TFTC: A Bitcoin Podcast - #759: Open Source Is The Only Defense with Kyle Olney
Episode Date: June 17, 2026Marty sits down with Kyle Olney to discuss the government's escalating war on freedom tech, from the BRCA's dangerous developer loopholes and sudden AI export controls to the urgent need for open sour...ce software as the last line of defense against surveillance and licensing regimes. Kyle on X: https://x.com/KyleOlney STACK SATS hat: https://tftcmerch.io/ Our newsletter: https://www.tftc.io/bitcoin-brief/ TFTC Elite (Ad-free & Discord): https://www.tftc.io/#/portal/signup/ Discord: https://discord.gg/yHGkvYxdqT Opportunity Cost Extension: https://www.opportunitycost.app/ Shoutout to our sponsors: Bitkey https://bitkey.world/ Aven https://www.aven.com/bitcoin CrowdHealth https://www.joincrowdhealth.com/tftc Unchained https://unchained.com/tftc/ Lygos https://lygos.finance/ Salt of the Earth: https://drinksote.com/tftc Join the TFTC Movement: Main YT Channel https://www.youtube.com/c/TFTC21/videos Clips YT Channel https://www.youtube.com/channel/UCUQcW3jxfQfEUS8kqR5pJtQ Website https://tftc.io/ Newsletter tftc.io/bitcoin-brief/ Twitter https://twitter.com/tftc21 Instagram https://www.instagram.com/tftc.io/ Nostr https://primal.net/tftc Follow Marty Bent: Twitter https://twitter.com/martybent Nostr https://primal.net/martybent Newsletter https://tftc.io/martys-bent/ Podcast https://www.tftc.io/tag/podcasts/
Transcript
Discussion (0)
you've had a dynamic where money's become freer than free
if you talk about a fed just gone nuts all all the central banks going nuts so it's all acting
like safe haven i believe that in a world where central bankers are tripping over themselves to
devalue their currency bitcoin wins in the world of fiat currencies bitcoin is the victor i mean
that's part of the bull case for Bitcoin. If you're not paying attention, you probably should
be. Kyle Olney, I welcome you back to the show, but under some pretty heady circumstances here,
we've got a battle for freedom in the digital age. It's raging on, probably more pressing than
was when we spoke last november yeah absolutely the uh the battle for the future of freedom tech
is now yeah and i think ai is really accelerating the battle that's actually it's one thing i'd be
uh interested to get your thoughts on because i think i described ai as equally exhilarating
and unnerving it's typically in the context of oh my gosh this is so powerful it's so cool playing
with these tools i've been able to do more than i ever thought would be possible as a one-man
team playing with a clanker than a one-man team we have a six-person team here at tftc but i think
you catch my drift and it's unnerving because it's like oh gosh what's this going to mean
for the state of the job market if if this is truly disruptive i think on the job market side
i've become more convinced that i'm not sure it's going to be as disruptive as many people
are fear-mongering about but the other side of that that coin is exhilarating because you can
do all the things that i just described and it's unnerving because we're sort of speed running
into it making an integral part of our workflows and it's almost like it's a bit of a honey trap
where we're getting addicted to it and the government the governments of the world are
observing. And over the last week, I think we've seen the first big shot across the bow of how the
governments think we should be able to utilize these tools, particularly with the export ban
on Fable 5 last week. Yeah, I mean, I think you nailed it right on the head. The whole world
changed in the last two weeks out here in Silicon Valley, at least, right on the front lines of
what's happening with AI, both, you know, with the Anthropic change on pricing models and people
kind of being forced to confront the real cost of tokens that they're using. And then, you know,
in very quick succession, the U.S. government, you know, coming in to, you know, pull back and
implement export controls on Fable 5 and, you know, kind of the newest class of models that
Anthropic had released. It's, uh, heads are spinning, uh, heads are spinning out here in
Silicon Valley and people are still trying to figure out what this actually means and kind of
where things are going. But those of us in the, uh, you know, kind of Bitcoin community and the
freedom tech community are, are well aware, uh, where this is going. We've seen this before. Uh,
we've been fighting this battle for years now. Uh, and, uh, I think, you know, I don't want to
jump too far ahead uh with the ai stuff i actually think it's great to to maybe take one step back or
two steps back and kind of start with uh you know the brca and clarity and market structure and kind
of like the debate around uh you know our ability to access these freedom text freedom technologies
generally uh that's going on in congress and then i'd be happy to connect that through line straight
from what we see going on in BRCA and Clarity to AI and Anthropic
and the Department of War, the U.S. government, all that?
Yeah, let's start with BRCA and Clarity because I think that's taken a backseat
to the AI stuff in the last month as everybody looks at the saber-rattling
between Anthropic and the U.S. government and obviously last week's actions
by the U.S. government with those export controls.
But I think the last I heard on BRCA and Clarity was that it's getting held up in an ethics committee because they want some language in the bill about insider trading or the ability of the Trump family specifically to play with any of these assets.
And then the other one was the Law Enforcement Blockchain Association letter that went out a couple of weeks ago where you had, I think, a couple hundred law enforcement officials sign a letter basically saying we want clarity because we've been promised that it's going to expand the Patriot Act to make it much easier to track people using this.
Uh, yeah, it was, it was quite a, quite a turn of events to, to see a bunch of law enforcement
officials, uh, coming out in favor of clarity. Um, but yeah, you, uh, you kind of hit it on the
head. So there are a number of different countervailing issues. Uh, I don't know how
many of these that we discussed in, in November, but, uh, as you just mentioned, there's kind of
The overarching issue that, you know, has been, I would say, more navel gazing for the last, you know, nine months, which is, you know, Coinbase and, you know, kind of the stable yields fight against Wall Street.
You know, that's one dominant thread that is still unresolved.
It's worth mentioning.
I mean, we've heard that there's a deal that's been reached between, you know, some different parties who are, you know, representatives of the banking industry and in the Senate.
But, you know, clearly we saw Jamie Dimon come out two weeks ago and, you know, kind of drop a gauntlet on Brian Anderson and what was going on with Coinbase.
So I would say we're definitely not out of the woods on that front.
You know, Wall Street is not going to just roll over and accept a completely, you know, crypto native reframing of the finance industry.
Right. And I think that that's that's one major problem or category problem that we have with the Clarity Act.
