Also, that study is extremely dubious/barring on fan-fiction.
Homer (if he was even a person) recited oral epic poetry (that we've lost most of) and we've written down an extremely lossy version of it.
Why should we take any of the mythical character's interactions with mythical gods as any insight for the brains of real humans???
It'd be like year 4000 historians citing World of Warcraft claiming the Lich King is a very obvious example that humans lack an internal monologue because he was not able to distinguish his own motives/actions between Ner'zhul/Arthas (both spiritual entities inhabiting the "Lich King" body).
I mean I definitely disagree with the book's argument, and you're right that it's making a different argument than the article, but I think straight up declaring it irrelevant is coming in way too hot. It's also an argument for a merging of separate mental structures that's sourcing the change in an event from recent natural history.
Again, I don't agree with the book's argument, but declaring an argument proposing a similar convergence on the basis of entirely different reasons to be not even mentionable in response to the article is just getting way too overzealous with comment section policing.
"too hot" and "way too overzealous" is ad hominem, and talking about comment section policing is remarkably hypocritical -- you're actually telling someone not to make certain sorts of comments, which pointing out that something is not relevant to the article doesn't do. Your "not even mentionable" is a complete fabrication -- the comment from dinobones said nothing of the sort. Your whole shtick about "getting way too overzealous with comment section policing" is a complete fabrication -- a projection of what you are doing onto a simple factual statement. And your "Uh, yes" comment below is more obnoxious trolling and attacking strawmen.
My immediate reaction to the comment about Jaynes's book and the split brain was that it's not relevant to the article, and it conflates three very different things. The comment omits the crucial fact that the split brain research was on people whose corpus collosum had been cut -- no surprise that "these two parts of the brain function independently" since they were no longer attached to each other (which is the complete opposite of what TFA is about). Jaynes's hypothesis was not related in any way to that research, and neither are related in any way to TFA. As someone else noted: "Okay but this article is about evolution hundreds of millions of years ago." -- the clear implication being that Jaynes and split brains aren't relevant. But if anyone actually dares say so, here comes the too hot overzealous comment section policeman to project his own attributes onto them.
And this is completely aside from whether Jaynes's argument is valid ... you saying multiple times that you disagree with it is also utterly irrelevant and reflects one of several fundamental errors in logic in your comments. But criticisms of your comments is not "comment section policing" -- how absurd ... and dinobones didn't even do that in simply saying that a comment wasn't relevant to the article ... only in your head was "and therefore you are not allowed to post it here". Presumably the comment was made in good faith. I can even grant the same about yours, with a bit of straining and extraordinary charity.
The premise (and this is how the GP comment put it) was that humans around Homer's time had fundamentally different brain structures, and that differences in poetry and thinking processes could be explained by looking at how brains changed over the last few thousand years.
That's not "overzealous comment section policing". It's rejecting a premise that is completely infeasible with everything we know about human evolution (or relative lack thereof) during that timespan.
What could plausibly cause that change in just a few thousand years?
I cannot see how this could be true without it following that not everyone will have this change as it is far too recent to have spread to everyone every where. Where are these people who lack it?
My read was that they were also disagreeing with it (and I personally disagree with it too for exactly the reasons you mention).
But the problem in this case isn't just that it's right or wrong, it's that it at least belongs to a related family of arguments that are fair game to explore in response to the article. Some of the best conversations take the form of knowing something is wrong but powering through the struggle of putting the why into words, and all the associated conversational branching that can happen along the way.
Uh, yes. It's overzealous because personally disagreeing with something doesn't mean that it's outside the scope of the subject matter that the article is speaking to. If anything, you're reaffirming the commonalities between the theory and what's proposed in the article because what they have in common is a proposed difference in brain structures followed by a convergence.
The fact that you disagree with the truth of the theory has literally nothing to do with whether it's on point in response to an article that's about the same subject matter. I would have hoped that me saying multiple times that I disagreed with it would have made clear already that there's more going on here than just a matter of whether you do or don't disagree with it. Declaring that it's "not overzealous" mirrors my language back at me but without attaching an actual argument to it.
Is this really that big of a flaw in this implementation? I don't think it's worth the additional complexity to address it. (Encoding depth information in some way, trying to detect "flat" surfaces, whatever).
Discerning a camera taken image of an image is typically very very easy. The collors/exposure/etc will all be obviously wrong in ways to a human, even without doing any analysis.
