Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
6gvONxR4sf7o
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
7 ms
·
61.
▲
by
6gvONxR4sf7o
2y ago
I use a similar notation, but never quite found a satisfactory notation for elementwise operations (e.g. `-M-a + -b`, especially broadcasted ones which I end up doing as `-A-B- + -c 1-`) or for denoting what derivatives are with respect to.
62.
▲
by
6gvONxR4sf7o
2y ago
That's the kind of thing that I could imagine could dramatically speed up certain tasks, but not enable particularly new abilities. A ton of the challenge is converting a sequence into a graph. So if you need a huge clever model to tur
63.
▲
by
6gvONxR4sf7o
2y ago
Human performance is much closer to 100% on this, depending on your human. It's easy to miss the dot in the corner of the headline graph in TFA that says "STEM grad."
64.
▲
by
6gvONxR4sf7o
2y ago
I've always disliked this argument. A person can do something well without devising a general solution to the thing. Devising a general solution to the thing is a step we're talking all the time with all sorts of things, but it do
65.
▲
by
6gvONxR4sf7o
2y ago
I'm glad these stats show a better estimate of human ability than just the average mturker. The graph here has the average mturker performance as well as a STEM grad measurement. Stuff like that is why we're always feeling weird t
66.
▲
by
6gvONxR4sf7o
2y ago
It's a shame because in 3D you can do much much cleaner layout in general.
67.
▲
by
6gvONxR4sf7o
2y ago
> so high performance "python" is actually going to always rely on restricted subsets of the language that don't actually match language's "real" semantics. I don't even understand what this means. If I
68.
▲
by
6gvONxR4sf7o
2y ago
The frustrating thing about prolog is that at first it seems you get to ignore the execution model, but very very quickly you have to learn to think of it in imperative terms anyhow, so you can ensure termination and stuff. Or at least, I n
69.
▲
by
6gvONxR4sf7o
2y ago
I'm late to the party, but also wondering about Metal, and I have a question for you all. Do you happen to know how energy use relates to utilization? If I run a super duper duper fast thunder kitten kernel on an iphone (pending Metal,
70.
▲
by
6gvONxR4sf7o
2y ago
It seems like in your first part, you're saying that proofs are the same as their normalized proofs, up to some rewriting system. So like how we say 3-2 is the same as 1, basically, or (more interestingly) saying that x-x is the same a
71.
▲
by
6gvONxR4sf7o
2y ago
Why use RL for this instead of plain old supervised learning?
72.
▲
by
6gvONxR4sf7o
2y ago
In honor of election season, I hope I can use the 'priority messages' stuff to better filter the political spam I get.
73.
▲
by
6gvONxR4sf7o
2y ago
I think this is where it's worth differentiating between different types of "believes in" (and why I think modal logics are cool). I can convince myself that a thing seems safe to believe, or I can tangibly believe it, or I
74.
▲
by
6gvONxR4sf7o
2y ago
The only AI thing that excites me on my phone is a battery-friendly way for app developers to run ML models. I don't want an AI assistant that sees everything that i do, but some things like computer vision and translation and OCR are
75.
▲
by
6gvONxR4sf7o
2y ago
They jump up in price pretty quickly as size goes up. The cheapest 5+ inch display I found at your link was over $40, and it's about 100PPI. It's certainly not prohibitive, but certainly priced high compared to "just jailbrea
76.
▲
by
6gvONxR4sf7o
2y ago
I think more of the issue might be the eink screens. As far as I can tell, there just aren't 5+ inch eink screens for cheap.
77.
▲
by
6gvONxR4sf7o
2y ago
Even the insurance side isn't really a free market. I've only ever gotten like two choices of insurance provider at any job I've had. More than two plans, but very limited provider choices. You can't get a job offer and
78.
▲
by
6gvONxR4sf7o
2y ago
> Server Side code running is mostly a solved problem I know what you mean here, but I think we're very limited in what we tend to run. Polyglot programming still isn't really a thing, and with things like WASI standardized (so
79.
▲
by
6gvONxR4sf7o
2y ago
Any time the message is "We should do X because if we don't, the bogeyman will do it," it should be a red flag. Seems to be one of the common themes in the article. It's always admitting that the thing you want to do is
80.
▲
by
6gvONxR4sf7o
2y ago
For a person just curious about all three, how do the ecosystems compare?
81.
▲
by
6gvONxR4sf7o
2y ago
Less about risk parallels and more about control parallels. The last 10% of fine grained control over a system is hard. Like every time I’ve done text to image prompting and it gives me a great starting point, but cannot get certain details
82.
▲
by
6gvONxR4sf7o
2y ago
Heh, I should have been clearer. I meant citation needed on the other sentence.
83.
▲
by
6gvONxR4sf7o
2y ago
> Google Glass originally failed due in part to public backlash at being recorded without consent in public spaces. However, it’s also true that in the decade since, people have become more accustomed to being filmed due to the rise of s
84.
▲
by
6gvONxR4sf7o
2y ago
Calculators are reliable and predictable, so losing skill at that kind of calculation is a safe, compartmentalized offloading. We offload an extremely clearly defined set of tasks, and it gets executed cheaply, immediately, and perfectly. L
85.
▲
by
6gvONxR4sf7o
2y ago
> A human still needs to consume their output and act on it intelligently. We already do this, except with other tools/mechanisms (i.e. other humans). Nothing really changes here... What changes is the educational history of those h
86.
▲
by
6gvONxR4sf7o
2y ago
I wouldn't call those the same. My wife's car has a camera and mine doesn't. Since having children, I prefer backing up in her car because I'm so much more aware of the blind spots I have without it. We also live somewhe
87.
▲
by
6gvONxR4sf7o
2y ago
> The best educators will adapt to AI. In some ways, the changes will be positive. Teachers must move away from mechanical activities or assigning simple summaries. So much of education always comes down to efficiency, scale, and stand
88.
▲
by
6gvONxR4sf7o
2y ago
So making libgen is illegal, but using it to train LLMs is legal? I know there's a whole issue of transitive liability (maybe you couldn't know you were getting an illegal thing from the thief, so it doesn't always make sense
89.
▲
by
6gvONxR4sf7o
2y ago
We didn't need to take people's music to build a record player, and when we printed records, we paid the artists for it. So yeah it had a profound effect, but we got consent for the parts that fundamentally relied on other people.
90.
▲
by
6gvONxR4sf7o
2y ago
Not just that. It's "the ends might justify the means if this path turns out to be the right one." I remember reading the same thing each time a self driving car company killed someone. "We need this hacky dangerous wa
More ›