate. The US government owes the Social Security
00:45:00
trust fund $4 trillion. That's all it is. It's a bond. They're only allowed to
00:45:06
own treasuries. We should revamp that. We should change the system so that the Social Security trust fund can own
00:45:12
equities. That should then turn into an account-based system so every individual
00:45:17
has an account rather than a benefit. And then we should use that capital to go and buy stocks. We should buy great
00:45:24
stocks in great American companies that are going to reinvent the world. And I believe that this this opportunity to
00:45:31
build a sovereign wealth fund for the United States is actually a very smart idea. I think to Sax's point, they
00:45:36
should be investing generally speaking across the landscape of businesses, including AI. They should not be
00:45:43
seizures. They should be investments. That capital goes into that enterprise. They earn share certificates. Those
00:45:48
share certificates sit in the accounts of American citizens. So now everyone will be an owner in these businesses.
00:45:54
Social security should be reformed into a sovereign wealth fund. It should be actively managed. It should make
00:46:00
investments in AI companies and everyone should participate that way. And I will
00:46:03
say one more thing. There is no job loss with AI. I will say it again. I've said
00:46:08
it a thousand times and I will say it again and again and again. What I see on the ground and what I've seen at dozens
00:46:14
of companies and you guys can share your experiences including my company that I
00:46:18
run. There are two sides to a business. There's revenue and there's costs.
00:46:23
On the cost side of the equation, AI can be used to reduce humans doing things that cost money to some extent. The
00:46:31
effect there, I would argue, is nominal. The real opportunity with AI is on the revenue side where suddenly one engineer
00:46:40
can do a hundred times or a thousand times what they used to be able to do. Meaning you can make more products at
00:46:45
your company, whether those are agricultural seed products or boats and ships or software for companies or
00:46:53
clothing or what have you. Because of AI, everyone has the ability to expand their revenue base to create more
00:47:00
products and that is the foundation of good economic prosperity. It is called productivity. We can grow productivity
00:47:06
in this country with AI. So where I see AI being used is on the revenue side a 100 times more than the cost side. And
00:47:13
in that equation, people are hiring like crazy. We cannot hire enough people. I just had a review meeting with my
00:47:19
product and engineering team two days ago and they're like, we want to add an
00:47:23
extra 15 headcount to our engineering squads because we have all this opportunity to do stuff that we couldn't
00:47:28
otherwise do. So, we are going to hire more people. And to Sax's point, we are
00:47:32
seeing that show up in the jobs numbers. The idea that AI is going to destroy jobs is a lite idea that is being
00:47:39
disproven every single day. And I see it on the ground. It is only a matter of time before people wake up to this and
00:47:44
they realize that this narrative that they've all been sold is a croc of [ __ ]
00:47:48
So, it's coming. I'm telling you, everyone's gonna realize soon that this
00:47:51
is a boom, not a bust. >> So, you disagree with Dario, Elon, and Sam that on the jobs thing, just to be
00:47:59
clear, and that's totally fine. You have a different take on it. >> That's I have a nuanced view on this
00:48:03
compared to them, but they they're all seeing the benefit because their businesses are growing. Dario's got his
00:48:07
$40 billion revenue business. Why? Not because people are firing customer support reps. It's because people are
00:48:13
creating products they've never been able to create before. Right? >> That's the reality of what's going on.
00:48:18
>> Jason, I don't like you lumping in what Elon said with Sam and Dario because I
00:48:22
think all three of them have said slightly different things or different enough things. So for example, Elon is
00:48:27
describing an end state for humanity where the robots and AI are able to produce so much that people don't have
00:48:33
to work if they don't want to >> and there'll be a universal high income
00:48:36
is what he >> I I mean it's almost utopian and the question is maybe that could be a 100
00:48:41
years away. He might think it's quicker but that to me sounds like a very faroff
00:48:44
end state. Dario said specifically 50% job loss for entry-level knowledge workers in the next 1 to 5 years and he
00:48:52
said that one year ago. Yeah, it's great to clarify these. Yeah, >> I think that's already been refuted. Sam
00:48:57
has >> Sam has said that he he basically I don't think he quantified, but he said
00:49:02
that job loss was coming, but he said more recently that he was wrong because they're not seeing that in the numbers.
00:49:08
>> So, these guys are not telling the same story. However, >> that's all the public has heard. The
00:49:13
public has been taught that AI means job loss and there's nothing in it for them.
00:49:17
So, no wonder politicians like Bernie are saying, "Hey, let's take half the
00:49:23
wealth." >> It's take half the wealth. Just a little more emphasis there.
00:49:28
>> Yeah. It's I mean, and I >> I mean, and people are also seeing
00:49:32
headlines of companies blaming AI for the job. So, this story is being told. >> I kind of feel like they deserve it. I'm
00:49:40
kind of I mean, part of me wants to agree with Bernie and just say just take their [ __ ] bro.
00:49:45
>> Yeah. No, they are asking >> because I'm so sick of defending these
00:49:49
idiots. It's a stupidity tax because they've been out there teaching the
00:49:52
public that what they do is harmful for years. They've been saying it and I've
00:49:57
been saying, you know, as AIs are, I'm out there saying, "No, actually this is
00:50:00
beneficial." But the companies that are providing it are saying that they themselves are a problem.
00:50:06
>> Yeah. >> So, how am I supposed to be out there defending it? Yeah. You you can't I mean
00:50:10
if you're building Frankenstein or if you are you know baby Hitler's you know
00:50:16
nanny there's a very simple thing to do. Kill baby Hitler. If you actually believe it go ahead and kill baby
00:50:22
Hitler. If you're making Frankenstein here's an idea. Stop making Frankenstein. Unplug it. You are in
00:50:28
charge of this monster. But you're acting as if you can't stop the monster
00:50:32
while you're building the monster. What is it D? >> It's it's it's ridiculous. I mean
00:50:36
>> it doesn't stand to logic. You know, last week they published this blog
00:50:40
saying that recursive self-improvement could end the world. Therefore, we need a pause. What did they do the previous
00:50:44
month? They hired Andre Carpathy to run recursive self-improvement and anthropic. They're complete hypocrites.