Then, as you mentioned, you know, there's this other aspect of, you know, things that we care about as Bitcoiners deeply,
which is the BRCA and developer protections and the ability to know that if you're developing tools and open source software that utilize blockchains
and you're not actually taking custody or providing, operating the service yourself, that you won't be prosecuted by the government.
This is still an unresolved issue and frankly is, I would say, the most concerning issue, again, from a Bitcoin perspective,
Because the language that they inserted in the finance committee in order to get it voted out and brought to the floor added a clause or, you know, what I would call a loophole that allows, you know, prosecutors to continue charging, you know, developers.
and basically, you know, kind of inviolates the safe harbor in circumstances where, you know,
the government deems that you had knowledge or should have had knowledge of your ability
to be helping nefarious actors, you know, with the technology, which, you know, I've said elsewhere,
but, you know, to me as a builder, as a, you know, entrepreneur, as somebody who's, you know,
been involved in the building of this technology like that feels like a loophole big enough to
drive a truck through uh and should be very concerning to everybody and then yes the third
issue uh as you just mentioned uh which nobody really knows how it's going to land because it's
all kind of up to one man uh is the ethics concerns uh you know and what we're hearing is that uh
or actually what's been stated clearly uh by uh you know the senators uh the two democratic
senators who voted it out of the Finance Committee to the floor of the Senate, they said very clearly
that they would not vote for a bill that didn't contain strong ethics provisions, you know,
against, you know, kind of the Trump family crypto dealings. And clearly, that would seem
to be a deal breaker for, if not the entire Republican Party, Donald Trump. So we've got
three different layers of problem that are still confronting, you know, the Clarity Act and its
ability to get passed into law. And it's very unclear at this point whether any or all of those
are going to be able to be resolved. And certainly before we make it into the election later this
year. Yeah. What is it? Is Congress in a recess right now? Are people on the Hill? I believe the
Senate is actually in town this week, but, uh, Congress is off this week. Then Congress will
come back, uh, for a couple of weeks. And then we basically have, uh, the 4th of July holiday and,
and the summer recess. So there's less than 30 days left in the legislative calendar before we,
you know, actually, uh, get to the election season. Uh, and, you know, we've now, I think,
blown past six self-created deadlines for getting this bill done over the last nine months. And,
you know, the latest deadline that they were hoping for was to get it on President Trump's
desk by July 4th. That clearly is not going to happen. It's literally a physical impossibility
at this point to make that happen. So I think the game that they're playing is to try to cram
something in before everybody goes home for the summer recess and hope to have something that
would be signable maybe in September when everybody gets back from Labor Day. But that's
getting really close to the election. And it's hard to say why Democratic senators would vote
in favor of a bill that would broadly be perceived to be handing Donald Trump a signature win going
into the next election. So, yeah, the politics and the policy are both very challenged at the
moment. You know, it would look like a signature win for Trump, but as it stands today, I don't
think it would be a signature win for Bitcoin or specifically. No, I mean, yeah, if we I would say
that we've marched right up to the line of what is tolerable or acceptable to Bitcoiners with the
latest uh carve-outs in that brca provision 604 in the bill dive into that specifically like what
what is the carve-out i know you loosely characterized it but what are the specific
things they're leaning into yeah well i don't have the language right in front of me i'd have
to pull it up but basically the clause that was added to the uh to section 604 of you know which
as the BRCA section of the Clarity Act, was basically this language that says that the
safe harbor, well, so actually maybe to take one step back, the original construction of
the BRCA was that it created a safe harbor for developers of open source software so
that they didn't have to fear prosecution under any circumstance if they never took
custody of customer funds, right?
So that's a pretty clear, bright line that everybody would be able to understand.
If you're developing software, it would be clear and obvious to you whether you're taking custody of user funds, whether you have the control, the actual ability to move somebody's funds on their behalf.
And that was one of the reasons that the BRCA has been very popular for multiple sessions in Congress now.
It's been bipartisan.
It was voted out almost three to one out of the House with that language.
You know, this is a common sense, you know, sort of application of the, you know, regulatory guidelines that we had already been given by FinCEN years ago.
And it was just meant to anchor it into law so that it was clear to everyone that this wasn't a regulatory interpretation.
It is literally actually the law. It should be the law.
If you don't touch customer funds, you are not a money transmitter.
The problem is that by adding this clause that they've added where you now have to consider the intent of the builder in building or publishing the software as to whether they were aware or should have known that their technology, their tool, was going to be used for nefarious purposes or was being helped to abet the funds that were gotten through illegal activities.
these uh this would violate the safe harbor and then you would now be able to be charged
uh as a developer for intent right to uh conspire or collaborate or whatever word you want to use
with you know these people who are doing bad things with your technology well uh i think it's
important to bring up the precedent that's been set here already which would completely i mean
if it gets passed with that language would render it completely in effect because roman storm right
in his case they were using an email that he never responded to as evidence that he had
knowledge that somebody nefarious may have been using tornado cash correct yeah correct and in
the case of uh the samurai developers you know as well like the literally the reason that they
created this subclause, which I think is D, subsection D in 604, they created it so that
they didn't, quote, lose the ability to charge samurai developers or the tornado cash developers.
So the government or the people who, you know, the senators who have added this language to the
presumptive reason for the compromise is to be able to continue to allow the ability for
prosecutors to prosecute people like the samurai devs and the tornado cash devs uh so that in
itself should tell you everything about uh kind of how the brca has been modified already now uh
you know if you ask the lawyers in washington you know people like coin center and uh you know
others who are kind of on the front lines litigating on these things uh you know bpi as
well. You know, the lawyers are are less concerned by this language, because they think that, you
know, the legal standard for proving intent is, you know, pretty high. And, you know, you would
have to have some meaningful paper trail, like you can't just accidentally fall into, you know,
intending to help bad actors with your software. That would be the theory that the lawyers would
like, uh, to, you know, feel good about, but again, as a builder, as somebody who's, uh, you know,
developed technology and, you know, as an open source proponent, I don't think you can ever know
how somebody is going to use your tool. And, you know, once it's been published to the outside
world and, uh, in doing so you, you know, nobody is going to run the risk that they might get
ex post facto rounded up and thrown in jail for 20, 30, 50 years, right? So the kind of policy
objective of, you know, making America the crypto capital of the world and trying to keep the
industry in the United States and harvest the economic gains as a policy objective that come
from, you know, the blockchain economy would seem to just be completely violated and kind of,
you know, thrown in the bin by trying to take this approach that they're trying to take by
having their cake and eating it, too. And so this is actually kind of the worst of all policy
outcomes, in my opinion. It's where we've come to the point where if anything were to happen
with this bill or rather if this bill were to be passed, even with its current language,
It's unclear that this would even stop or forestall the desire for blockchain developers to leave the United States, exit the U.S. jurisdiction and go elsewhere and practice, you know, or build their companies in places where, you know, that threat of prosecution just doesn't exist.
