traject_

If you mean > What is missing is an intelligible proof that human mathematicians can understand and use to advance the aims of mathematics.

Then that is not necessarily subjective either if an AI can produce an actually intelligible proof. The problem is that as mathematicians with PDE expertise have mentioned on Twitter the actual solution seems to devolve into an unreadable mess focusing on irrelevant details after a more readable first few pages in the proof. If it wasn't a Lean compiled proof and presented as a human artifact, it would be hard to assess if the deviser of the solution had any actual understanding of the solution.

I don't think that's particularly true. The whole point of the Millennium Prize problems as the article states was not on absolute difficulty of the problems but on the high chance of a proof producing fruitful results leading to new concepts and theories. Any pursuit of capital T truth will of course aim for better and more clarifying abstractions and is not a subjective turn by any means.

The contention is usually that empathetic policies that favour outgroups can be exploited by that outgroup to benefit themselves. For example, a universal policy of hiring from all groups can be subverted by allowing in a group that hires only their own members.

But that belies the real contention of what exactly is the truth? What is even more fundamental that has been lost is trust in nation and institutions due to a growing values divide. We must first trust before we can accept in good faith. Perhaps in some sense we should be thankful that we now have societies robust enough that even if major institutions lose legitimacy with half the population we don't get a civil war.

It should also be noted another paper also came out today with new ancient DNA from a modern human whose bones carbon dated to around 45kya. The data suggests the individual was only 80 generations from the Neanderthal interbreeding event suggesting it happened around 45-49 kya. The event may have happened over a few generations realistically but I think 7000 years is unlikely due to the rapid expansion of humans out of Africa (or to be more precise an extremely successful expansion of a population of modern humans whose signal is difficult to filter out) all sharing this same introgression signature.

The timeline overlaps with the Peace of God movement and the Gregorian reforms which legal historian Harold Berman calls the Papal Revolution in his book Law and Revolution: The Formation of the Western Legal Tradition. Also coincides with the beginnings of renavatio that Charles Taylor mentions in A Secular Age. You could argue though that perhaps it took some time until the population fully Christianized; it is precisely in this time period we start seeing the use of familiar Biblical along with a set of standardized Germanic first names all across Europe while old Germanic naming conventions start to disappear.

> And within those pieces of fiction they will generally refer to their headcanon, but it's no more special than the fans'.

If it's the same author writing these pieces of fiction, then speaking by definition, the author's opinion is more special by being the creator of those works and therefore can create fiction that reinforces their headcanon (which is why it's called canon). So I think calling the author's opinion less special is wrong for the author's tie to the work will always be more special than the consumer by virtue of being the creator.

> And the thing is Windows has not enforced such direction, or even pushed hard for it AFIK, and that is something you can very well blame then for.

Even here it's pretty hard to blame them due to antitrust concerns. Just google the word Patchguard.

Yeah, a lot of these articles conflate use of zero as placeholder, numeral and number. But the real critical conceptual step is the last step of using zero like any other number (mostly) in arithmetic.

> It's almost as if he's being overly negative about them to make a point or something to push back against all the unbridled optimism.

I don't think it is like that but rather Chollet wants to see stronger neuroplasticity in these models. I think there is a divide between the effectiveness of existing AI models versus their ability to be autonomous, robust and consistently learn from unanticipated problems.

My guess is Chollet wants to see something more similar to biological organisms especially mammals or birds in their level of autonomous nature. I think people underestimate the degree of novel problems birds and mammals alone face in just simply navigating their environment and it is the comparison here that LLMs, for now at least, seem lacking.

So when he says LLMs are not sentient, he's asking to consider the novel problems animals let alone humans have to face in navigating their environment. This is especially apparent in young children but declines as we age and gain experience/lose a sense of novelty.

We don't actually know if it is SOTA, the previous SOTA solution also got around the same on the evaluation set.

Yeah, this is a point a lot of people simply don't understand. Having billions of dollars in GDP generated by the service industry (like financial services or informational technologies) does not map one to one to generating a functioning arms industry to produce artillery ammunition for example. You need manufacturing facilities, a large pool of candidates with potential expertise in technical hardware skills to run these factories and logistical lines to keep them running. These prerequisites existed in the West during the earlier part of the 20th century which was why the transition to the war economy was relatively painless but no longer exists now. It is simply irrelevant to talk about multibillion dollar GDP economies specialized in unrelated industries if you don't have the actual physical resource and staffing requirements.