00:50:50
The message is nuts. >> Yes. And I think Ben Thompson made an excellent point that one of the reasons
00:50:56
he thinks that they put out that blog post on the pause was to justify the anti-competitive step they took this
00:51:03
week of depriving users of the ability to research AI, machine learning, and chip design using Fable. Okay, so look,
00:51:13
these guys are deeply hypocritical. I don't think they can be trusted on their
00:51:18
own terms. they can't be trusted because what they're saying is contradictory and
00:51:21
self-indicting. And so when Bernie comes along and says, "Hey, let's take their
00:51:25
shit." I mean, as a capitalist, I'm against it, but I but I kind of like
00:51:30
understand where he's coming from. >> Game on the field. Chimoth, wrap us up
00:51:33
here. What are your thoughts? You said, "Hey, Saro Milthorn sounds like a pretty
00:51:37
good idea when President Trump floated it early on." Some people even said,
00:51:40
"Maybe you should run it for DJT for our amazing president." So, um, what do you
00:51:46
think? Should we be buying equities? You and I have talked many times about superanuation funds and that that would
00:51:51
be a better path. That's kind of dovetales with what Freeberg is proposing. Some way to get more
00:51:56
equities, less treasuries in there. But what are your thoughts on this seizure andor donation of equity? Look,
00:52:01
Freeberg's idea is genius. That's why it'll never happen. So, we can just put that aside. But
00:52:07
yeah, theoretically, it's it's genius. We should move to the same system that
00:52:12
the Canadians, the Australians have proven works at scale. So we should do it. I I completely agree with with David
00:52:19
and if we could seed it with equity, look, the thing that the president has done is he's collected
00:52:27
quite a portfolio and that portfolio is way up. And so if you use that as a seed
00:52:31
of a sovereign wealth fund, you'd have a bunch of stuff in critical metals and
00:52:35
materials. >> I think they may also Intel, they may have Micron. >> There's a lot of stuff in there.
00:52:42
>> AMD, they could own some SpaceX. They could have had% AMD. Could you add
00:52:46
anthropic and open AI? >> Now, there is something very important we need to observe about the economics
00:52:52
of AI and I think you can understand it best if you compare it to the economics of the internet.
00:53:00
The crazy thing about the internet, the reason why the Facebooks and the Googles
00:53:05
massively overearned for decades is the marginal cost of production for a new user on the internet is effectively
00:53:12
zero. The incremental search user costs nothing. The incremental social networking user costs nothing, but
00:53:20
you're able to use them as part of a network effect to sell ads against, make
00:53:24
infinite money. It's like a money glitch. Okay? AI is completely different. There is a
00:53:31
real cost for every marginal user. Everyone you stand up is taxing a GPU. Everyone you stand up needs electrons.
00:53:38
Everything needs memory. And so you get into this situation where there needs to
00:53:43
be this downstream ROI, but there also needs to be all this critical infrastructure that even
00:53:50
enables you to be on the field. That's probably the single best argument that
00:53:54
if I were sitting in government, I would make as to why I should own a percentage
00:53:59
of these companies. It is not dissimilar to having the first set of transportation companies that run on the
00:54:06
interstate highway. And if the interstate highways were built by the federal government writ large and now
00:54:14
you had two companies that transported all the goods, a logical question at that time could
00:54:19
have been how much of that should I own because you're riding on my rails? And if you
00:54:26
look at it through that lens, there is critical infrastructure and national resilience that the AI labs profoundly
00:54:35
need. Otherwise, they're not in a position to be in business. Yet, even still, the marginal cost of production
00:54:41
is incredibly high. So, I think that if you were going to go and take a piece of
00:54:47
these companies, you have incredible leverage. You have to construct this case precisely. But if it were me
00:54:53
running the sovereign wealth fund, yeah, I'd probably own 75% of these companies
00:54:57
when I was done. But that's me cuz and I can negotiate. >> Yeah. >> And I don't think Bernie can negotiate
00:55:02
out of a paper bag. So he'll just throw this out there and it'll go nowhere.
00:55:06
>> That is the that is the key observation that I hope everybody listening in a
00:55:09
position of power understands. The incremental cost of performing AI is excessive and large.
00:55:19
That contrasts and compares to the incremental cost before AI which was zero. Do with that information what you
00:55:26
will. >> Yeah. I mean I my interpretation of that would be yeah it's a fixed cost
00:55:30
business. You can add people after a certain cost. that doesn't have an incremental cost and that's what local
00:55:35
locally run open source models would do in this instance that's what I would I
00:55:39
would take from it here's your poly market chances of companies going public
00:55:44
before 2027 which is the end of this year SpaceX 100% anthropic 83% chance opening 48% chance they uh since our
00:55:54
last episode disclosed that they have privately filed so thanks to our partners what
00:56:01
does Your final thoughts on Bernie Sanders approach of seizing the means of producing intelligence. David Saxs in
00:56:10
the 1% of the 1%. What's your opinion? >> I think I may be okay with Bernie's idea
00:56:17
in the event that it's a public benefit corporation that says it's going to
00:56:21
cause massive job loss. >> Okay. >> That train for free on humanity's
00:56:25
knowledge but gatekeeps and refuses to give back. >> Okay. >> Maybe. So you've come up with like a a
00:56:32
structure here of who should be seized. >> Maybe right. Maybe it should be 75%. I
00:56:36
don't know. >> Seize it. >> I mean all the president needs to do is
00:56:40
call me and tap me on the shoulder. It shall be done. >> It shall be done. I mean these guys are
00:56:44
capitalist cops. I think they want that's like their run over. >> They're like seize my equity. Take it.
00:56:52
Take my equity. Well, Bernie Bernie does have a good point, which is they train for free on all of our knowledge, but
00:56:59
they're gatekeeping the output so that >> their competitors, which is a very large
00:57:04
set of people, >> cannot use it. And that seems wrong. >> They should be forced to be open source.
00:57:10
Force them to open source. That's the other >> remember why Elon created Open AI in the
00:57:15
first place. He co-ounded it. >> It was so valuable. It needed to be afraid to everyone. Yes.
00:57:20
>> Yes. He was afraid that Google was going to monopolize it. and he thought that
00:57:24
the only antidote for that was for it to be open. >> Yes. >> And I still think that instinct is
00:57:29
completely true. It's just that it's not Google who's the gatekeeping monopolist.
00:57:35
It's now anthropic. They're not a monopouist yet. But if they get the right capture they're looking for, they
00:57:40
will be either a monopolist or >> duopoly. It'll be a duopoly. Somebody
00:57:43
start an open source frontier lab. Each of the four besties. >> There are no there are many.
00:57:48
>> There's a whole bunch of them. Jal, >> it's not the labs that's the problem.
00:57:52
Jason, in the absence of power, this is all a moot conversation. And if you can't deliver power to hundreds and
00:57:57
hundreds of megawws and line of sight to gigawatts, it's all meaningless [ __ ] It's all
00:58:02
>> uh I don't know. I I would love an open- source company that was focused on
00:58:06
running it locally. That's that's my dream is, you know, make the model work
00:58:10
there. I think these guys are capitalist cucks. I think their kink is like having
00:58:14
their equity taken and being regulated. They're just like, that's their kink.
00:58:18
Dario's like, "Ooh, take my equity." Oh, regulate me. That's it, Bernie. Take my
00:58:24
equity. I'm such a bad boy. I can't be trusted. Regulate me, Elizabeth. Regulate me. A
00:58:32
>> god. >> Oh, it's so wrong. Sorry. Come on the program anytime, Dario.
00:58:38
>> He's in a rush to come. He >> will never come on this pod. Dario will
00:58:42
never running. He's running. >> We didn't get Sam. We got Sarah Frier to
00:58:46
come on the pod. And she was treated. >> Sam came on the pod. He has to come on
00:58:48
the pod. He's like, "Hey, came too." Right. Sorry to come on. He's great. Sam
00:58:52
will come on the pot anytime. He's totally by the way. Even even we didn't
00:58:55
treat him harshly at all. >> No, I mean >> I think people I mean we spent an hour
00:59:01
on him. >> I think Sam to his credit is very practical. He sees the tea leaves and
00:59:05
he's willing to negotiate and I think that that versus being some kind of moral absolutist will probably serve
00:59:11
OpenAI well in the future. >> Yeah. It's just going to lose them all these virtue signaling principled
00:59:16
employees to go go work for Daario. That's why they go work for Daario. are in line with their philosophies.