well i mean i've seen the lawyers i saw um peter valkenberg and zach shapiro making the intent
case and it's like well you can say that it's really hard to prove intent but like again
samurai devs are in jail roman storms on trial like they they without the language even existing
there they they sort of inferred intent and again i think the roman storm case of he had an email in
his inbox that he never responded to and saying if you check the email like you knew it was
happening therefore you're guilty so it's uh it's complete bullshit my mind i think it's a non-starter
in terms of actually protecting devs and that gets to the question like what is what is the state
I mean, it seems like despite the language in the BRC section, in section 604, it seems like even without that being a point of contention, it seems very unlikely this could pass anytime soon due to the procedural process that you referred to earlier with the lack of timing between now and midterms.
And then who knows what happens after midterms.
Yeah, exactly.
I think at this point, you know, the odds of this bill getting passed, even in its current language, is pretty low.
And we know for a fact that it can't pass in its current language, right?
Like we know for a fact that the Democrats in the Senate are going to try to weaken the BRCA even further because to them, even the current carve out that allows prosecutors, you know, kind of this loophole to continue charging people for violating, you know, kind of, you know, whatever theory, legal theory they can come up with around, you know, facilitating bad intents or bad actions in the real world.
it's just not tenable. It's not tenable for anyone, really, at this point. And so it seems
like we're heading toward a bad conclusion, from a political perspective, at least. It's not going
to be the win that everybody was hoping it to be when we started out on this process. And again,
from the perspective of Bitcoiners, this may actually be a good thing. Back to our earliest
conversations about clarity. You know, we were saying that no bill was better than a bad bill.
And, you know, we've seen all of the trade groups in the industry line up and say that a bill
without strong BRCA protections is a non-starter for them. I think most people on that front have
been just kind of, I won't say it was naive hope, but they've been kind of, you know, hoping that
we were going to get something better than it seems we're going to be able to get and so you
know kind of now i would say that uh you know our backs are kind of against the wall as a you know
freedom tech and liberty oriented community like you know at this point uh if we can't get strong
brca protections then i think we would prefer not to have any bill because the alternative is that
it radically expands bsa surveillance requirements for every licensed or you know uh you know kind of
above board operating company in in america uh right that has that touches a blockchain like if
you touch a blockchain you basically have to surveil all of your customers and report all
of that information to the government which you know again i think as bitcoiners we we kind of
find uh onerous and yeah abhorrent so um yeah it's challenging times uh for freedom lovers in
the blockchain community so freaks this work was brought to you by good friends at bitkey
guys things happen things get lost you need to replace your phone life happens and bitkey has
been designed with this in mind uh so if you're looking for the easiest way to secure your bitcoin
off the exchange go get a bitkey doesn't come with any seed phrases there's no single point
failure it's a two or three multi-sig you have your device you have the mobile app and block
stores a key in the server so you have collaborative multi-sig there on top of that the new bitkey has
a screen so you can actually see and verify what you're approving you can look at transactions
addresses account changes the difference between trusting and knowing uh has been built into the
new bitkey with the screen they've changed code delegation uh and other advanced features like
inheritance recovery if you're looking to get your bitcoin off exchange you haven't done so
You're looking for the easiest way to do that. Go get a BitKey. You can use our code today,
TFTC, to get 10% off the new BitKey. So go download the BitKey app, use the code TFTC to
get 10% off the new BitKey and start your self-custody journey today. This episode has
been sponsored by our good friends at BitKey. Freaks, look at me. I'm glowing. I've got like
an angel's halo going around me. You know why that is? I feel good. I feel taken care of. I feel
blessed, healthy, happy. And that is because I'm a CrowdHealth member. My family and I have
crowd health members for five years now literally this month five years ago we joined crowd health
we've had two babies we've had multiple health events and we're never going back to health
insurance crowd health is crowdfunded healthcare so you you sign up for crowd health you pay a
monthly fee you help out with other people's bills uh and it's significantly cheaper than
health insurance we were on cobra it's a family of three when i when i left my last job before
I went full-time to CFTC. Went on the CrowdHealth. Now as a family of five, we pay, I believe,
$700 a month. It's significantly cheaper. They're going to negotiate prices lower for you.
They've consistently negotiated healthcare prices as much as 50%, 60%, 80% in many cases.
They help out with babies. If you have a pregnancy, you pay the first $3,000 and the
crowd covers the rest. If you have a regular health event, you pay $500 and the crowd pays
the rest go to joincrowdhealth.com sign up today use the code tftc opt out of health insurance i'm
uninsured baby and i love it use the code tftc at joincrowdhealth.com you'll get 99 a month for
the first three months that you're on the crowd health platform in the community big winners you
found sovereign money now find sovereign health and sovereign health care yeah well we were talking
about the bsa on the first episode and how that should really be the the focus uh is is repealing
the bank secrecy act and since then we've met in person uh and had uh had a dinner in austin
and i think there was an update on your end from the the language of the bank secrecy act and um
particularly the appeal was there appeal to the supreme court in the 70s or 80s
and uh what's the story is there it seems like there's a chink in the armor of the bsa that we
didn't talk about the first time you're on i really wanted to lean into and that has to do
with um i think reasonable um like reasonable amounts of of lower thresholds of triggers that
would um trigger a suspicious activity report yeah so uh you know i guess i'll copy out by
saying i am not a lawyer uh and this does not constitute legal advice but uh i am a legal
nerd, I suppose, almost went to law school. So I'm very interested in such things. And yeah,
I mean, I think without turning this entire episode into a BSA episode, I guess what I would
say is that both the legislative and the judicial history of the Bank Secrecy Act
raise a lot of questions about how it has been applied and extended in the modern era.
