Far outside my expertise, but it feels like the shaky assumption here is that the understanding and proof need to come in a specific order to be valuable. Can't we get all the meaty goodness by simplifying and generalizing the proofs now we know they exist? Sure, some insights will live in the discovery itself, and I guess it's nice to be the person who got to a proof first, but the real work (according to the mathematicians, afaict) is in the understanding and processing. In this way, it feels like it's moving towards being like most other science: mostly understanding things that are already there.
> Can't we get all the meaty goodness by simplifying and generalizing the proofs now we know they exist?
Perhaps, but there is a danger: it is difficult to un-see things. At least for now, AI-generated proofs are likely to be of a somewhat brute-force nature. As each such proof pops into existence, two things happen: (1) it is much harder to maintain an untainted mind and go on to discover alternative, perhaps deeper and more conceptual, proofs that cannot be obtained by massaging and cleaning up a more brute-force approach (think of this as getting 'stuck in a local minimum', if you like), and (2) at a societal level there is thereafter much less incentive to attempt to do so (announcing a proof of a previous unproven result is, for now, far more prestigious than announcing a much better, more elegant, proof of a known result).
One could ask why we prefer conceptual explanations to brute-force proofs; after all, a proof is a proof, isn't it? I suppose one reasonable answer is that conceptual explanations lead more readily to new questions and that there's also intrinsic beauty in such explanations -- though, of course, many outside the field will simply not care.
A coin flip can make infinite decisions on all possible input taxonomies.
Zero-shot is indeed very cool. And it's super useful/useful for the developer masses that don't know, care, or work pressures don't allow, for proper evaluation and calibration.
But it would be good that we don't over-hype these things.
I ain't watching that video or clicking on Twitter, but from the comments, this is about "ingredient splitting", where they "split an undesirable ingredient into several varieties so that each appears further down the list", because in the US ingredients are ordered by weight. So you can "hide" the amount of sugar by claiming it's from several sources, and listing those sources distinctly.
> if this was an academic paper submitted to a journal, it should be issued a desk rejection for the quality
If Mr Tao had submitted a shitty, poorly-written proof of a famous outstanding problem, no journal would reject it. That extends to anyone with sufficient credibility. They might ask him to keep at it and fix it up, but nobody would begrudge him putting his shitty (but ultimately correct) draft of arXiv while he did so.
We're all getting disrupted, we all have feelings about it, but from the perspective of a software engineer who's been dealing with all of this for several years now, this post is just cope.
This is awesome. I would note that there are similar charities (based in various Western countries) that do similar things if people like the idea but for whatever reason don't want to donate to this particular one, just a simple Google search away.
Vetting them is a challenge though. I always prefer some sort of personal contact like it's available here. And for anyone who doesn't trust OP, I invite OP to visit me in Tanzania, then I can verify and vouch for his existence (and he can then verify mine :-)
Yeah, as I wrote in the other comment, I moved to Africa 4 years ago. My wife in China had passed away, and I felt that I could not sustain living there anymore. I was stuck in my wife's small town doing remote work that didn't pay enough to move to a big city where I would have more opportunities, and also my children, being Chinese citizens were not allowed to go to school just anywhere. It was also during covid so there wasn't even a chance to travel and look for opportunities. By the time covid was over I was ready to leave.
My options then were to go back home, or to try Africa and see if it is possible to live there for a while. I'll be staying here as long as I can financially sustain myself with online work or through sponsors. And even if I can't stay much longer, we have now been here long enough that Africa has made an impact in our lives, and we'll hopefully be able to use that to have a positive impact on others too.
I wonder if it makes conceptualization simpler for models too, given that they're trained already on human-speak. And I'm also curious as to whether humans currently have an innate advantage into simplifying and contextualizing proofs, or will the machines get good at that as well?
My point is that they will, and that is always going to be useful. Bundles of simplified knowledge (in whatever form) that make research simpler will be useful to share whether humans can understand them or not...
Yes, but why be the politician or civil servant responsible for dealing with it when it will most assuredly be someone else's problem in a few years? The political upside of getting it done is low, and the risk and cost are high.
reply