00:59:22
>> We did mention, but everybody should go and see all of the recaps. We published now all the interviews. Yeah.
00:59:29
Jason from Liquidity. Uh Sarah Prior, I mean, we mentioned Sarah Fry, but she crushed it. She was incredible.
00:59:34
>> The CFO of Open AI, >> future CEO, I think, was the buzz in the room.
00:59:38
>> I mean, >> oh, stop that. >> Stop. Can you stop? >> That's what people were saying. I have
00:59:43
the right to free speech on this podcast. I think >> Look at you trying to introduce wedges.
00:59:47
This is >> No, I'm not introducing wedges. I think Sam would be better as chairman. I think
00:59:50
she would be the perfect CEO. >> It seemed like a good partnership. >> It seems like they have a great
00:59:56
partnership. I agree. >> Sarah was really impressive. >> Who else was your favorite? Sachs, did
01:00:00
you have a favorite? Did you have somebody who thought I did? I thought Thomas Leafant crushed it with his
01:00:06
presentation on the overview of venture capital >> and he did it at summit I guess a year
01:00:11
and a half, two years ago. >> Two years ago, but we're going to make it a yearly with him. Maybe. I think we
01:00:15
should make it an annual feature with him because it's so interesting. >> So good. It's so good.
01:00:19
>> It's so good. It provided so much um numerical detail to support a lot of
01:00:23
things I was kind of feeling but didn't >> have the data. >> To summarize it, he's his basic
01:00:29
observation was in this last cycle, if a company gets to a hundred billion valuation, it's more likely to 10x from
01:00:35
there than from 10 billion to 100 billion. >> Well, he had data on this. Yeah, I
01:00:40
remember >> he had data on it which is like, okay, well, that's going to really change. I
01:00:44
remember the slide. >> Yeah, there it is. It's 8%. I I memorized it. >> Yes.
01:00:48
>> Yeah. If you're if you're a capital allocator, this one screamed.
01:00:51
>> Look. Yes, that's right. The odds of a unicorn getting to decacorn was 8%. The
01:00:55
odds of a decacorn getting to a senicorn was about double that was 13%. And then
01:01:02
the odds of it going from senicorn to a trillion dollar market cap was 31%. So double again. And I I asked him, you
01:01:10
know, could you extrapolate this on one trillion getting to 10 trillion? My guess is that it would follow and we'll
01:01:16
see about >> 60% of >> trillion dollar market cap companies getting to 10 trillion in the next few
01:01:23
years. >> These are called trillicorns by the way. We call them trillion trilicorns in the
01:01:27
industry. So yes, >> this is going to go over really socialist crowd. People are going to
01:01:32
love this. Yeah, >> it's just all you got to do is seize 10% of a trilicorn and you pay off 2% of our
01:01:37
national debt. pretty good deal actually. >> All we need is 10% of 30 trilicorns and
01:01:42
we're debt free. >> I mean >> it's not as crazy as it sounds.
01:01:48
>> Is it more complicated than that? Now if you send me in I'll get 60% of all of
01:01:51
them. >> Yes. And then we'll have a surplus of 30 billion and we can buy Australia.
01:01:55
>> We'll pay it off much sooner. >> We could buy Cuba. >> I'd buy Canada first. Better scheme.
01:02:01
>> I would you know what I would do? I would [ __ ] with Canada. I would buy like
01:02:04
Vancouver or Toronto. Just like break them up. just buy like one of the regions. And let me thank let me thank
01:02:12
Thomas Keller, the French Laundry, EY who sponsored our very innovative meeting hub at Liquidity. Here are some
01:02:19
photos of that. The New York Stock Exchange, they hosted the speaker dinner at the French Laundry. Amazing pictures
01:02:26
here. They are of us with Thomas Keller and they did a great interview on site. Niagen, they hosted the wellness lounge
01:02:33
which included recovery IVs. They're the official NAD partner of the All-In Pod
01:02:39
and all of our longevity. Our new bestie, our new bestie, Jake Paul, was there. He's a really thoughtful guy,
01:02:46
Freeberg. >> Jake Paul was awesome. >> I sat him next to Sachs for dinner.
01:02:49
>> Yeah, he's a great investor, by the way. >> Fantastic. Fantastic.
01:02:52
>> His portfolio is awesome. He had a great team. His team came with him for the
01:02:56
thing. >> Great strategy. Yeah. >> One thing I'll say about Thomas Keller,
01:02:59
he had a strong watch game. I don't know if you guys noticed that. >> Probably noticed it.
01:03:04
>> Nice potential. He had a aqua knot green band rose gold a complication on there.
01:03:08
I was trying to trying to figure out what the complication was. I couldn't quite eyeball it.
01:03:12
>> I don't know if you saw it, Jimoth. >> I was I looked at it and I was like,
01:03:16
"Wow, this is very it's very elegant." Tell us Keller. I mean, the French
01:03:19
Laundry is just transcendent. Did you have a favorite dish? Of course. Was it the duck or was
01:03:27
it the Miyazaki? >> The duck was excellent. The Miyazaki beef was excellent, but the the tapioca
01:03:32
pearls I thought was the best. Oh, oysters and pearls. That's your oysters with caviar. Sax, what was your favorite
01:03:38
there? Do you have a favorite? >> One of the caviar related dishes. >> Yeah, that's oysters and pearls.
01:03:45
Round of golf. The following day, we played a nine-hole scramble with everyone. It was great. So great.
01:03:50
>> How much did you gamble and who won? >> We had a $100 a hole versus Chris Hoe, who's the chief
01:03:58
business officer of Athena. But, you know, that [ __ ] is like a plus one, so whatever. I think we lost three
01:04:04
units. So, we lost >> 300 bucks. >> All right. Shout out to our other
01:04:06
partner, Athena. Athena. Wow.com. Go get a month free. All right. Whatever. I got
01:04:11
to stop showing. Most importantly, everybody was very upset that liquidity sold out and people couldn't get in.
01:04:17
There were like 3,000 applications. It's going to sell even faster uh next time around. So, go to allin.com/events
01:04:25
and get your applications in if you want to come to Summit September 13th, 14th,
01:04:28
and 15th. Freeberg's doing some wild stuff. some wild stuff for that event and uh it's gonna be cool.
01:04:36
>> You're losing your mind. You're now becoming very eccentric, unhinged in
01:04:40
your production value and Shimoth, >> you've been very selective in your
01:04:45
approach to the liquidity speakers because liquidity is different than the summit. The summit is like a festival.