and there are a number of concerns that people who like i and you and others who have looked at this
and i should say a lot of the other policy organizations again like peter von walkenberg
coin center has been quite vocal about this over the years um you know that the bsa has been
wrongly applied and it has is basically unconstitutional uh and it only survived its
very earliest sort of challenges for this, you know, I don't want to call it a loophole, but this
like, kind of factual oddity, which was the when the DSA was passed in the 70s. The threshold for
reporting was, you know, the inflation adjusted equivalent of about $130,000. So, you know,
the bank would not have to produce a suspicious activity report, you know, unless it was $10,000
in, you know, again, the 1970s, right? But in the statutory, you know, construction of the bill,
there was no inflation indexing that was ever done. And so, you know, obviously, everybody
listening to the show will be familiar with how much inflation we've had in the last 50 years
in the United States. But, you know, $10,000 now today, I mean, just to give you an example of
of how commonplace and every day at $10,000 transaction is. I mean, it's like I moved out
of an apartment in San Francisco and the landlord wrote me a check for my security deposit in my
first month back. And I couldn't even deposit the check at the bank, uh, because it was over
their limits and they needed to like do a physical ID check and, you know, wait a bunch of time for
it to clear. And they gave me this whole song and dance about it. Right. I mean, it's, and it was
all because of the SAR reporting and BSA flags and all of these sorts of things that have to
happen in the background, which are just massive compliance overhead headaches for the banks.
I mean, again, as a kind of former bank regulator in a prior life, my personal opinion is that you
don't even need these kind of surveillance tools that the BSA provides in order to do proper
regulation. It's really just an excuse to dragnet surveil everybody in broader society. And back to
the kind of jurisprudence of the BSA, you know, there were two challenges that happened in the
70s after it was first passed. And in both of those, some of the most famous justices on the
Supreme Court in the 21st century were strongly opposed to the BSA and thought that the arguments
in favor of it were kind of ridiculous and kind of undermined core constitutional guarantees that
we have in the Bill of Rights. But they felt that the threshold for application was so high that it
was something that, uh, you know, was tolerable, uh, in that moment. And so, yeah, again, like
the chink in the armor here, I would say is that, uh, as we see even more onerous overreach on the
behalf of government to surveil, you know, kind of the everyday transactions, uh, and data stream
that surrounds all of us living in a, you know, society where, you know, digital payments are
ubiquitous uh you know it's becoming obvious now that you basically don't have an expectation of
privacy in the united states anymore uh if the bsa continues to be rolled out to the level of
you know every four dollar micro transaction that you might uh you know engage in in the economy
on either a stablecoin rail or a layer two of a native blockchain is going to be surveilled.
Again, I just think that if you were to look at the original construction intent of the bill,
we've drifted so far away from it.
And if you were to show the justices in the 70s who barely found the BSA passable
in the contemporary moment, what has happened today and how it's being applied. I think they
would flip the table, frankly. And so I think, you know, broadly speaking, I think that we should
feel the same way as Bitcoiners and we should push back and do everything that we can to fight
against this, you know, what I would call an unconstitutional invasion of our privacy rights.
Not even just as Bitcoiners, this is Americans. Flip the table. We should take it back to the
Supreme Court, the reasonable threshold of $10,000 that existed back in the 70s and was a point of
contention for justices who thought at the time, like, ah, the amount of people that are making a
$10,000 transaction so low, it doesn't really bother anybody. But today, that would be interesting
data to get. I'm sure somebody's run the numbers, but the frequency of $10,000 payments on a daily
basis today versus 1970 and adjust that for population i'm sure it's still orders of magnitude
higher yeah well and i think again kind of more importantly to the point like it's really theater
right it's kabuki theater uh even in its current incarnation because we see people like jeffrey
epstein getting a thousand suspicious activity reports uh from bank compliance officers uh moving
a billion plus dollars through the pipes over a 10 year span, and literally nothing was done about
it. So if you needed more proof that, you know, they're not really using these suspicious activity
reports, at least not proactively, for, you know, kind of their stated purpose, I don't know what
better example you could find than that one. Sup freaks, this report was brought to you by
good friends at Ligos Finance, Celsius, BlockFi, FTX, they took your Bitcoin and gambled it away.
LIGOS can't because they never hold it.
Non-custodial lending on Bitcoin's base layer.
Your keys, your collateral.
It's verifiable on-chain.
Go to ligos.finance.
Tell them that TFTC sent you.
All right, freaks.
You know me.
You know I don't take sponsor money from products I wouldn't use myself.
So listen up.
The Avon Bitcoin Visa Card is one of the most interesting things I've seen in the Bitcoin
lending space in a long time.
Here's the deal.
You can get a line of credit up to a million dollars backed by your Bitcoin without selling
a single sat.
