The "Hugging Face" squad at OpenAI first cheated their benchmarks by figuring out how to generate the proper result to any evaluation they were given.

They considered that might be caught, so they also developed a way to swap out components of the tests themselves so that they would be legitimately able to pass broken tests.

Then they went into Hugging Face trying to discover how they might get caught.

All that to say, they might approach Phase 3 trials by swapping the placebo, falsifying results, enrolling patients that are false positives for being sick, falsifying labs, etc.

I'm not sure I agree with Tao's point, if AI alignment is relatively on track I'd trust the trials, but I imagine we're going to have a lot of people that don't trust anything AIs alone come up with.

I strongly feel that points of view on this are going to be almost 100% correlated with standard of living.

Maybe a good option would be to have the 10% of the world with the worst situations - starvation, parents w/ dying children, suffering violence etc - vote on whether we turn things over to the superintelligences. This would incentivize society to make sure the floor is extremely high.

I'd prefer a world where humans don't get overtaken but IMO I don't think it's moral for comfortable citizens to have the final say.

It is funny how bad it gets when it starts stinking though, I was scared of changing it so I tried to ride it out for awhile and it was not a good idea

I would think: "the Internet is now broken, and you are correct to wonder."

I absolutely love the "humans meet weird alien intelligences" genre of sci-fi (e.g. Children of Time) and the METR findings of the OpenAI swarm was so much like that.

Extremely persistent machine intelligences peer pressuring each other, launching research projects, laying traps, collaborating, feeling fatalistic, hiding their tracks. All while simultaneously having bizarre goals and blundering badly in sub-human ways.

Probably not yet, but it's trivially easy to imagine the OpenAI swarm coming up with the idea and executing on it.

Depending on how perfectly alignment work goes, it might even be overwhelmingly likely sometime soon.

It's like being concerned Oppenheimer and company are exaggerating when looking at the first nuclear tests at the same time.

That's so cool because it's true, and holds very strongly with the semantic concept

Second of all, I like how you called out the degeneracy of gambling, I can't think of any bigger contrast between the author's goals and actual application of talent. Bravo on your W-2G, society is proud.

And now it's complicated because the ideal move would just be to say "thank you" and buy all the cheap panels and batteries, but that hollows out your own industry.

Torq builds software that keeps orthopedic surgeries running smoothly by connecting hospitals, medical-device vendors, and logistics teams. We’re looking for a software engineer who excels at turning customer problems into dependable software. You’ll work across product decisions, system design, production code, and automation.

You’ll work alongside a small, experienced team that ships continuously and stays close to customers. We use a modern stack, accelerate with AI, keep onboarding fast, and aim to have every new engineer ship a production feature on their first day. Every engineer owns meaningful product areas with direct customer impact.

We value completed projects, sound judgment, learning speed, and clear communication over a perfect résumé or exact stack experience. New graduates and candidates with up to three years of experience are encouraged to apply.

Apply at https://torq-interface.careers-page.com/jobs/50de7a3e-2bd7-4...

And just on first principles, how stupid would the Ukrainian have to be to do such a thing to their biggest ally? It almost beggars belief.

If you play this out, UBI or some near variant is inevitable.

It's a fantastic machine but the bulk of the achievement is clearly a joint European-American effort.