atemerev

Yeah, let's also ban 3d printers, drone motors, and YouTube chemistry videos. If you want to be dangerous, there's a world of opportunities already, with existing open models quite sufficient for a lot of interesting cases.

No. Conclusions are.

"Lock the weights, weights are dangerous" again. Sorry, keeping such power in few hands is inherently more dangerous than distributing it.

AIs do not kill people, people kill people.

Technofeudalism is a misnomer, as feudalism requires people to exploit their labor, which are linked to some territory or domain. When labor loses its value, it is no longer "feudalism", as there is no value to extract anymore.

No one personally uses API prices. If you need to do something yourself, you need a subscription. Ans subscriptions shinkflated at the same time, but a lot.

Amazing how this considers people will still be needed to direct AIs for something.

Well now the accounts are tracked with your real biometry which will be more and more common every month thanks to Persona.

And then, when they ban you, you will not be able to create a new account and continue. Every ban will be forever. Like getting on the no-fly list.

Quantum _mechanics_ cannot be understood, it is full of paradoxes. It contradicts classical EM and special relativity, among other things. A toy theory that was an important step to what came later.

Feynman's own QED is a part of QFT, but this is no longer quantum mechanics. Special relativity and quantum physics are reconciled since then. Now, we only need to reconcile GR and QFT, but this leap was so far beyond our attempts to scale it. Yet, QED/QFT/SM can be "understood" as there are fewer obvious paradoxes that plagued QM.

"Was", unfortunately. He died this September. What a loss, he earned a Nobel long ago.

When I learned about causal sets program and algebraic quantum field theory, I finally (somewhat) understood quantum physics.

Tl;dr spacetime does not exist as a global background where things are "placed", it is not ontic. "Things" are first, and spacetime is derived from interaction statistics of "things". This eliminates nearly all naive paradoxes like double slit etc. Now, these "things" are somewhat complicated and non-intuitive (e.g. there are no individual "particles", they are a limit of discrete distributions), but that's a good start.

Originals are in the torrent

Well, this is exactly how people voted for Hitler, and I would have assumed people know from history how it ended when "moral approach" was thrown away.

Apparently, everyone forgot already.

I wonder why no one writes that all UWB devices - airtags, phones, headphones - form the opportunistic P2P network perfect for mass surveillance. Best case scenario, all phones and airtags can be localized on demand by FBI - yes, the current one. Worst case scenario, by FSB.

I think that AI is the best thing that happened to art. It elevated _all_ human art to this mystical realm. All human work became sacred, as it is unique and bears the Secret Fire. Whatever you do as "art", as long as you didn't use AI, it automatically bears the sacral stamp. Can go to a gallery, whatever. All human art is now elite art.

To Russian war machine.

If you have the compute, I have the expertise.

Well, if the solution is a manual workaround that forces a better query plan, this needs to be a part of Postgres itself so it builds better query plans automatically.

Then it is the https://en.wikipedia.org/wiki/Bellman%27s_lost-in-a-forest_p... - also unsolved

But we can't program it yet.

Just recently I used Fable 5.1 and Astra to solve and prove a long-standing mathematical problem I was always interested in: the shoreline search problem (a ship in total fog is at unknown distance from the shore [infinite line], what is the optimal trajectory?) A particular kind of logarithmic spiral was conjectured 33 years ago; a few days ago, I have obtained the proof. How routine it has become.

https://arxiv.org/abs/2609.24454

Real-world Factorio, aka composable manufacturing, aka "factories building drones building factories", aka universal constructor, is the next trillion dollar company.

I had some exposure to architecture and structural engineering. I am so very sorry to disappoint you, but strictness there is much overhyped.

Well sorry, still have to get decent AI somewhere. Productivity without AI is about 5x less. I am not comfortable with paying Chinese companies, and no Western companies provide subscription-based pricing for open models.

I think recursively self-improving universal intelligence is a good thing, and tool alignment is what creates instrumental convergence and therefore a bad thing. The paperclip parable is about tool alignment gone wrong.

If he was an agent from the start, I doubt we would have learned this much publicly. The materials would have been much more valuable to Russia as pressure points, compared to the public release (yes, of course they spinned it properly, US is watching you and they are hypocrites etc etc, but this was damage control and closure from their part).

As a Russian, I am not 100% convinced that US are evil imperialists [at least before 2025]. Have some humility, there are many more evil imperialists. You are probably not even in top 5.

Absolutely not a joke.

As someone who emigrated from Russia and don't want to ever return here (nor I will be able to unless I want to spend my life in jail), I don't understand how anyone could possibly want to voluntarily "flee to Russia". Doesn't compute.

I have published a LLM-assisted proof of a long-standing problem in convex geometry (70 years) while not being a specialist in convex geometry. I am a scientist and I have education and scientific experience in computer science and systems biology, but I was never a mathematician. So, well, you can definitely work across fields at least.

von Neumann could. The others, I doubt it.

Initial definitions of AGI (as of 2014) are long surpassed. Then, AGI was defined as something that is general enough to work in many fields and orient themselves. It did not include being smarter than humans or even having comparable intelligence to humans. By original standards, we have AGI already for some time.

Current AGI definitions are intentionally vague.