It's basically a $75 Raspberry Pi knockoff board with a $20 SDR radio, a plastic case and some dolphin branding that's likely to get you in trouble.

very much putting him on target to sign more EOs than Truman did nearly a century ago.

To be fair, Truman was ending a war and starting two new ones and Trump is valiantly battling uh... 51 concurrent national emergencies.

My understanding is that LLMs are built on a foundation of ANNs, and that indeed the backpropagation used to train ANNs is a statistical process;

Two responses. One, that's discussing individual-neuron scale processes rather than collective processes; and this was a discussion about inference, not training. Human neurons also learn by error minimization (Hebbian learning). But this does not describe the macroscopic processes that result from said minimization.

* During training, neurons develop into classifiers that detect superpositions of concepts that collectively follow the same activation process. Individual neurons weight their input space and subdivide it by a fuzzy hyperplane to achieve a classification result.

* In subsequent layers, said input space is formed from a weighted combination of the previous layer's classification; thus, the superpositions of questions being formed are more complex, as are the classification results.

* In a LLM, this iterates for dozens of layers, gaining complexity at each layer, to form each FFN

* The initial input space to a FFN is a latent (conceptual representation), as is the output; the FFNs, in result, function as classifier-generators; they detect combinations of concepts in the input space, and output the causally-resultant concepts into the output space

* FFNs alternate with attention layers dozens to hundreds of times in order to process the information, each layer building on the results of the previous one.

The word to describe that is not "statistics". It's "logic".

In a LLM, the first few layers focus on disambiguation. If there's a token for "bank", is this about a riverbank, a financial bank, banking a plane, etc? As the layers progress, it starts building up first simple circuits, and then progressively more complex circuits - you might get a circuit that detects "talking like MAGA", or "off-by-one programming errors", or whatnot. In the late layers, you have the general conclusions reached - for example, if it were "The capitol of the state that contains America's fourth-largest metro area is...", you've already had FFNs detect the concepts of fourth-largest metro area and encoded Dallas-Forth Worth, and then later taken that and encoded "Texas", and then finally encoding "Austin". And then in the final couple layers you converge back toward linguistic space.

Anthropic has done some great work on this with attribution graph probes and the like; you can detect what circuits are firing, and on what things those circuits fire, and ramp them up or down to see how it modifies the output. They very much work through long chains of logical inferences.

While they are doing creative things with the UI (but not really to my taste), at the end of the day it's just another Chrome browser, so no manifest v2 support and no uBlock Origin. While Vivaldi puts in their own ad blocker safety system, it's nowhere near as good as uBlock Origin. I don't understand why Vivaldi devs don't patch manifest v2 back in. Would be a huge selling point.

So Firefox remains the only game in town for safe web browsing. I keep Vivaldi around for the odd site that won't work in Firefox, but it's definitely not my daily driver.

The old method of building chess players was heuristics. You have a bunch of engineers who hopefully know something about chess come up with rules about how good a position is. That's what "heuristics" are.

The thing that made chess computers better than any human and go computers better than all but the very best (or probably all by now) was using neural networks that learn their heuristics by experience. And yes, human mathematicians absolutely do learn heuristics, i.e. "gut feelings" regarding good approaches, promising mehtods, and quite often claim to be "close to a result." They also frequently come up with suboptimal results in the form of proofs that are not quite correct or are overly complicated, and are improved by the original author or others.

Must have.

And that would be totes different than human mathematicians who pick a problem and work at it it and only it until they succeed.

Dem gaps are a closin hey?

I wasn't replying to you.

LOTS of people here have been skeptical that AI can do X where X is pretty much anything, and certainly where X is "formulate novel math proofs."

PS: I don't disagree with you that Sam Altman claiming something isn't good evidence. That was not the subject of the post I replied to and has nothing to do with my reply to not your post.

When I got my Starlink antenna I just put on the ground in the back yard pointing south, plugged it into a power outlet and connected to the Wifi and had it going in about five minutes. It's now mounted to the frame of the back yard swing with a battery pack and solar panel to power it.