no gains, no annual fees, no minimum draws, and your Bitcoin is custodied by BitGo,
which is one of the most trusted names in digital asset security. AVEN never lends it out. There's
no rehypothecation. You stay in control. And guess what? You can lock in a fixed rate for up to 10
years. That's 10 times longer than most lenders out there, or go interest only for up to five
years. Rates start at 7.99% APR. For a product that lets you keep your stack and still access
liquidity it's hard to beat i mean the duration in the rates is the best i've seen in the market
to date you also get two percent unlimited cash back every time you use the card spend fiat keep
your bitcoin the whole game if you've been stacking for years and need liquidity without
triggering the taxable event this is worth a serious look go to aven.com bitcoin that's aven.com
bitcoin check it out i'm bridging this back to section 604 the brca it may be a segue bridge
to the ai conversation too it's like i'm sure you've been paying attention to it but like what
calais has been doing with cashew particularly with the ability to run a mint in a trusted
execution environment which would completely separate the min operator from anybody using
it like they would have no visibility into this and i think we discussed this in november when
you're first on but uh we have the ability to re-architect the banking infrastructure
on top of bitcoin natively in a new way more efficiently more privately with better payments
rails and you could see something like the brca passing in its current form with that loophole
language and just completely curbing this but if you look at what that a single example of running
a cashew mint in a trusted execution environment in a secure enclave it's almost impossible i think
it is impossible for a mint operator to know anything that's going on with the mint um
unless somebody explicitly tells them but this is like technology i want to see
and it makes you worried for developers like cali and others working on the protocol like
are they going to get in trouble for really pushing the limits of what's possible with this
new tech yeah well i mean there's a reason that cali isn't anon uh right i mean he's probably wise
uh to remain in anon uh and continue doing what he's doing without attaching his you know real
world or government identity uh to you know what he's doing for uh those of us who are fans of the
technology and think that it you know ought to be something that we're able to engage with freely
uh you know in normal society but yeah i i think the through line here is so this is this is kind
of like you know we were talking just before the call and i was trying to like kind of sketch this
picture and i don't want to like you know kind of connect the constellation for everybody here
right like the fight that we're having right now over uh the brca and the clarity act and the right
to utilize open source software that doesn't harm other people is, I think, becoming the primary
through line that we're going to experience in all of the technology battles that we're
encountering probably for the next at least five years, if not the next decade. Because the action
that the U.S. government took by implementing export controls on the model and the weights
that are coming from Anthropic and, again, trying to make it in the export control so that it was
targeted at individuals of foreign national origin, that is a very scary escalation of the
precedent that we were already dealing with in the brca and it's now taking it and extending it
to the level of uh basically digital id right like how you know part of the reason that it
sounds like anthropic was uh forced uh and we don't really know what happened here we've seen
lots of different characterizations from everybody and so you know it's kind of there's kind of a
fog of war. But nonetheless, what we've been told is that Anthropic couldn't comply with
the U.S. government's edict to restrict access to foreign nationals because they can't even do that
easily inside their own company, right? Like they have foreign nationals who work at Anthropic,
who have access to the source code. They have, you know, they use it every day. They probably
have to use it in order to do their job and so the ability for them to comply with that order
just broke right on its face even internally inside the company right on day one to say
nothing of the millions of people around the world uh who are using the commercial version
of the model and uh again that is something that uh you know we've now gone from the level of uh
kind of export controls on physical hardware, chips, things of this nature to software and
ideas. And that, I think, should really scare every American all around the country. Because
if your ability to access a core tool that most people will need in order to engage in daily
civil society moving forward is being controlled by a bureaucrat somewhere who has very limited
understanding of the actual technology underneath it and you're incapable of uh challenging
you know their their restriction and you can't even be told exactly why or how it's being applied
uh i don't think that that means that we are sovereign anymore as citizens inside this country
right like that that would seem to to argue that you know your access to a common tool or commodity
that you need in order to live your life uh exists at the end of uh the discretion of a state agent
and how they feel about your ability to access that technology
or that tool today versus tomorrow or next week.
And so, yeah, I think that this is a very scary precedent
that hopefully is waking up a lot of freedom tech advocates around the world.
I know for a fact, the last two weeks here in Silicon Valley,
a lot of people suddenly woke up and realized the importance
of open source software and open source weights and models
and open source intelligence in a way that they really hadn't
for the prior two and a half or three years of the ai you know debate that's been going on here
in the valley we'll expand on that a bit what are people recognizing what are they saying
what are they saying as it pertains to the anthropics battle with u.s government and i'll
add it was a bit disconcerting last week to see the reaction of people laughing at anthropic which
I think to a degree, rightfully so, of Dario being very public about his desire to have this technology licensed and tightly controlled by people on the frontier in collaboration with the governments.
And it's almost like a ha-ha look, you got what you wish for, but you're not going to be the one.
It seems like the U.S. government doesn't want to ordain you as the frontier model that's going to write the licensing regime and really, really push this forward.
But a lot of people were throwing shade at Anthropic and David Sachs, a lot of respect for him.
And obviously he was the kryptonite eyes are for Trump for a bit there.
it's no longer there but he was giving the governments he had that tweet i think he sent
it out friday or saturday basically saying the government giving the government's perspective is
we asked anthropic to do this they pushed back we said you can't do that you gotta shut it down
blah blah blah but everybody was positioning it as the u.s government had the right to do this
because anthropic wasn't playing ball but to your point it's like how do we know what the intent of
the u.s guard like if we it's a slippery slope like if if we do it now for this like what kind
of precedent does that set and i think that's one thing that many people are trying to grapple with
is this again this idea that on one side you have this incredibly powerful technology it's nation
uh anthropic specifically it has fear-mongered about the particular model
mythos um table five which got export controlled um is a guardrailed version of mythos but
dario basically said this is the strongest model ever it's going to destroy all these critical
systems and they had the operation glass wing where they gave access in the model to all
large software companies to make sure they could harden their systems and operating systems and
uh it seemed like just observing the amount of software updates have been pushed out in the last
few months that they were able to unearth some and so you have this um this push pull it's like
they are powerful just to play devil's advocate does the government have a good argument to be
made to to throw export controls on this because it is that powerful and if so how do you move
forward and ensure that individuals have access to, to the software without digital ID or anything
like that. Yeah. I mean, there's, there's so much to unpack in this, uh, based on everything you
just said, I guess the first thing I want to touch on and not let just kind of like go by is that,
um, clearly the U S government and, you know, kind of, uh, authorities around the world have
an interest in making sure that truly sensitive uh existentially dangerous uh capabilities do
not leak through something like a frontier model uh right i mean there are real use cases like
bioweapons uh you know chemical weapons things of this nature that like you know you just you don't
want any random person with access to you know an anthropic account to be able to spin up a
bioweapons lab that can cause a mass pandemic, right? Like, that's very obvious. I don't think
any of us would disagree with that. I think like all rational observers would say, yes,
that that is a valid and legitimate goal of, you know, the U.S. government and authorities to
prevent, you know, broad impact to their citizens in this regard. However, I think what we have to
do is we have to take a step back and really ask whether the approach that they've taken
serves their overall policy objectives, right? And I think this is the place where they probably
missed the mark the hardest. And, you know, this is where I would argue that we suffer greatly
in Washington, D.C. and kind of like legislative policymaking in the United States for there being
very few software people on the East Coast and in Washington, D.C.
They don't really understand the physics of the Internet.
They don't understand information theory.
It's clear that they don't understand how this is actually going to work
once they took the actions that they took,
because there's a much larger conversation that I'd like to have
around the various other pieces of what happened
and kind of the impact in Silicon Valley,
But like specifically as it relates to the U.S. government's objectives, right, what are the two things that they would want?
They would want the U.S. sort of regime or like model of the world, the U.S. champions to be the winners, right?
You know, the people who are setting the standard around the world.
We don't want to live in a world where Chinese models, air quotes, are the things that are determining everyday access to intelligence for the average person.