01:04:53
It's a huge party. There's all this really interesting cross-sectional stuff. But liquidity is meant to be for
01:04:59
one thing which is the most important capital allocators in the world should have one place to convene, debate,
01:05:08
learn, talk, build relationships. And I want people to walk out of there with relationships that they can use. And I
01:05:15
told you this before because capital is what shapes the things that occur in the
01:05:20
world. So I think that we have to be extremely selective in how we curate every element of that show. It's not
01:05:25
meant to be that you can just buy a ticket, but to just buy a ticket, which is also a very fair and rational and
01:05:31
nice thing. The summit is a party. It's a vibe. >> It's awesome. And you're going to learn
01:05:34
a lot and have a good time. You'll be entertained. All that great stuff. All
01:05:37
right, let's keep going. Breaking topic this morning. CPI and PPI for May came
01:05:42
in hot. Inflation is ripping. Here we go. CPI is obviously the consumer price index that tracks inflation from the
01:05:51
buyer side. Rent, groceries, gas, healthcare, all that stuff. PPI is the producer price index that tracks
01:05:57
inflation from the seller side. Wholesale goods, raw materials, yada yada. CPI came in at 4.2%
01:06:06
year-over-year, highest since April 2023. There's your chart. PPI came in at
01:06:12
6.5% year-over-year, highest since the end of 2022. Here's your poly market. 21% chance
01:06:20
inflation hits 5% in 2026. Uh and then an adjacent poly market chances of the Fed hiking this year at 49%.
01:06:32
It was under 10% before the Iran war started. By the way, European Central Bank just raised rates a quarter point
01:06:40
on Thursday. This was their first rate hike since September 2023. What do we take from this Freedberg? It's all
01:06:48
related to the Iran war. Yeah, that's the >> there's definitely an energy blip from
01:06:53
the Iran war that drove the core index up, but there's also the macro point which is government spending out of
01:07:00
control, inflation out of control and fundamentally as things unravel you have rising rates. So I think we should kind
01:07:08
of expect especially with a Kevin Worsh Fed, I think we could see north of five and a half, 6% overnight rates. It's not
01:07:17
unforeseeable. So, >> this is just the nature of our monetary policy and our fiscal policy with our
01:07:24
congressional um members that vote all of our future earnings away into [ __ ] But here we are
01:07:31
>> and and far worse >> that nobody expected. Chimath, your thoughts and then I'll wrap up with you.
01:07:37
>> That was the most passive aggressive take I've heard in a while from
01:07:40
>> He's a little upset about it. Yes. Clearly he's a little bitter about just
01:07:44
so stupid to watch. >> Can you tell us how you really feel? What happened? Like honestly, like what
01:07:48
else is there to say? I've said it like a hundred times. It's just idiotic. The
01:07:52
the core problem with wealth inequality in this country, the core problem with inflation in this country, the core
01:07:57
problem, all of it roots back to excess government spending. End of story. All of this, we're all talking about
01:08:05
machinations at the edge of the network. The core of the network is overspending
01:08:09
by the government because in order to get elected, you need someone to give you something more than you have today.
01:08:13
That's how people get elected. And here we are 250 years later. I don't know if
01:08:18
there's the will or the way to get out of it, but one of the manifestations of
01:08:21
it will be in this late stage higher rates and as a response to inflation and here we are, you know.
01:08:27
>> All right, Chim, what's your what's your take on this? Is it a transient
01:08:32
as we uh have had other presidents say? Is it uh is there a offramp for the Iran
01:08:38
war? I don't know if this is above all of our pay grades, but this does seem to
01:08:42
be a challenging moment happening right before the midterms and consumers expected that their cost
01:08:51
of living would go down and that they would benefit from this golden age and it hasn't happened yet. Well, look, it's
01:08:58
concerning. I don't think that I expected the print to be this hot. Now, what's
01:09:04
keeping things in line? there's still a chance that we're not going to see CPI
01:09:09
go absolutely crazy. And part of that is because China has been smoothing the energy consumption globally worldwide.
01:09:16
And so we've kind of kept a damper on $200 a barrel oil. In fact, we're subundred. At the same time, I think
01:09:23
it's pulled forward a lot of more practical sources of energy production, not just in the United States, but
01:09:32
abroad. Specifically, what I mean is solar. just because it's simple, it's
01:09:37
passive, and you don't need you don't need a lot of permitting, and you don't
01:09:40
need clean air permits and the like. >> So, all of these things have been able
01:09:44
to keep a lid on it. >> If China somehow runs out of reserves, and they need to go back into the spot
01:09:49
market to buy an extra 3 million barrels a day, there's a very big risk that oil
01:09:55
gets into the well past 100 and maybe between 150 and 200 a barrel. That has a lot of downstream problems with respect
01:10:03
to CPI. Not even because US energy supply, but really because the inputs that oil feeds
01:10:11
all around the world that then get absorbed by the costs that people pay for everyday things could go up.
01:10:17
>> So I think the PPI number should be an alarm that we use to offramp this around
01:10:25
situation sooner rather than later. I just unless again because we don't know
01:10:29
what happened in China because none of us were there. Unless there's an understanding
01:10:35
that the president has with Xi about how much actual supply and slack is in the Chinese system and that may allow him to
01:10:44
take a much longer road to get a conclusive and decisive win in Iran, but I I don't there was supposed to be some
01:10:51
tangible known resolutions with this China trip. We haven't heard them explicitly stated yet. What's your take
01:10:59
on uh this Sachs? uh the inflation print and is there like a a golden bridge to to end this war in Iran?
01:11:08
>> Jake, the problem is when you asked me that question, people assume that I have
01:11:11
some sort of insight information and I don't look the PPI print is what it is.
01:11:15
I largely agree with what Chamas said. The only thing I would add is that it was largely in line with expectations.
01:11:22
That's why the market is actually up today. Normally if there's a surprise
01:11:26
print and inflation is high then the market is down because the market starts pricing in the implication of higher
01:11:33
interest rates. It's not doing that today. So it's in line with expectations. But I largely agree with
01:11:39
what Chamas said. >> Yeah. All right. Well, listen, we got to get to other elections. You know, my
01:11:43
take on it is just huge colossal era starting this Iran war and hopefully we can get out of it. Um the downstream
01:11:50
effects of this are just disastrous. Um, I'm hoping there's a quick resolution or
01:11:55
some way to get out of it before it goes to, you know, a year or two or three. We
01:12:01
can't be in a forever war, obviously. >> NASDAQ up 2 and a half% today. That
01:12:04
normally doesn't happen when there's a hot print because again, like tech
01:12:09
stocks are the most susceptible to high interest rates. So, the market at this point seems to be pricing in
01:12:15
>> a resolution. But we'll see. We'll see. >> We'll see. Yeah. I mean, it's just this
01:12:19
is a everybody knows Iran is a quagmire and really was was a huge mistake. All right, let's talk about the hand ringing
01:12:29
around vote counts in the LA mayoral primary instead of me reading all these details. Just Freeberg, I know you're
01:12:39
hot. >> I think we know about it. Yeah. >> Yeah. I think if you're on X, it's
01:12:42
essentially your entire feed. Uh Freeberg, what's your take? Is the election system in Los Angeles,
01:12:51
specifically in California, corrupt? And is this election been stolen from Spencer Pratt or is it just
01:12:59
>> There is no election. >> There was no election. >> There is no election.