> Not really. Bonobos share roughly the same amount of DNA with humans as do Chimps. And Bonobo societies are matriarchal and mostly lack the group warfare dynamic.

Unless you mean group warfare as in terms of internal warfare; this is incorrect. Bonobos are noted to be peaceful within their groups but are similar to chimpanzees in being observed engaging in conflict with other Bonobo groups to a more limited extent.

One man's altruist is another man's fanatic. I, for one, would prefer an evil bandit over an evil fanatic because at least a bandit sleeps once in a while as that C.S Lewis quote goes.

A lot of that billions of capital is simply Azure compute credits though.

They have the GPUs and there's a GPU shortage already. What do you think is going to happen when they lose that access in $10bn compute? Given the shortage and other vendors being mostly locked up already, there's not much OpenAI can do. And that's not considering that the board is facing immense pressure internally starting from the very temporary CEO that they put in place two days ago to various members of senior leadership threatening to resign (with many employees being paid in options dependent on valuations that the board just destroyed). The board may have de jure power but that is nothing without the de facto power associated with it.

> But it boggles my mind how people can assume, with a straight face, that they are equipped to educate their child alone - something which is normally a profession for which you have to study O(years) (and even then most people aren’t really good at).

The key assumption here is that the expertise and years of studying translates to desirable outcomes. The large expansion of homeschooling suggesting by the article suggests a great rise in people who no longer believe that the expertise promoted by teaching schools is actually relevant to teach their children.

After all, mass public schooling is only about a century old and default human experience in many ways has always been closer to what we call homeschooling.

There were actually metaphysical materialists similar to Machiavelli in Newton's time who were paragons of something similar to atheist thought and were free of " this cancer on humanity". But it's telling that none of them ended up developing anything of worth in the natural sciences and anyone who knows their philosophy would know why that's the case.

Why is "reactionary" inherently wrong or "progressive" inherently correct? Why can't people use more clear language rather than relying on Whig history to make their points. If what they are saying is morally wrong according to the truth (and that might be so), clearly state why that is the case rather than obscuring with terms like "progressive", "social justice" or "reactionary".

“We are fond of talking about 'liberty'; but the way we end up actually talking of it is an attempt to avoid discussing what is 'good.' We are fond of talking about 'progress'; that is a dodge to avoid discussing what is good. We are fond of talking about 'education'; that is a dodge to avoid discussing what is good." - G.K. Chesterton

There's no reason to believe that LLMs are necessarily similar to humans. Just think of flight; both birds and planes can fly despite being very different. It's understandable that because humans are our main reference for intelligence that we see LLMs speaking language as a proof of similarity but there's no concrete reason to assume that.

That could mean it is a necessary but not sufficient condition; I don't think your observation by itself invalidates anything the author said.

I mean he implicitly threatened to hack Bing; I don't know how that's not an aggressive action. A LLM's response is just a reflection of the user writing the prompt and so acting with outrage is perfectly described as "botsturbation" as mentioned by another commentator. It's like being shocked searching Google for sexual or violent things will return such things; it's a product in beta and will likely be patched out which is already happening. An LLM can't do anything anyway so I really don't get the whole feigned outrage here.

I don't know; it seems there is a clear learned aversion to the word "stealing" but doesn't change the unethical nature of the crime is equivalent to stealing royalties deserved for the consumed work. I pirate some times sure but I do so with the understanding that what I am doing is unethical and try to avoid it.

Consuming media/entertainment is no human right and if it is too expensive/too inaccessible/whatever and you wish to be ethical, don't pirate it in the first place.

Humans have great difficulty controlling their impulses especially in connection to crimes that are undetectable and easy to perform but the honest will at least own up to what they do.

>And I do know that video-games have been available for a few decades now, but the thing is that they've become quite compartmentalised from society's point view. More exactly if you like video games or if you're into video games most probably that means that you're the kind of person that is spending tens of hours per month (week?) hooked to a PC or a game-console, ignoring the outside world, hence the outside society. Culture, by definition, was meant to provide some "glue" to society, to embrace it, so to speak, today's video-games (or most of them, anyway) run counter to all that.

Culture has various definitions that it can be interpreted by but by your comments, your views on culture appear to be that relevant to the youth of the mid to late 20th century but are much less relevant to the youth of today. The culture around video games is more dynamic and popular compared to that of cinema which has serious issues drawing eyeballs outside of the repetitive blockbuster films. There's a reason why the CEO of Netflix was more concerned about the threat from video games to their product.