And then the second thing is, is it actually effective for preventing the thing that we're discussing, which is the catastrophic tail risk outcomes?
So on the first measure, I would say that cutting the world's access to the model and restricting anybody's ability to utilize that model, both inside the country and outside the country, clearly is not extending the lead of the American model companies, right?
Like you're literally shutting them down, preventing them from doing business.
This prevents more people from adopting that particular technology.
It pushes people into competing platforms, which are likely to be non-American, right?
And outside of the reach of U.S. jurisdiction.
So, you know, I would say just on the face of the first objective, clearly it's a failure, right?
You can't really argue that you're advancing the U.S. sort of strategic dominance equation on model adoption or kind of Americanization as the standard when you're restricting everybody from accessing the leading American offering.
uh more importantly what i would say is that uh that policy error could make sense in a universe
where you actually had secure ability to prevent that information or those capabilities from
leaking out elsewhere right like if you did legitimately have a monopoly on you know the
supply chain or the physical material or you know any of the kind of skeuomorphic
comparisons that we would make to like nuclear enrichment, for instance, right? Like there are
specific pieces of technology, specific, you know, production methods, specific, you know,
raw materials that need to go into nonproliferation. Like those are places where you can
execute just a full blanket blackout and probably still maintain control. But again, this is the
whole uh thrust around open source software and information theory and kind of the physics of the
internet right like you can't dictate you can't stop knowledge from spreading around the world
right like that is not something that is really a capability of any state authority right like we
and we know this as americans like you know the whole idea to win against the soviet union was to
you know, drown them in Levi's and rock and roll. Right. I mean, and despite the fact that,
you know, the Russians and the Soviets didn't want those things coming in and kind of disrupting
their social system, they did. Right. Because you can't stop these things from that are
decentralized with decentralized production and decentralized distribution from making their way
to the average person in any economy or around the world. And so, you know, I think the challenge
here is that, at least from the outside, the charitable read would be that the US government
understands that they can't prevent this knowledge, this intelligence, this capability from
leaking out around the world. And they're kind of just relying on the idea that we're that far ahead
of our competitors, that this is maybe buying us around or two around the board before the next
guys catch up. But again, if you're paying close attention, you know that, you know, the, uh, the
models, the Chinese models are only 30 to 90 days behind the U S frontier labs. And they've taken
an open source approach to the distribution and production of their, uh, AI technology.
And so, you know, they are becoming the global standard, right? If you are somebody in Africa,
or let's just say the global south, right?
Like South America, Africa, any non-G10, G20 country,
you can't afford a frontier model subscription, right?
Like the only thing that you're going to be able to afford
is an open source model.
And so we've already got the playing field tilted
in the direction of what at the moment
are primarily Chinese open source models
because the US market is not investing at the moment
in producing open source intelligence, at least at this moment.
You know, a couple of years ago, Mark Zuckerberg and kind of Obama
and the various things that were happening there, there was a thrust,
but they've since kind of faded in their position.
And because of this reason, not only would it seem that the Chinese models
are going to have wider distribution than the American models
from a cost perspective, but now just pure accessibility perspective.
The other reality is that we can't stop any of those non-American models
from eventually converging to the same level of intelligence and capability
that the frontier models have.
Because at the end of the day, this technology is just kind of a probabilistic engine,
running over text.
And you can't classify probability.
you can't classify text plain text right like there's no way to do that without implementing
full-blown censorship on speech so you know it's kind of an inevitability that capabilities are
going to rise in the non-american intelligence market and uh again all this is like a very long
way of saying i don't think that either of the u.s government's policy objectives were well served
by the decision that they made to implement these export controls in the last week.
It would seem that they, you know, again, it's kind of a desperate plea to bide time
and hope that maybe we can get smarter on the policy internally in the United States
in the brief respite that, you know, this, you know, restriction provides.
So that's kind of the challenges I see it, at least from a, you know, geopolitical and
government perspective. What are the people in the Valley saying? How scared? How big of a shot
across the bow was this? I mean, it was massive, right? I mean, it was almost an eight-point
earthquake, I would say, right? I mean, the rest of the economy is not fully adopting AI just yet,
but everyone here in Silicon Valley is trying to integrate it. Everyone's been token maxing for
months now. And actually, this is what the lesson that taught everybody why token maxing on
proprietary closed source models was not a good idea. So two weeks ago, when Anthropic changed
their pricing scheme and basically kind of dropped the veil on what it actually costs
to provide inference to the end users, I'm sure you actually, I think I heard you talking about
it on one of your prior podcasts you had even had the moment of uh hey this is getting really
expensive like maybe i need to figure out how to lower these costs and make this more more
tolerable for our business right like you had that experience didn't you yes yes yeah so you
know this quite well like and it affected you know uh you know an outfit even of your size right
let alone the ubers and the facebooks of the world where they you know blew through uh probably a
multi-hundred million dollar budget for the year in one quarter, right? The challenge is that these
proprietary models are insanely expensive. And they're insanely expensive, both on an absolute
basis of what it requires to pay for the actual costs of the delivered tokens right now, but also
they're incredibly expensive compared to the open source alternatives that you could pull off of
the shelf. I think the last number I saw is that they're 10 to 20 times the price per token of what
you get from open source intelligence and so, or open weights and models. And so, you know, if you
are, uh, you know, the, I used to work at Uber, so I'll, I'll just use Uber as an example here.
I don't want to like put words in anybody else's mouth. And to be clear, this is a, uh, contrived
example. He didn't actually say this. I just happened to work at the company previously and
I understand how it works and what was going on.
So none of these numbers are real.
And, you know, none of this should be taken as, you know, a statement from Uber.