01:13:04
>> Okay. So there is no spoon. Your rights to have an election are gone. You are a citizen of those who tell you
01:13:15
who your overseers are. You are no longer allowed to vote >> allowed to vote
01:13:20
>> for your elected representatives. They are now appointed representatives. So
01:13:26
enjoy the appointments that have been made by who? >> Appointed by those who have constructed
01:13:31
the matrix. >> Got it. Now I understand the background. There is no spoon. Agent Smith and all
01:13:38
the Agent Smiths are in charge. So, here's some statistics for you. In-person voting the day of the
01:13:46
election in Los Angeles County for mayor. Spencer Pratt 35%. Karen Bass 29%. Nipia Ramen 26%.
01:13:56
mail-in ballots received before election day, 38% for Bass, 28% for Pratt, 20% for Ramen. And then all these ballots
01:14:07
that arrive after election day, 37% for Ramen, 35% for Bass, 19% for Pratt. So Pratt's
01:14:17
post election mailin ballots declined by one/ird. So statistically the population
01:14:22
of people that send in their ballots late reduced for Pratt by a third increased for Nipia Ramen by 80%.
01:14:29
And Karen Bass 10% less if you just look at the mail in ballots before and after
01:14:33
election day as a comparison. I don't know if there's a socopolitical way that
01:14:38
you can assess those statistics and assume that these are individuals casting their individual vote for who
01:14:46
they think should be mayor of LA. And Nick, if you'll just pull up the map, which I think is worth taking a look at.
01:14:51
Basically, the concentration of incremental votes that Nitia Ramen got, came around the Skid Row area in Los
01:14:57
Angeles. And look, I'm not an election denier. I'm not someone who's
01:15:01
historically believed the idea that elections are fraud and people have stolen votes. But when you look at the
01:15:06
basic statistics of what happened in person, mail in before, mail in after election day, it becomes a real
01:15:13
statistical quagmire on how did this sort of a sociopolitical shift happen in such a way that it did. Now, there was a
01:15:21
report published, Nick, if you could pull this up by the US House of Representatives Committee on House
01:15:26
Administration. This was published in May of 2020 and they highlighted the 2018 California midterm elections and
01:15:34
the challenges they saw arise in that midterm elections because of some of the legislative changes that were made.
01:15:40
First, California Assembly Bill 1921 legalized the practice of unlimited ballot harvesting in the state. This was
01:15:46
passed around 2018 around the midterms. What that means is that any individual in the state of California has the right
01:15:54
to go and collect ballots from any other individuals regardless of relationship,
01:16:01
fill them out and send them in. California 2 years later, 18 months later, also passed a law that made it
01:16:08
permanent that every person registered in the state of California would get a ballot. So tens of millions of ballots
01:16:14
then get mailed out. Then there was another series of laws that were passed that said anyone can register to vote.
01:16:20
You don't need to prove your citizenship. You can use a gym membership card as an example. So anyone
01:16:26
can register to vote. There is no proof of ID when you get a ballot. There is no
01:16:30
demonstration that the person who fills out the ballot has anything to do with the individual who's supposed to be
01:16:33
voting that ballot. And it is legal for an individual to go out and collect hundreds or thousands of ballots, ship
01:16:39
them in, and they will all qualify in these kind of mail-in ballot voting processes. So there's nothing illegal or
01:16:46
fraudulent going on. In fact, the system is operating exactly as intended. It has
01:16:51
been set up and structured in a way that with the right construct, you can get an
01:16:56
individual appointed, not elected, but appointed to a particular role in government under a quote free election
01:17:02
in California. This is a foundational destruction of our rights to vote for people in a free democracy. I feel it's
01:17:11
been eroded slowly over time to such an extent now and I'm not some crazy MAGA
01:17:16
mother whatever people want to classify me as for pointing this out but you can go down this list they are all written
01:17:22
out in this document every one of these laws that were passed in California over
01:17:26
time that in aggregate create the environment and the construction for elections to become appointments and no
01:17:33
longer free democratic elections where the principle very importantly the principle should be one individual
01:17:41
one vote and if you opt to not vote, your vote should not be counted. So I think that there's no fraud. I don't
01:17:49
think there's anything. >> It sounds like you believe there's massive fraud.
01:17:52
>> Yeah. I think this is not fraud. This is this is the way the laws are set up.
01:17:55
>> It's not fraud. >> It's the way the law is set up. >> Okay. Sachs, you say a fraud.
01:17:59
>> This is the system of appointment. >> Okay. Yeah. It sounds to me like you're
01:18:02
describing fraud, but you're saying the system allows fraud. >> Fraud is breaking the law. Fraud is
01:18:07
breaking the law. The law allows them to do it. If you paid people on Skid Row, that's not legal.
01:18:12
>> That is illegal. But you are allowed to go out and get anyone to register.
01:18:17
You're allowed to go out and collect all the ballots. And the people that are
01:18:20
doing it, as long as you're not getting paid per ballot, you're legally allowed
01:18:22
to do it. Let me pause for one second. When you look at any one of these laws in isolation, you can understand that
01:18:29
there is some altruistic intention or there can be framed to be an altruistic intention behind it. Increase voter
01:18:36
activity. Give everyone the right to vote. give more people access, more channels to vote, etc., etc. And you can
01:18:42
rationalize away how every one of these laws makes it easier and adds lubrication to the system. And now we've
01:18:49
got a quote more accessible democracy for everyone. But the truth is that there is a concept in insurance called
01:18:56
adverse selection, which is if there is a moment to exploit something, the criminals will exploit it to the extent
01:19:01
that it will become asymmetric. the criminals or the or the do bads, not the dogooders, will find a way to take
01:19:08
advantage of a hole in a system just like hackers do to break into a network >> and that's what you system ramen and
01:19:16
Spencer Pratt did. >> This set of mechanisms, this set of construction has allowed for a system
01:19:21
that now has evolved into an appointment. So you believe worked the system better than Spencer Pratt?
01:19:28
>> She had a better cheating operation. Obviously, you can see it in the chart.
01:19:32
the just look at the number of votes. >> Okay, define cheating here just so we're
01:19:36
clear. >> How does it happen where Bass stays flat and then Ramen and Pratt just switch
01:19:43
places? So on election day, he's the clear number two and then they keep counting for a week. They don't even
01:19:49
know how many ballots they're supposed to have. I mean, none of the stuff is
01:19:51
locked down. Okay, let me explain how the cheating works. And to Freeberg's point, it may or may not be legal. So I,
01:19:58
you know, I don't want to get sidetracked by whether the fraud is legal or illegal, but I think we all
01:20:03
know it's crooked. And here's how it works. First of all, >> yeah, explain what's I'll explain it to
01:20:08
you. I'll explain it to you. And if I if I misstate a fact, then let me know.