> To give you an idea of how sophisticated Ontario is, most Hollywood comedians and a large percentage of writers are originally from the Toronto, ON area because of their writing and self-production abilities.

Haha, that's just because Ontario has a large population with respect to Canada as a whole. Sophistication is not the right word to use as someone from Ontario.

> Seeing how African populations are extremely diverse and we're just seeing the extent of it I would refrain from making such definitive statements.

There's no reason to not. Holocene expansion of farming/pastoral populations all over the world (including Africa) has largely homogenized human ancestry into identifiable distinct strands that is in varying proportions. Africa is indeed diverse genetically but it is not magic or anything. It most likely comes from the earlier mentioned multi-regional model within Africa through drift and admixture with highly drifted populations. And almost all of these populations outside of those of North Africa were outside of the Neanderthal range and we have archeological evidence to support this as well. There is no logical reason to believe in an (outside of the obvious example of historic West Eurasian admixture in the Horn/North Africa) African population with non-trace levels of Neanderthal ancestry. High levels of diversity in Africa does not imply a significantly large population with non-trace levels of neanderthal ancestry.

> The science is constantly evolving to this very day on the supposedly 'trace levels' of Neanderthal DNA in Africans (especially as we gather a more diverse range of cohorts) so I'll leave it at that.

No, the trace levels of Neanderthal DNA in Africans is very unlikely to change by gathering more diverse range of cohorts. It is a matter of identifying major strands of ancestry within Africans (by admixture over thousands of years to now be in various African populations) which all are distinguished by a lack of Neanderthal DNA outside of West Eurasian admixture.

And for the last point, I did not condone scientific racism. To repeat, interest in human populations and such phenotypic differences does not imply scientific racism once you realize the basic scientific principle that humans are animals and consider how animals exist in populations with phenotypic differences.

>Africans do have Neanderthal DNA, up to 0.3%.

That ancestry comes from later interactions with West Eurasians and is at trace levels compared to the substantial 2-3% in non-Africans. This does not change the point that non-Africans received input from Neanderthals just before expansion outwards.

>The post tries very hard to make it look like 'Out of Africa' is wrong and not the mainstream accepted by the majority of scientists. Admixture doesn't change that.

You've misread the article if you believe that. The point is that the total replacement model of out of Africa imagining a small band of hunter gatherers expanding out of an East Africa giving rise to all of humanity popular in the 2000s was proven wrong. The point was that single locus markers like mtDNA and Y-DNA can create biases that allowed for such a consensus that was changed by the whole genome of the Neanderthal. The ancient DNA we have now suggests a multi-regional model for modern human evolution within Africa.

The remainder of the post (other than the first nitpick) is non-substantive ideological claims that appears to be largely because Razib's politics lean conservative. Interest in human populations and such phenotypic differences does not imply scientific racism once you realize the basic scientific principle that humans are animals and consider how animals exist in populations with phenotypic differences.

> Systems level thinking is unpopular.

I think it is the opposite really. Materialistic thinking is the baseline in modern society and in many ways the tendency to dismiss the unquantifiable as non-existent is one of the greatest weaknesses of modernity. It is certainly far easier to quantify the average temperature in a region using Antarctic or Greenland ice core data and use that to determine how that may affect agricultural production in a historic time period or analyze average skeleton height sizes to assess quality of life metrics than to analyze how cultural changes affected problems.

>At the root of this exists a fallacy of composition where the argument is that if each individual is virtuous, then the whole[system] is virtuous

The Romans may not have worded it correctly but it should not be surprising if we replace the word 'virtue' with 'social norms' and re-analyze how changes in social norms can affect social trust, how institutions are structured and even how families are structured(e.g. a change in norms from the religious change from paganism to Christianity meant a change from the city focused philanthropy of euergetism to the poor/sick focused ecclesiastical charity).

Social trust, for example, is very important as it greatly reduces transaction costs yet and its core much of it relies upon the concept of common shared knowledge (in the form of traditions, values and principles) that is difficult to quantify. Yet at its core, the change in social norms would be deeply related to changes in 'virtues' as virtues, in practice, are ideals that society pressures individuals to follow. So, the Romans may not have the scholarly data on material changes in the environment to be sure but they also have a better intuitive understanding of the importance of virtues than their modern successors.