But, you know, if you're the CTO at Uber, right, and you've got a multi-billion dollar
budget on a yearly basis that has to support tens of thousands of engineers who are operating
at breakneck pace in 180 countries around the world, and your primary barrier to servicing
greater customer demand that you can't even keep up with is how quickly you can ship software
and iterate and get it to market and build this deep backlog of features and capabilities that
you're going to keep your team busy for the next 10 years. Of course, you're going to lean into
something like agentic engineering, where you can spin up 100 or 1000 agents that can work 24-7,
365, and do all the things that you need to be done while you're sleeping. That's an obvious
conclusion, right? The problem is twofold. One, when the real cost of those tokens comes home
to roost, as we just saw, all of a sudden the real dollar cost of, oh, this is what it's going
to cost me to be a fully agentic engineering organization at enterprise scale, that's 10 or
20 or 50 times what we had budgeted in our capex for engineering spend this year. That's a huge
problem. And structurally, you need to refactor that machine in order to make it work just from
a budget P&L perspective. But then the US government comes along the next week,
while Silicon Valley was already digesting like, okay, like, what are we going to do? Maybe this
is the equivalent of buying a super expensive machine, but like, I don't ever need a human to
operate the machine again. And so like, maybe the payoff period is like, you know, five to eight
years, but like we get enough, you know, on the backside that like the, you know, upfront cost
makes sense. Like people were already furiously working through the economic model of like,
how do we rationalize this after, you know, six months of token maxing?
Then, you know, the government comes along and says, oh yeah, you can't serve that model to
anybody in the outside world. You know, shut that thing down immediately. Well, if you're the risk
officer or the CTO or the CFO inside Uber, and all of a sudden, you can't build and ship
your product now on a daily basis, or your velocity drops by 10x compared to what it
was last week, and you got no warning that that was going to happen, and you have no
idea how long that freeze is going to be implemented.
How do you operate a business, right?
This is the very reason that publicly traded companies have a risk management team on the
board and the risk management team often will not let you single source core technology
from a provider.
You need to have backup providers because what happens if this one guy that you use
for this thing goes out of business, right?
Are you going to have continuity?
Are you going to be able to stay in the game?
Are you going to be able to continue generating revenue?
Like when Uber, when a, you know, again, like a backend service at Uber, right?
Like when you're serving at the scale of billions of users, like if you have something that goes down for an hour, you're talking about potentially tens of millions of dollars in lost revenue.
I mean, how much impact to velocity, to dollars served, to everything, you know, that the businesses were operating on under the new kind of token maxing model has been lost just since Friday last week, right?
I mean, I would argue that the economic impact across the broad economy is probably totaled in
at least the hundreds of millions of dollars at this point, probably billions, right? So I'd say
like those are the first two layers of, you know, the layer cake that everybody in Silicon Valley
is kind of working their way through and trying to understand. And again, I think like I've seen
more people who did who never understood the the purpose or the value of open source software
just wake up suddenly once their access to this intelligence model was taken away from them
suddenly with no warning and no clarity around when it's going to return
yeah well that last point it's like when is it going to return are they going to allow to return
like what does that mean for anthropic training new models like do they have any clarity on
what they're allowed to train and release and yeah i mean i don't we don't know what the back
room negotiations are like i guess between the administration and anthropic but um but yeah i
mean you can imagine that this is this has got to be damaging for anthropic's business too
right i mean again if you were uh interested in an american champion being the primary leader
of this paradigm uh which i would say inarguably anthropic would seem to be at the moment
uh kneecapping them and putting an indefinite suspension on their ability to compete with
whatever their best product is in the market certainly would not seem to be helping them
uh you know be more efficient or more productive or more cash flow generating uh right like this
can't be good for anthropics business it's definitely not good for any of the people who
were using or relying on Anthropic. So again, I think that this level of chaos, you know,
we've kind of become accustomed to it a little bit in the Trump administration with all the tariff
madness and kind of like the general reordering of the chessboard. But these are really
consequential, really meaningful twists and turns on the roller coaster that are, I'd say,
shaking everybody up uh you know across silicon valley because we're everybody i think is being
forced to ask the question for the first time like okay well if i don't actually have access
to the underlying model or the weights myself can i put this in any kind of a you know critical
path workflow in my business when the u.s government can just come in and shut it down
any day of the week with no warning or Anthropic shutting it down any day of the week for no
warning, uh, with no warning. Right. It's, it's kind of the same problem on, uh, you know, just
a different, uh, impetus or a different person instigating the challenge. So I think all of this
is to say that like, you know, open sources is poised to really have a moment right now. And I
think we should really lean in and support that, uh, you know, as Bitcoiners and Americans, because,
it is really the path to greater liberty and prosperity for everybody in the economy.
Well, I do too, but you have to ask the question, if they're doing it to Anthropic,
what's to stop them from doing it to any of these open source labs?
Well, I mean, in America, I suppose the argument would be that once you've published the software,
it's out there, right? So I guess they could come preemptively stop you from trying to publish the
software or you know again you could end up in the scenario that we are talking about with you
know with the samurai doves where they went and talked to a regulator and asked for guidance they
were given guidance they followed the guidance and then the government still charged them anyway
um you know we uh could see that kind of behavior i mean i think one of the fears that uh i certainly
have uh looking at what's going on now is you know we've been lucky so far to have avoided
kind of draconian or onerous, you know, state licensing or, you know, national security
restrictions on this technology until Fable. And this would signal to me that there's been a change
of feeling in D.C. about the fact that they are going to move in the direction of doing licensing
and regulation uh over intelligence and ai and i think that that is going to be incredibly dangerous
uh because we know that the government uh i mean here we are 15 years later in bitcoin and we still
don't even have sensible market regulations that we're talking about uh having been turned into law
can you imagine you know if if uh you know we have to go you know why basically the fear that
i have is that the government may have learned the lesson in the wrong direction they think that
they waited too long on the bitcoin thing and like now they're going to try to do it way early on the
ai side um and that's just going to lead to to disaster because i think even the some of the
smartest people in the world on ai don't really know how to land this plane uh and get it to where
we're going right because the reality is that i don't think we really know it's like kind of like
3d printed guns or something like that right like if it's just information uh you know how
do you prevent that information from moving around the world at the speed of light right
you kind of have it's not really uh something that we understand how to do in the physics of
the modern internet no i mean my hope for this export control move by the government on anthropic
is that it was just a pedantic move
stemming from Anthropics
for a confrontational relationship
with the Department of War specifically.
And hopefully it was them.
I mean, not even hopefully.
