Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
naasking
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
35 ms
·
661.
▲
by
naasking
1y ago
> So, the point of the entire article is that he didn’t like the title of paper he’s criticizing? I think the article makes a good point when it says that nobody else he spoke to about it had noticed this qualifier either. There is a lon
662.
▲
by
naasking
1y ago
The point has nothing to do with speed, but with expressive power / what is achievable and learnable, in principle. Again, if you accept that a computer can in principle run a program that qualifies as AGI, then all I'm saying is
663.
▲
by
naasking
1y ago
> Powerful pattern matching is still just pattern matching. Higher order pattern matching is Turing complete. Transformers are Turing complete. Memory augmented LLMs are Turing complete. Neural networks can learn to reproduce any functio
664.
▲
by
naasking
1y ago
Depends how much tinkering you want to do. Out of the box, Prusa and Bambu are king for a higher price point. More of a closed ecosystem though, the remaining options are more open. If you want to assemble your own premium machine, Voron fo
665.
▲
by
naasking
1y ago
Programming languages with type erasure would have no runtime, just raw program code and the WASM GC. Languages that have runtime types still need a runtime for that functionality.
666.
▲
by
naasking
1y ago
Did someone in this thread claim that?
667.
▲
by
naasking
1y ago
The difference between ML models and computers since the 1960s is that the ML models weren't programmed with predicates , they " learned " them from analyzing data, and can continue to learn in various ways from further data
668.
▲
by
naasking
1y ago
> but rather the ability to reason in the general case, which requires the ability to LEARN to solve novel problems, which is what is missing from LLMs. I don't think it's missing, zero shot prompting is quite successful in man
669.
▲
by
naasking
1y ago
I agree LLMs have many differences in abilities relative to humans. I'm not sure what this implies for their ability to reason though. I'm not even sure what examples about their bad reasoning can prove about the presence or absen
670.
▲
by
naasking
1y ago
> It's trivial to demonstrate that LLMs are pattern matching rather than reasoning. Again, this is just asserting the premise that reasoning cannot include pattern matching, but this has never been justified. What is your definition
671.
▲
by
naasking
1y ago
> are they really learning to reason, or are they just learning to pattern match to steer generation in the direction of problem-specific reasoning steps that they had been trained on? Are you sure there's a real difference? Do you
672.
▲
by
naasking
1y ago
This is a product of programming languages continuing to ignore the lessons from capability security. The fact that packages in your programming language even have the ability to any of the things listed in these articles by default is an e
673.
▲
by
naasking
1y ago
> It's a very helpful way of saying it shouldn't be bothered to be read. Not really.
674.
▲
by
naasking
1y ago
Helpful hints are all you can rely on from client-side. Client-side validation doesn't really exist, true validation only happens server-side.
675.
▲
by
naasking
1y ago
> it divides something that would be otherwise a monolith into subsequent small calls This is often at odds with page responsiveness, and can increase server load considerably.
676.
▲
by
naasking
1y ago
> That just not a metric that matters. So you said that fuel costs dominate the cost of energy delivered, I'm saying that you can deliver the same energy while purchasing less fuel , but now you're saying that that's not
677.
▲
by
naasking
1y ago
> Total death count is a straw man argument, what matters here is the economic costs. Paying out lawsuits is an economic cost. Regardless, disposing of low level radioactive components of the reactor had to happen at some point, and the
678.
▲
by
naasking
1y ago
> cleanup of vast quantities of nuclear waste The total high level, dangerous nuclear waste of the entire world since we started playing with nuclear power 70 years ago fits in an American football stadium with plenty of room to spare.
679.
▲
by
naasking
1y ago
> It’s really not, nuclear inherently requires extreme costs to operate. Compare costs vs coal which isn’t cost competitive these days. Nuclear inherently need a lot more effort refining fuel as you can’t just dig a shovel full of ore an
680.
▲
by
naasking
1y ago
He's named because his rebuttals are weak and Hossenfelder actually makes legitimate points that "Professor Dave" just evades.
681.
▲
by
naasking
1y ago
> The collapse isn’t in the fields themselves, it’s in the layer of journalism that used to connect experts to the public. Sure, but what do you do if the subject matter is the question of whether the experts are being myopic and the sys
682.
▲
by
naasking
1y ago
Judging a perfect solution requires first asking, whose problem are you solving? If you only ever insist on addressing your own aesthetics and properties of interest, you're only solving your own problems, not a problem anybody else ha
683.
▲
by
naasking
1y ago
And impure engineers would say that that pure engineering produces too much bankruptcy because you never ship a viable product. ;-)
684.
▲
by
naasking
1y ago
Maybe ironically, the idea that curves don't exist hasn't gone away, it's still a serious idea: spacetime could be discrete, in which case curves really are a series of discrete line segments. Loop quantum gravity is one such
685.
▲
by
naasking
1y ago
The reason purity is good is because you've distilled something to its simplest, most essential and parsimonious form. It is only in this form that you can truly understand something, and only from this form do you have maximum flexibi
686.
▲
by
naasking
1y ago
> I think there's something healthy to the boredom the kids describe, which ultimately leads to socialization and introspection. This. People these days talk about boredom like it's the worst thing ever.
687.
▲
by
naasking
1y ago
Windows 10 keeps telling me my very capable computers can't run Windows 11, so why would I upgrade? I suspect this is fairly common.
688.
▲
by
naasking
1y ago
They need it operationally for calculations and such, but those calculations don't necessarily need to be statically validated beyond simple things like type and unit checking.
689.
▲
by
naasking
1y ago
If you can't fail students and hold them back, poor students will continue and pull down the average of later grades. News at 11.
690.
▲
by
naasking
1y ago
> The point is that you always have to think about merging behaviour for every piece of state. CRDTs can't eliminate the requirement to think about what the consistent states are.
More ›