Stephane Philipakis

"Code is Proof"
Friendly Tech Lead | Angel Investor
Anything code | AI & accountability | Privacy

[Stephane Philipakis @sphilipakis]

This proof-of-age shit is ridiculous. We already know how to do age verification without leaking your ID everywhere. A trusted issuer (DMV, bank, whatever) checks your ID once, issues a signed “over 18/21” credential, and stores it in your wallet/phone. Then sites only get a cryptographic yes/no proof. not your name, not your DOB, not your ID number. And the issuer doesn’t need to know where you used it. This is a solved technical problem. The problem is policy and incentives.

[Hank Green @hankgreen]

Just listened to a clip of Jason Calacanis talking about how most tech companies now "support Trump on various endeavors" with the exception of Anthropic and "this is america you're allowed to have an opinion BUT! This administration takes things personally, and they're full contact."

The overall thesis of the clip as I understood it was "Anthropic is a bunch of Democrats, they hate Trump, Trump hates them back. Why isn't Dario doing the thing all the other tech leaders are doing? Of course Trump is fucking with them."

OK.

I think it's fair to say that the average person would say that Jason Calacanis is more of a capitalist than me. Certainly he's made much more money than me and his job title contains the word "capitalist."

But what Jason is talking about being the new structural quality of the American economy here is not capitalism.

If the President of the United States picks OpenAI to win the AI race even though Anthropic's tech is better, that's not "Oh, y'know, this administration takes things personally!" shrug.

That's actually not capitalism anymore...it's a pretty well-understood different system called oligarchy. That becomes very obvious if you just replace any of the characters in the story with names you don't recognize.

Just in case it needs saying, oligarchy sucks. It is bad for innovation and bad for economies and bad for consumers and ultimately also often (eventually) very bad for the oligarchs! This is why we're supposed to be apalled when our leaders start actively encouraging oligarch-like behavior.

Because, look...funny thing about capitalism that I've noticed...most capitalists (there are some exceptions) have no real ideological allegiance to capitalism. They don't /want/ to compete. Competing is annoying, it's hard, you can very easily lose!

If they can make more money by getting the state to protect them than they can by innovating and competing, they will absolutely pick state protection.

Most capitalists will happily become oligarchs. Their allegiance isn't to any particular system, it's to margin.

Now, people to my left will say that oligarchy is, in fact, the final form of capitalism...it's where capitalism will inevitably ends up. I don't believe this because of how we have lots of examples of capitalist countries that are not oligarchies.

But watching it happen in my country (and watching influential capitalists shrug as it does) doesn't strengthen my position!

It is honestly kinda astounding to see such a staunch pro-market guy just say, "the administration is full-contact, Dario needs to play ball." We all get that that's not a defense of markets, right?

It's a defense of one guy distorting a market because he dislikes any power that is not his power. If you find yourself defending that, it is possibly because you, yourself, strive to be an oligarch. Be real careful with that. You can follow your incentives into an evolutionary dead-end. It happens all the time.

I've just been thinking about it all morning. Just a real jaw dropper.

You can watch the video here: https://t.co/eU29LH9xoR

[Stephane Philipakis @sphilipakis]

everybody says token cost will drop... has it ever dropped? it seems like it's been going up consistently. #ai #llm

[Stephane Philipakis @sphilipakis]

$300 billions question: who will get the contracts for rebuilding ?

[Stephane Philipakis @sphilipakis]

@AlexanderKnigge @FrenchForces @EmmanuelMacron @europeanunion @MistralAI it's satire :)

[Stephane Philipakis @sphilipakis]

Markets and talent flows are what actually open and close capability windows. Barring foreign nationals from frontier models like Fable/Mythos doesn't just protect US advantage, it walls the current leading edge off from large parts of the global market. That gives players with serious capital and national incentives even stronger reasons to accelerate their own development in more shielded environments. Curran's analysis is compelling under open, integrated conditions. These kinds of interventions change the incentive structure and may reopen competitive dynamics he underweights.