You mean that it is sometimes very easy. But it is also sometimes impossible. You seem to be thinking only of poor quality photos of poor quality prints, but there's no basis for assuming those characteristics.
The model can't reason comprehensively (e.g., like Sol XHigh would to solve a complicated problem), but it's designed to be able to answer anything a human reasonably could quickly and intuitively, i.e., system one thinking: https://en.wikipedia.org/wiki/Thinking,_Fast_and_Slow
You play White, and Jev plays Black. Rather than asking an LLM to generate move strings or JSON, the backend feeds all server-validated legal candidate moves into Vercel AI SDK's experimental_evaluate(). Jev picks Black's move and outputs its probability distribution across all legal candidates in a single forward pass (~300ms, ~$0.00004/move).
Struggled to understand the use case till after i came across this https://eu.36kr.com/en/p/3991165197188099 article, which referred to them as the "referee for agents."
I had a different initial confusion - it seems this company has no relation to the company formerly known as Typesafe https://en.wikipedia.org/wiki/Akka.io
I definitely agree it's underexplained in type safe.ai's materials.
I have to assume it's a reference to the fast, heuristic, intuitive "system 1" process in humans, as opposed to the slow, procedural, reasoning "system 2".
This theory is recognized, among others, in Daniel Kahneman 2002 Nobel prize on Economics.
It can't hallucinate, but it doesn't mean it can't make wrong decisions. Just because it adheres to a specific output format at all time, while LLMs have the output format at their mercy, then the claim of not hallucinating is made technically true.
I think that this specific part is not super interesting if your harness just recovers from invalid LLM outputs.
The latency and cost - yes, those are super interesting.
AI researchers: Yeah our AI models are a bioweapon threat and a nation state level cybersecurity threat and also a genius math researcher and very dAnGeRoUs!
The world: Cool, can you rewrite this email with a professional tone.
The places where people want to live have no land or archaic zoning laws.
I wouldn't want to live two hours away, but if I could sleep/work during my commute, and pay $400/month for it, I might do it. My housing costs would be like less than a third of living in the metro where I work.
Ive never liked this framing; "the public good" is too nebulous an idea to justify how tax should be spent.
I, and people like me create wealth. Some of that wealth is pooled at an administrative level that makes sense across a town, state, or nation. Its still wealth I created; and I expect it to be used to benefit me, ideally proportionally to the tax taken.
Unemployment can be justified as crime prevention; Disability and healthcare systems limit individual worst-case outcomes (especially if due to chance). Super-commuter subsidies do what?
these things will be evaluated based on the net benefit such a service would provide. more people having affordable commute to dallas would end up increasing the net benefit and value that dallas has.
cities grow in capabilities in direct proportion to the number of people who have access to it and their capabilities. it might be cheaper labour leading to better roads, plumbing etc
This is like the “why not use FTP yourself” response to the YC Dropbox page.
Yes, but good, standardized developer tooling is valuable.
Postman could’ve also been a “just use CURL and dump to JSON” for example.
I actually am bullish on this. Tmux and ssh are somewhat clunky and probably under utilized by most people. I struggle to find commands, navigate sessions, see history, and so on.
Also in an “agent” based world where you might ask an agent to execute things for you, how do you manage that context well beyond just giving it its own Tmux session to drive?
It’s a good idea, and I’m surprised no one’s built it sooner.
Devs don't use Dropbox, they use git. Git is probably the poster-child example of an unfriendly interface, and yet it's the standard way for devs to share files.
Emacs is downright friendly compared to git & tmux IMO.
Just start pricing in bytes input/output. This whole "token" and "tokenizer" thing is an implementation detail that shouldn't even be leaking out into the API.
Providers change tokenizers all the time with model updates, and it's often not even possible to query/figure out how text is tokenized without actually just sending the LLM a request.
Just switch to charging for bytes of intelligence. Please. Claude Shannon figured this out decades ago.
Also, that study is extremely dubious/barring on fan-fiction.
Homer (if he was even a person) recited oral epic poetry (that we've lost most of) and we've written down an extremely lossy version of it.
Why should we take any of the mythical character's interactions with mythical gods as any insight for the brains of real humans???
It'd be like year 4000 historians citing World of Warcraft claiming the Lich King is a very obvious example that humans lack an internal monologue because he was not able to distinguish his own motives/actions between Ner'zhul/Arthas (both spiritual entities inhabiting the "Lich King" body).
reply