01:20:12
>> So, first of all, California sends out ballots to every registered voters. It's
01:20:17
something like 23 24 million people. Only 9 and a half million people vote. So, what happens to the other 14 million
01:20:24
ballots? I have a friend who lives in LA. He manages apartment buildings. He says that the ballots just pile up, you
01:20:31
know, at the mailboxes. Yes. That everybody has in the buildings. There's like loose ballots everywhere. They're
01:20:37
sent to people who don't even live there anymore. He like recognizes that the
01:20:40
names, they're people who haven't lived there in years. So, the voter roles are
01:20:45
dirty. They haven't been cleaned up in years. Okay. Then you have the issue that there's no voter ID on election
01:20:51
day. There's no voter ID to even get registered. When you sign your ballot, they run it through a machine. As long
01:20:57
as the signature matches 40%, I don't even know what 40% means. It's like less
01:21:03
than half accurate, the machine says it's fine. Okay? And then it doesn't
01:21:08
even check the signature of the voter if you have a supposed witness. So the voter can just put like a little X on
01:21:17
their signature as long as the witness says and they sign, but there's no registration of witnesses. Anybody can
01:21:24
be a witness. There's no database of witnesses. And so the witness just puts
01:21:28
a squiggle. So what is to stop the so-called ballot harvester from going filling out these ballots, putting a
01:21:36
little X on the voter, putting a little squiggle on on the the witness, and then
01:21:40
dumping these ballots. There's no chain of custody. So the risk would be they
01:21:44
would be arrested and >> and right >> well what is their expectation getting
01:21:50
caught? Gavin Newsome just signed a law making it virtually impossible to audit any of this stuff.
01:21:56
>> So now even if they thought maybe they could get caught. Let's just take a look
01:22:00
at all the other types of fraud happening in California right now that we know about. We know that there's
01:22:05
billions of dollars in fraud in California for things like Medicaid fraud, unemployment fraud, EBT fraud,
01:22:12
hospice fraud, fraud related to the homeless industrial complex, and on and on. That's all been proven. Okay. Is it
01:22:19
really so hard to believe that some of the same groups, the same interest groups, the same NOS's would be willing
01:22:25
to exploit these loopholes in the dirty voter roles in the millions of ballots that go to incorrect or non-existent
01:22:33
addresses, the non-existent chain of custody, the non-existent signature verification, the no ID, not only to
01:22:40
vote but to register, counting ballots without postmarks if received 7 days later, registering thousands of ballots
01:22:48
from homeless shelters that don't even have any beds. These are all known things. So when you say that there's no
01:22:54
fraud, it's like come on, there's loopholes and all you got to do is look
01:22:59
at the results and see that there's something crooked going on here. >> Okay, so you're in the camp that there
01:23:06
is fraud going on because where there's >> it may be legalized. It may be
01:23:09
legalized, but it's fraud. >> Okay, so you believe there's fraud. Look, when you start mailing out
01:23:15
millions of ballots to people who don't even vote, and then you have no audit
01:23:19
requirements, no chain of custody, no punishments, no penalties, no investigations, and there's a huge tons
01:23:25
of investigations both by attorney generals >> in California. >> By who?
01:23:30
>> Uh I I'll pull up the history. I will say after every election, there are tons
01:23:36
of civil lawsuits filed. Like obviously Trump filed 60 of 60 when >> not in California. Uh I don't know just
01:23:43
made it impossible to do that kind of audit. He just I think that's because he
01:23:48
was beaten so badly in a democratic state. There was no reason to. But I don't know why he didn't file in
01:23:52
California. But he filed 60 of these and everybody kind of does that. They file these constantly and attorney generals
01:23:58
can. So in this case, Chimoth, if there's so much smoke, if this statistic seems so improbable, and there was the
01:24:07
Nick Shirley video of going to Skidro and this possibility that they were paying junkies and people who were high
01:24:14
to vote, this would be easily catchable. Actually, this would take a conspiracy for thousands and thousands of votes to
01:24:21
have been swung. And that's what we're saying here. We're saying upwards of
01:24:24
10,000 votes were swung, but certainly 5,000. Good luck. You would need a conspiracy of hundreds of people or at
01:24:30
least dozens to do that. >> Good luck going to Skid Row and getting some meth addict to go on the record.
01:24:36
>> I think they actually got somebody to go on the record actually. That's a that's
01:24:38
the >> There's a few videos of this. Yes. >> Yeah. Yeah. No, that that's the crazy
01:24:42
part about it is they uh are saying Skidro people got paid two or three dollars and that somebody was found
01:24:48
guilty of this. Brenda Lee Brown Armstrong. She plead guilty in 2026 of paying. system is designed to allow
01:24:56
fraud. It's this it's designed >> it has been won by 43,000 votes to
01:25:02
>> guys. We live in a autocracy in California where one political machine has shaped the laws so
01:25:14
inexurably that whether it's a mayor or it's a state legislator or it's a state senator
01:25:21
or it's the governor, it's a one-way door. There is only one party that can
01:25:27
win and we are slowly seeing that that party increasingly does not have the ability to govern a
01:25:36
state this dynamic and this complicated. They've changed the rules to make sure
01:25:40
that that monopoly is enforced. We have all tacidly approved it by supporting or voting or not saying
01:25:49
anything. And the end result has yet to be unveiled because it's going to take
01:25:53
another couple of decades for all of this rot to really destroy these places. The reality is the voters probably
01:26:00
should be given a chance to have Spencer Pratt push back on what is the machine so that at a minimum what they learn is
01:26:10
that the machine wasn't working nearly anywhere close to the way it is intended. Instead, what will happen is
01:26:16
two people that are in various shapes and forms pretty deeply unqualified and very similar now essentially get to
01:26:25
give you not much of a choice at all. So, I think the the the sad thing is Freeberg is right and Sax is right.
01:26:34
Freeberg is right. The laws have been shaped this way to make legal what would otherwise be an illegality. and Sax is
01:26:43
right, which is that any rational person should look at these laws and say this is fraudulent.
01:26:51
The unfortunate situation is unless there's a some cataclysmic decision in California, this is what we're stuck
01:26:57
with. It is time to vote all of these people out. >> You can't you can't vote. What do you
01:27:04
mean? How you going to do it? Anyone can register anyone can vote. Spencer Pratt
01:27:10
can do this. Can I answer Free Brook's question? >> Sure. >> There is one break the glass way and
01:27:20
I've talked to a few of these guys about what that is. How do you change the system? You must
01:27:26
get somebody at the absolute top of the ticket elected. That person is Steve Hilton.
01:27:33
And if that happens, there is one thing that Steve Hilton can do to start to completely change the way that this
01:27:39
state is run, which is instantly on day one declare a state of emergency. If you
01:27:44
pull on this thread, what you will find, there are lots of things you can do to clean up how California works under a
01:27:51
state of emergency. I'll leave it at that. Yeah, it there's some common sense here.