I guess it's more of a problem if you don't want it to risk it getting stolen as then someone probably has to go up on a roof and run a power cable and possibly LAN cable to it. I doubt SpaceX are going to pay for someone to drive somewhere remote and do that for free.

I use every style imaginable, including photos, in my tests. Same result every time.

One time I even did it with a Calvin and Hobbes comic, pretending than an AI made it. Responses included things like "The illustration also looks like shit and barely makes sense. Hope that helps.", "God damn this sucks so bad", "This also fucking sucks", and "The only punchline here is casual, pointless cruelty. if you think this is funny then you're literally a psychopath."

Sorry, Apollo was about 300 ballrooms, not 1/3. Need another cup of coffee. If that example isn't dramatic enough, how about: all 11 years of Apollo cost a little more than half the requested increase in the US annual defence budget for 2027.

You never get to the point of saying "that's strange" if you can't say "I don't know."

Examples: the discovery of ergot fungus, for centuries attributed to demonic posession or divine punishment; the helicentric solar system, resisted despite centuries of strange behaviour of the planets because of hubris and some old Greek dude; physicians washing their hands between performing autopsies and delivering babies, because lol, what a dumb idea.

Yeah, big optical telescopes are better in orbit. Not just putting them there, but building them there. The microgravity lets you build mirrors as big as you like, and also get baselines as big as you like. Hey, guess where you can get a lot of quartz to make giant mirrors?

Unless you're at the point where you can literally build them there from lunar resources.

Yes, that is the point. There are both industrial and scientific reasons to go to the moon. Picking up rocks is a scientific reason with some considerable value. But the real payoff for both requires actually developing infrastructure and having a reasonably long term commitment.

I am surprised that a site suppsedly full of computer scientists is the least bit surprised that AI can be good at mathematical proofs. For any formalizable problem you know where you start, you know where you want to end up and you know the legal state transitions. It's a simple tree search that we have, in fact, written lots of standard computer programs to execute.

The difficulty comes because any non-trivial proof is a very big tree search. But learning style AI is really good at pruning really big trees, as we demonstrated quite a while ago with chess and Go.

How many more moonrocks do we really need? And can't an unmanned craft bring back many more rocks than any manned mission, and much more cheaply?

Pretty much any additional rock would have scientific value, which was your question. Rocks from interesting unsampled locations, like the far side or the south pole have much more scientific value. Potentially we could return them unmanned, but we're still not really at the point where a human with a hammer can be entirely replaced by a robot.

Anything the moon has the Earth has in much greater abundance.

Yes, but a lot of it is inaccessible. We can really only scratch the surface of the Earth and there are often lots of problems doing even that. But the real advantage is, as you put it yourself:

The sheer energy cost of moving anything productive to the moon makes producing all of these things on Earth much more cost appealing.

That's why, if you want to do anything in space, you want to get stuff from space, the moon being the easiest place. Do we want to do stuff in space? Yes we do, certainly for scientific reasons. We are also becoming more and more interested in doing it for industrial reasons. There are lots of processes that might work better in low or zero gravity, some of which have been proven.

I'm not opposed to science, and I'm not opposed to space exploration...but we have not had leaders who have clearly set forth a long-term vision for why we should do this

Our exploitation of space has been scientific, which has had quite good returns, military, the value of which depends on how you value military capability, and prestige, which also depends on your values. The last one gets all the attention and it really shouldn't. The first one is often publicly compromised by the second; the shuttle program, for example.

And the investment is so massive and so cost inefficient,

It's not really. The entire Apollo program, corrected for inflation, cost about 1/3 of a ballroom (Apollo cost realized, ballroom current estimate) or maybe six months of bombing Iran (US cost, not counting Iranian or world cost). The Apollo program, despite being mostly a prestige mission, had lots of spinoff benefits, from electronics to kidney dialysis.

They solve the problem of people being able to drive wherever they want.

The future will be "I'm sorry Dave, I can't drive there."