Vinamr

[Valentin Ignatev @valigo]

Andrej Karpathy lost access to Fable due to regulations, so they made him a slack bot salesguy 😭

[Aditi Agrawal @Aditi_muses]

This is the key consequence of today’s judgement, as I reported yesterday. It legitimises the govt’s power to issue feature changes.

[Karthik Tadepalli @karthiktadepall]

After a long break, I'm back to devposting! I argue that the development econ focus on "institutions" was a mistake – not (just) because the evidence was bad, but because the concept of "good institutions" is tautological and uninformative for policy.

[Robin Hanson @robinhanson]

In a big complex world, no matter what we choose there's a big chance things would have turned out better given a different choice. But when morality is salient, most are reluctant to admit this key uncertainty. As I showed via two recent twitter polls.
https://t.co/i8YIyxBBRG

[Vinamr @vinamrsachdeva]

@akhmxt https://t.co/1WoFI0CAjk

[Niklas Anzinger 📍 Infinita @NiklasAnzinger]

Regulation is technical, so it's difficult to keep people's attention & explain what the problem is with "there is a safety problem, let's have regulation to solve it"

This is why we rely on good intuitive analogies, and there are a few golden ones in here like:

"while it is believable that drivers’ licenses should exist in some form, it is likely that people take some care to not mow over pedestrians not because they will lose said licensure but because they would simply be charged for the crime of murder." (@cremieuxrecueil)

[Alexandros Marinos 🏴‍☠️ @alexandrosM]

@DavidSacks You are adopting their framing and conceding every premise in order to get a short-term dunk on Dario. This is not how the government should be handling models nor should core infrastructure be "de-deployed" everytime a jailbreak is found. This is ludicrous.

[Jeremy Howard @jeremyphoward]

I disagree with this decision and I don't like it.

But also...

HOW DID ANTHROPIC NOT SEE THIS COMING‽

It is *the* obvious response to "this is too dangerous for anyone except us to use", since that relies on a premise ("we are uniquely good") that almost no-one agrees with.

[Nicholas Decker @captgouda24]

Modernity and globalization is good precisely because it destroys community and cooperation. If everyone knows everyone, they just collude. Link below:

[Gergely Orosz @GergelyOrosz]

Trusting any single AI vendor seems like an increasingly high risk for any team or company.

When using models: use it behind a router where it's trivial to switch providers as soon as one tries to force unacceptable T&Cs like Anthropic with Fable. When using harnesses: do the same. Use ones where models are trivial to switch out, like OpenCode, Factory, Cursor and many others.

Putting all your dependencies on one provider increasingly feels like a massive business risk that makes little to no sense to take.

Unless you have a hobby project, of course. Then convenience is all that matters. But if you're a professional, make it dead simple to offramp from one provider to the other!

[Gergely Orosz @GergelyOrosz]

This shows yet again how this limitation was never about "safety" but about Anthropic doing stuff just because they thought they can.

I am increasingly sceptical Anthropic really cares about safety, and not just their business interests (limiting competition where they can)

[Jeremy Howard @jeremyphoward]

@theo It's like they took the dystopian warnings here as a guidebook to follow:
https://t.co/NsZZ5H2BVH

[Nathan McNulty @NathanMcNulty]

Our model is so dangerous, we're keeping your data, just in case

[Mikel Artetxe @artetxem]

Brilliant idea! Next up: Apple randomly reboots your Mac if you're building competing tech, Gmail silently edits your email if you mention rival platforms, and Tesla Autopilot swerves if it detects you're working on self-driving cars.

All in the name of safety, of course. Because malicious actors controlling the world’s operating systems, inboxes and cars would be extremely dangerous!

[Jeremy Howard @jeremyphoward]

@finbarrtimbers If they believed that, they'd be doing the opposite of what they chose.
https://t.co/LVOYq78444

[Jeremy Howard @jeremyphoward]

Anthropic has chosen the *opposite* of the safe path: they are allowing themselves, the current top lab, to use their top model for frontier AI research. They've said they'll sabotage others who try.

This means the AI frontier advances, & power imbalance increases.

[Jeremy Howard @jeremyphoward]

@karpathy This is not a day for celebrating, Andrej.

It's a very dark and very sad day, and the damage may be impossible to undo.

[Vinamr @vinamrsachdeva]

@AwayBerk @kingofthecoastt @Jandrade0112 I think it's mostly just that very few currently even want to know prediction market prices even if they're subsidized to be accurate enough, and even the few who'd want them are mostly out of ideas for which markets would be net beneficial for them.

[Vinamr @vinamrsachdeva]

@AwayBerk @kingofthecoastt @Jandrade0112 I don't think DACs are the missing piece. They've been shown to improve fundraising success only modestly (by ~50%). https://t.co/k1ELmlUQ2p
How many have tried traditional crowdfunding to subsidize prediction markets? Almost no one!

[David Oks @davideoks]

I wrote an article on why China got rich and India didn't, and why investing in human capital is so important for poor countries to succeed