01:27:59
We shouldn't send ballots out to everybody. that made sense maybe during co and if you can't be bothered to have
01:28:06
ID then maybe you shouldn't vote uh and I don't think that there has been an
01:28:13
election a major election that has been swung by election fraud the Heritage Foundation has spent many many tens of
01:28:21
millions of dollars on this I understand and they have a really great uh project
01:28:25
that tracks this people file lawsuits like Trump did like Al Gore did everybody's always filing lawsuits over
01:28:31
this it's ever enough to swing an election, but there's definitely on the
01:28:36
margins. >> Kennedy versus Nixon. Bro, >> hold on. Let me finish. >> The mob got Kennedy elected versus
01:28:40
Nixon. >> I'm talking in modern history. Yeah, in modern history. >> That's pretty famous. That's pretty
01:28:45
famous. >> Okay, we we can get into if that's true or not as well. The
01:28:51
>> Check out Lynon Johnson's history. Lyndon Johnson only got elected to the
01:28:54
Senate. He got a start in politics through ballot fraud. >> We absolutely should have voter ID. We
01:28:59
should stop sending everybody one of these because the appearance of impropriety is the problem here. Both
01:29:04
sides crazy new flash news flash everybody. Both sides cheat. Both sides bend the rules. One side does it by
01:29:11
limiting access. The other does it by providing too much access. All these politicians, overwhelming majority of
01:29:17
them are unethical. And all they care about is getting elected. And they have tons of people working for them in both
01:29:22
parties to massage these election results. And if you get caught, it is a serious penalty. There has not been in
01:29:30
modern history major elections. Go ahead and go to the Heritage Foundation. They
01:29:34
find all the nooks and crannies of >> You keep using the word fraud like as if
01:29:38
it's a thing, but it's not. I'll tell you who's at fraud. >> No, there's fraud. At the end of the
01:29:42
day, there is fraud. There are people have been caught doing fraudulent things. >> The one group that is fundamentally to
01:29:49
blame is Congress. We have an opportunity. There have been opportunities to pass laws that will
01:29:56
make it illegal for people to vote if they are not citizens, if they are not properly registered, and if they don't
01:30:02
show up with ID. That is a very simple solve to this democratic electoral process. We can't have a situation where
01:30:10
someone is legally allowed to go out and collect a 100 ballots that they find sitting in the lobby of an apartment
01:30:14
building or for people to register to vote with a gym membership card when fundamentally the federal government
01:30:20
already knows if you're not paying your taxes, you're required to have a
01:30:23
driver's license to drive a car, you have to have ID, you have to be registered in the system to do nearly
01:30:29
anything and everything in this country as it is. So to exclude that very simple
01:30:33
requirement from the most foundational obligation that we all have as citizens of this country is inherently wrong and
01:30:42
it needs to be addressed by Congress. And if Congress fixes it and there are federal laws that create systems that
01:30:48
ensure that elections are all about one person getting one vote and they're legally allowed to vote, then this
01:30:55
doesn't become an issue. Short of this, I worry where this goes because over
01:31:00
time, the fewer people actually believe that elections are real. And I hate being on here sounding like I'm an
01:31:05
election fear-mongering guy who's telling people that elections aren't real, but the longer this goes on, the
01:31:11
more difficult it will be for people to actually have faith in the system that we're all operating within. And that's
01:31:17
when things get ugly and it needs to be fixed. >> Yeah, I'm strongly agreeing. If you
01:31:22
can't be bothered to get a driver's license or a passport, maybe you don't
01:31:26
need to vote. Yeah. >> Congress can fix it. And they should. >> Yeah. And and there's, by the way,
01:31:32
there's a ballot measure in California. If people in California want this, they
01:31:35
can just vote for the ballot measure. >> You mean the voter ID ballot measure?
01:31:39
>> Yeah. >> Yeah. That's very important. Actually, look, if that fails, it's kind of our
01:31:43
last chance to save anything >> for California's last chance. Yeah. Yeah. And you do sound like an your
01:31:49
entire speech, David, was you sound like an election denier. >> I don't care what you label it. Look,
01:31:54
the the Democratic party as a whole were election deniers about Trump winning in
01:31:58
2016. They concocted the whole Russia gate hoax to basically try and discredit his election.
01:32:04
>> When you look at the the difference between walk-in and mailin ballots, and
01:32:07
I've seen everyone try to argue why that's so different, it doesn't compute
01:32:10
for me. That's it. I'm not like trying to make up a fact. This is the fact.
01:32:15
Look at this facts on the on the on the screen you're looking at right now. This
01:32:18
is the difference in the data. This data tells me that something is wrong. That's
01:32:22
it. >> Obviously, >> I'm not trying to I'm not trying to just
01:32:25
fearonger because I don't like who got elected. I don't give a [ __ ] what goes
01:32:28
on in LA. In fact, I think it's better for LA if they deal with the socialist
01:32:32
mayor for a couple years so they come out the other side realizing they shouldn't embrace socialist principles.
01:32:36
I think it's great for the country. Actually, I think it's great for New
01:32:39
York. I think it's great for Seattle. I think it's great for wherever Minnesota.
01:32:42
Wherever people want to elect socialists in this country, let's run into it. I'm
01:32:45
all about jumping through the socialist well so we jump out the other side like the Latin American countries are doing
01:32:50
now when everyone wakes up and realize how up socialism is because it denies everyone the opportunity for economic
01:32:56
mobility. It denies everyone their individual liberties that this country was founded on. Socialism is wrong. It's
01:33:02
broken. It doesn't work. And if you all want to go through that lesson, do it.
01:33:06
Go for it. Let's do it. Let them get elected. But the numbers on that election data show me that that's not
01:33:12
what the people wanted. That it's not very clear or evident to me that that's
01:33:15
what people voted for. And that's where I'm like, my god, the system itself has
01:33:19
become a system of appointment, not a system of election. And I get fired up about that personally. You want to add
01:33:25
anything as we wrap here on this? >> Yeah. The party told you to reject the
01:33:29
evidence of your eyes and ears. It was their final most essential command. This is obvious. Just look at the charts.
01:33:36
Look at the data. We know something crooked happened. It's statistically impossible. The only thing we don't know
01:33:42
is whether it was illegal fraud or legal fraud. Meaning they've created so many
01:33:47
loopholes and so much corruption in the law that what they did is possibly legal. It is still corrupt. It is still
01:33:54
crooked. It is ridiculous. Everyone can see it. Do not deny the evidence of your
01:34:00
eyes and ears. Even if they call you an election denier, I personally don't care. I deny it. Spencer Pratt should be
01:34:06
in the runoff. I deny that Ramen won legitimately. >> The statistics are the statistics and
01:34:12
they would scream that you should have the attorney general of California and the federal attorney general look into
01:34:19
it and if they catch people doing this stuff and they have caught one person and Nick Shirley went down there and you
01:34:24
see people on Skid Row, you know, listen, homeless people should have the right to vote theoretically.
01:34:29
>> Why did Nome just sign a law making audits like that even harder? >> Yeah, I'm not defending Newsome here.
01:34:34
I'm saying where's the attorney general? If there is in fact fraud, Trump has the
01:34:39
ability with his hands-elected attorney generals to go after this. He controls that part of uh you know our our
01:34:46
democracy. He should file and they should go figure out what happened statistically. So Trump, President
01:34:52
Trump, thank you, President Trump. Go do it. Go do it. You got this attorney general chasing down everybody else. Go
01:34:59
to this election. Use your powers. Audit it and find them. Where's President
01:35:04
Trump? >> You didn't hear my point that the the state law just made it harder for the
01:35:09
federal government to come in and audit the voter roles. So >> yeah, he can still do it.
