good institutions:
- subsidize shared infra where private markets won't
- improve coordination by reducing transaction costs
- protect people from fraud and coercion
bad institutions:
- create artificial barriers to entry
- socialize downside asymmetrically
@visakanv world at war 1973 doc should be required watching in high school - the modern docs on wwII are soft in comparison
it's unfortunately euro-centric and hopefully there's an asia-centric version I'm not aware of
https://t.co/Llx4uuehrq
Should code ever lose First Amendment protections?
This podcast covers free speech and software, including cryptography, cryptocurrency and AI. It should be the internet's most comprehensive.
You'll hear from experts like @valkenburgh, @prestonjbyrne, @LazPieper, Sam Enzer, and Professors Eugene Volokh (@VolokhC) and @JaneYakowitz.
We trace the history of Cold War export controls through to the recent restrictions on Anthropic's AI models, and explain why the Supreme Court has never definitively answered the question above.
Timestamps:
0:00 Anthropic AI export controls
6:17 Intro
9:39 The First Amendment
10:38 @VolokhC on his free speech philosophy
13:42 Speech vs function
19:24 @valkenburgh interview
25:50 The Cold War, munitions list and Phil Zimmermann
29:57 The Bernstein case
34:53 Strict scrutiny
38:48 The Corley case
49:17 Sorrell, Stevens and where the Supreme Court is heading
53:25 @prestonjbyrne on the Supreme Court's AI case
57:23 Defense Distributed and 3D-printing
1:04:00 Where publication ends and conduct begins
1:15:20 Lowe v. SEC and the agency line
1:45:45 Compelled speech
2:04:17 @LazPieper on Sorrell and third-party liability
2:11:29 @prestonjbyrne on the GRANITE Act and foreign censorship
Thank you to the presenting sponsor of this episode, @altitude.
Nothing in this podcast is legal or investment advice.
They’ve always known that age estimation is not accurate or reliable. Technical experts made that clear from day zero. This is not news to them. They don’t need a “report”.
They know the only effective and reliable way to verify a person’s age is for a company to verify their identity.
How do you force every company to verify everyone’s real identity without forcing every company to verify everyone’s identity?
@jayhinz question: How to get adversarial parties to agree on the state of an open database?
option 1: you don't
option 2: designate an overlord
option 3: economic consensus
@jayhinz question: How to get adversarial parties to agree on the state of an open database?
option 1: you don't
option 2: designate an overlord
option 3: economic consensus
@Arnaudschenk revitalising enlightenment values that underwent a growth spurt in the US and now are looking for a new physical home in the digital age
(while certain parties are trying their hardest to make this impossible)
I'm genuinely confused by the argument here. It seems to be: open-weight AI models reduce the profits of American VC-backed companies; many leading open models currently come from China; therefore open AI is somehow "communist."
We've never applied that logic to any other technology. Linux, Apache, PostgreSQL, and countless other open-source projects destroyed proprietary software margins and dramatically reduced returns for some investors. We called that competition and innovation—not communism.
If the concern is reliance on Chinese AI, make that argument. If the concern is national security, make that argument. But "this technology makes it harder for incumbents to earn monopoly-like returns" has never been a definition of communism. It's often just what technological progress and real free markets look like.
at its heart, education is about asking questions from a place of genuine curiosity, and refusing to settle for flimsy explanations. everything else is project management
"Without freedom of thought there can be no such thing as wisdom and no such thing as public liberty without freedom of speech"
- Benjamin Franklin, writing under the PSEUDONYM 'Silence Dogood' in 1722