[Stephane Philipakis @sphilipakis]

wow, you can't even buy a mac studio with more than 96Gig of mem anymore? what is happening ? 🤣

[Stephane Philipakis @sphilipakis]

Don't take me wrong, I like your approach (see https://t.co/IEECk05Wcj for a different use-case) I may have used your tweet to shit-post the crypto-sphere who has been super quick (you'd certainly recognize it) to roll over and forget their "values" in the past couple of years :)

[Stephane Philipakis @sphilipakis]

This proof-of-age shit is ridiculous. We already know how to do age verification without leaking your ID everywhere. A trusted issuer (DMV, bank, whatever) checks your ID once, issues a signed “over 18/21” credential, and stores it in your wallet/phone. Then sites only get a cryptographic yes/no proof. not your name, not your DOB, not your ID number. And the issuer doesn’t need to know where you used it. This is a solved technical problem. The problem is policy and incentives.

[Stephane Philipakis @sphilipakis]

Crypto: "We’re building tools to escape government control"
Also crypto: "Hey Dario, we built the perfect tool so you can comply with the government’s directive to only let Americans use the good models" #crypto #ai #fable

[Stephane Philipakis @sphilipakis]

I'm slowly starting to think that AI alignment efforts & research will paradoxically mostly serve to secure power for the elite and make the rest of us disposable biomass. #ai #alignment

[Stephane Philipakis @sphilipakis]

Truth is he largely got what he really wanted:
1- Favorable tax environment.
2- Huge payout via a16z’s SPCX position after yesterday’s IPO.
Doesn’t mean the rest of his portfolio or performative principles suddenly became totally irrelevant, but it definitely reduces any personal downside for him.

[Stephane Philipakis @sphilipakis]

When you realize that the model that is supposed to be able to bring down a whole country has been hard at work vibing $10,000 three.js 3D demos for the past 7 days 🤣 #fable #mythos #anthropic

[Stephane Philipakis @sphilipakis]

@pmarca What does this mean? Anthropic is based or the USG is based? may I remind you of what you were crying about before the elections? https://t.co/zRdPvaPvIC

[Stephane Philipakis @sphilipakis]

@Scobleizer @BrianRoemmele that reminds me of the early encryption key days...

[Anthropic @AnthropicAI]

The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees.

The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance.

Access to all other Claude models is not affected.

We apologize for this disruption to our customers. We believe this is a misunderstanding and are working to restore access as soon as possible.

Read our full statement: https://t.co/bwn0sximKZ

[SemiAnalysis @SemiAnalysis_]

What's the better business model for an AI lab, subscription or API? (1/4)🧵

[Stephane Philipakis @sphilipakis]

It's a good book, sharp, contrarian and useful to founders. but as soon as you start digging it starts reading more like an idealistic cult manifesto for tech optimists. plenty of over-simplifications, straw-man arguments with a generous sprinkle of survivor biases all over (like most books tech investors and founders to be fair).
Some monopolies can become those "creative monopolies" ... for a time... but Thiel wants us to ignore the rent-seeking / gatekeeping attractors that have been shown to be powerful destructive socio-economic and innovation black-holes (let's not make those young founders think too much about the what ifs). History is littered with those monopolies too.
Innovation is messy, incremental, accidental. If we tend think that AI-humanity alignment is a hard problem, we also easily forget that humanity-humanity alignment is even harder. Thiel's answer almost sounds like we need to erase choice because choice sometimes leads to bad outcome.
The EU decision is just one parameter in a chaotic game with plenty of disruptive or accelerating events (and innovation is just a tiny tiny emergent element of the game). Personally, I think that we have all been slowly cooking for a while... and that a quick rise of temperature is not necessarily a bad thing :)

[Jeremy Chone @jeremychone]

Rust #genai v0.6.4 Released 🎉

WITH MiniMax provider

More than 25 providers, 200 models

Rust native protocol multi-AI provider.

Built in Rust, for Rust

Give it some stars 👇👇

https://t.co/1QSmlRDyrK

@rustlang @MiniMax_AI

@rustlang