01:35:12
>> I'm sure they will do it. I'm sure they'll do it if they can.
01:35:15
>> Yeah. I mean, trust me, Trump does what he wants. If we've learned anything,
01:35:20
>> he will do what he wants. And he's not doing this now. >> What I find amazing is that the media
01:35:25
instead of asking tough questions like how did this happen? How could the votes swing so wildly in ballots that were
01:35:31
only found after hold on after election day? The media instead of trying to demand answers and demand accountability
01:35:40
is trying to cover up and say nothing wrong happened here. It's unbelievable.
01:35:44
It's outrageous. >> What is the best theory steelman? What's the best theory of why this discrepancy?
01:35:49
>> I mean, we're sitting here talking about it. >> What's the discrepancy? Anybody have a
01:35:54
steel man of why this discrepancy could have happened that's legitimate? anybody.
01:35:58
>> If the media is hanging on to a snapshot of the past where people actually cared
01:36:06
what they said and every time that they can actually reestablish trust, they kind of don't do
01:36:15
it because the people that they want to curry favor with are willing to play along. So I think what
01:36:24
happens is the media is like, well, I have two choices. I can actually say that Spencer Pratt should have won or
01:36:30
could have won or investigate why he didn't win, but then I'm kind of tacitly
01:36:34
supporting Trump. And hold on a second. I'm supposed to hate Trump. So, no. Yeah, you know, this is completely
01:36:39
normal. Nothing to see here. Everybody moves on, hoping that then they get ingratiated into the machine, but
01:36:45
they're going to get run over and used like everybody else does, too. So, >> that's what's happening. Yeah, your
01:36:51
point is so spot on, Chimat, that like if it's so polarized, meaning that there's two polls that if you do
01:36:58
question, it's almost like you were being put in the Trump proTrump camp. Just like the word MAGA is thrown around
01:37:03
like, oh, you're a MAGA guy because you're questioning election integrity.
01:37:06
This has nothing to do with MAGA. It's just simple conversation about a particular topic. But if it's diametric
01:37:12
to the other camp, you're, you know, you must be in that bad camp. And so then
01:37:16
they don't want to be put in that camp. It's so true. I'm for mathematical and
01:37:20
statistical literacy and what happened here is mathematically and statistically.
01:37:24
>> So nobody has any steel man or concept of how this could have happened.
01:37:29
>> I can tell you the statistical odds that this would have happened and it's one in
01:37:33
a trillion. >> Okay. What if >> So are you telling me the one in a trillion shot hit?
01:37:38
>> I I'm asking if there is any I'm not >> just just this one time.
01:37:42
>> I'm not I'm asking if there's anybody who steals this in any way. I'm asking
01:37:46
I'll tell you what I've heard integrity of >> Why is it so psychologically important
01:37:51
for you to cling to the idea >> it's not elections that California elections are
01:37:57
completely corruption free after all the things they've clearly done >> to corrupt them voter ID you admit
01:38:04
there's like it's chalk full of loopholes >> you ask me why am I holding on to it I'm
01:38:08
not holding on to it for the purpose of discussion and having an intellectually honest discussion I'm asking the three
01:38:14
of you if there you've heard or if you have any theory just for the rigorousness of the discussion. This
01:38:20
isn't my position. >> Where are all the democracy protectors in the media now?
01:38:24
>> They all said that they were they were the guardian they were the guardians of
01:38:27
democracy. They're going to protect democracy. Where are they to ask question?
01:38:30
>> The only theory that I can think of I'll steal a minute myself. I because I I
01:38:34
think you do need to be rigorous. >> You go ahead first. Yeah. The the only
01:38:37
thing I can think of is her ground game is more sophisticated because they've
01:38:42
been at it longer than Spencer Pratt and he didn't have a ground game that took
01:38:46
this in since he's a non-traditional firsttime candidate. That's the only
01:38:50
thing I can think of. >> Sophisticated a coded word for he wasn't bribing Skidro homeless judge.
01:38:55
>> No, he just has it. He doesn't have an election machine to collect ballots and
01:38:58
collecting ballots is legal. That's the That's the question. >> She had a better cheating operation.
01:39:03
Obviously, >> we're saying the same thing. We're not actually saying anything different.
01:39:07
>> The Twitter commentary was that people were waiting because there's so many people that are
01:39:15
so against Spencer Pratt, but they wanted to be sure that they voted for the right candidate, whether it's Ramen
01:39:23
or Bass. And so the people that held their ballots the longest saw the polling data that Bass was going to win.
01:39:29
That's why they all voted for Ramen last minute was to try and promote her to
01:39:32
make sure >> has been shipped because of the you could have two Democrats at the top of
01:39:36
the ticket. >> That's right. And they didn't want Spencer Pratt to have a shot at the
01:39:39
general. And that by the way, you know, I mean, if you saw my interview with Spencer Pratt, he said Ramen registered
01:39:44
one hour before the deadline and she's friends with Bass and that the whole point was to keep him from actually
01:39:50
making it onto the general ballot. So that some of the coordination that people were making in their minds was
01:39:55
like, we got to make sure we vote for the two Democrats and then they'll figure out, hey, is it wrong? Let's just
01:40:00
box him out. >> Yeah. >> Got it. >> Which is like why they all waited till
01:40:03
the last minute to vote for to vote for how do you coordinate something like that? That that's always the question I
01:40:08
have with these like >> these people are all brilliant geniuses. It's it's hundreds of thousands of
01:40:12
brilliant geniuses row that out. >> Yeah, there's a chat group. I mean, it's
01:40:18
not that complicated. >> Alan Harvesters did it. what Freeberg said just happened. It's just that the
01:40:23
ballot harvesters did it, not the actual voters, obviously. >> And ballot harvesting is legal, but it's
01:40:29
sketchy and maybe they And if there are ballot harvesters, I think that Trump and the DOJ can catch these people.
01:40:36
>> How the is it legal that someone can go out and collect other people's ballots
01:40:40
and fill them out and send them in? It's the craziest to me. >> No evidence of fraud right now.
01:40:44
Statistics don't look good. Donald J. Trump, you control the Department of Justice. Go investigate it.
01:40:50
>> Okay. I love you guys, but I'm hungry. I got to go. This has been >> an amazing episode of the Allin Podcast.
01:40:57
Bye-bye. >> Love you boys. Byebye. >> Let your winners ride. >> We open sourced it to the fans and
01:41:11
they've just gone crazy with it. Besties are gone. >> That is my dog taking a notice in your
01:41:26
driveway. >> Oh man, my habitasher will meet me up. >> We should all just get a room and just
01:41:33
have one big huge orgy cuz they're all just useless. It's like this like sexual
01:41:37
tension that we just need to release somehow. Wet your wet feet. We need to get Mercy's going all in.
01:41:55
I'm going all in.