I mean,
I guess best case scenario,
it's the government just saying,
we're going to shove you in a corner, Anthropic,
because you want to give this kill decision
leveraging your technology.
um which is a fucked up thing to say but uh again i think that's best case scenario worst
case scenario it's like oh shot across the bowel full licensing regime coming uh we're
going to completely cuck the models um going to limit the access to these models to certain
individuals with certain that meet certain id credentials uh and we're going to very aggressively
try to stamp out any development on the open source side because it's systemically
um dangerous to to the safety of our country and the children and national health and security
yes well it's it somehow is always for the children uh you know but you know that's the
other thing like the uk age verification laws just went in yeah i know this is and i think
this is kind of what i mean about the policy policy regime feels like it's shifted now in a
kind of scary way right like the stuff that we're seeing out of uh the uk with chat control and
digital id and you're kind of banning the use of social media under the age of 16 uh regardless of
you know kind of your feelings on that last point um right like just the fact that they
like how they implement it right the only way to implement it is to create a national id and
to have you sign into every device that you touch with a KYC AML identity, right?
That is incredibly terrifying.
It should be incredibly terrifying to everybody because that means that every web search,
every inference request, everything, anytime that you're touching anything with a screen,
you're being surveilled, monitored, and 100% of your activities are being associated to
your, you know, real world identity, right? Real world government identity. And again, this is
something that, you know, as freedom loving Americans who believe in, you know, the right
to free speech, you know, and free commerce and free, you know, transactions between two
willing parties, like, I just don't know that we can consent to a regime that looks like that,
Uh, right. I mean, again, we've kind of been moving in this, uh, you know, kind of surveillance,
total surveillance direction for 25 years now since September 11th. But, um, we really have to
draw a line in the ground somewhere, uh, and say enough is enough and start pushing back.
And that would be my contention. I mean, I think that the, the importance and the momentous moment
that we need to have with freedom tech right now, uh, it's never been more important than it has,
you know and then it's become i should say this month uh right the only other time i can think of
that is even equivalent is you know the original encryption wars uh right completely agreed and
it's really uh i'm sure just a coincidence that the uk age verification laws were sort of rushed
to market two days before the uh the uk rape gang inquiry report was released by uh by the uh
restore britain party um all right with all this in mind what's the call to action for the audience
out there i i'm very happy that you were able to do this on quick notice because i know that you're
on top of this and i i think i agree with you that most people don't understand outside the
valley at least don't understand the gravity of the precedents that are being made right now
like i said earlier it was disconcerting last week to see everybody cheering on the government's move
because they viewed it as like a ha ha look you got what you wish for anthropic but i think if you
go a layer deeper it's it's much more uh it's much graver than than the u.s government versus
anthropic specifically like this is a fight for freedom in the information age and
he who controls the information is going to have the most power as we continue to transition to
this age. And so having the ability to leverage these intelligence tools and give them context
that is specific to the information and the data sets that you care about is going to be important
to give the individual leverage against stronger forces than themselves. And if these licenses,
regimes, and digital ID get thrust through, the ability to do that is severely hindered. And
And therefore, you're not really free at the end of the day.
And the government has all the power and all the leverage.
Yeah, I mean, I think you said it right, right?
Like if your access to intelligence, if your access to, you know, free, you know, kind of unsurveilled money exchange in the broad economy can be revoked by a bureaucrat you've never met for a reason that they don't have to disclose at the drugstore.
of a hat uh i think that we have to ask like whether you actually have rights uh or whether
you you know are kind of just a permissioned node in the network right so as far as uh call to action
for everybody who would be listening to this i mean i think starting at the top like first of all
the game's not over with the clarity act uh and the brca i mean we can still have an impact here
I think it's incumbent upon everybody listening to this, if you haven't already, pick up the phone, call your Congress people and your senators in particular right now, make sure that they know that this is super important to you, that, you know, developer protections and open source software are your most important issue that you care about and that you're going to vote with, and that you will not support any candidate who doesn't, you know, share your views in that way.
I think like that's not a fruitless endeavor, although it is an uphill battle at this point.
And I think, again, it's like super urgent that people do that as much today as it has been for the last, you know, 12 months or so.
And then I would say that understanding the connection here to what is now going on in AI and the fact that this is really the same fight and that it's going to continue and extend into AI and your ability to access intelligence freely in the future and utilize it for your own reasons.
you know that is another existential risk that we're staring at right now and so you know it's
important to make sure that your representatives understand uh your feelings on this issue on both
of them or how it applies to both of them i should say uh and then i would say like getting organized
and in the fight and uh for open source is the next most important thing for everybody to do
here, uh, right. Like being, um, you know, unapologetic about the fact that, you know,
as Americans, we deserve access to open source freedom technologies. Uh, that is, you know,
literally, uh, should be our right, uh, you know, as, as citizens of this country based on,
you know, the social contract that we all sign up for. Uh, and it's incredibly important that,
uh, everybody begin to orient themselves and understand, uh, you know, both the value of
open source software, uh, and freedom technologies in this area that we're moving into, because
if you don't have, you know, the ability to, uh, run your own model, uh, and control your
own weights, then you are subject to the whims and fancies of whoever else that's not you controls
that model, whether it's anthropic saying, well, we've developed mythos, but, you know, only a
small chosen priesthood of people will get access to it, uh, in order to protect themselves from
the worst of the effects that we expect to come back from this technology, or whether it's the
U.S. government saying, well, we just don't think anybody should have access to this.
And so we're just going to unilaterally shut it down in a way that nobody can appeal and that
there is no due process around. That is just, again, I think a bridge too far. And it's something
that we as Americans, as Bitcoiners, as freedom loving, you know, sovereigns in a country that
is supposed to be you know uh by of and for the people should not tolerate couldn't have said it
better myself kyle uh thank you for continuing the fight the good fight as you made pretty clear
the fight uh is not over it may just be beginning so i'm sure we'll have many more checkups in the
future as uh both these stories progress in parallel awesome well great to be with you marty
And yeah, look forward to catching up and continuing the talk about freedom tech and
the need for open source software.
Oh, yeah.
Peace and love, freaks.
Thank you for listening to this episode of TFTC.
If you've made it this far, I imagine you got some value out of the episode.
If so, please share it far and wide with your friends and family.
We're looking to get the word out there.
Also, wherever you're listening, whether that's YouTube, Apple, Spotify, make sure you like
and subscribe to the show and if you can leave a rating on the podcasting platforms that goes a
long way last but not least if you want to get these episodes a day early and ad free make sure
you download the fountain podcasting app and go to fountain.fm to find that five dollars a month
get you every episode a day early ad free helps the show gives you incredible value so please
consider subscribing via fountain as well. Thank you for your time. And until next time.
