sdenton4

16 ms·

271.
by sdenton4 1y ago
Yeah, I wrote this in a bit too short a hand to meet the critics where they sit... There's an immense history of humans studying animal intelligence, which has tended pretty uniformly to find that animals are more intelligent than we p
272.
by sdenton4 1y ago
My favorite Samuel Delany story is about a woman in a village who invents writing, and teaches it to all the children. She makes a rule that you're never allowed to write down people's names, as it will inevitably lead to keeping
273.
by sdenton4 1y ago
It's an umwelt problem. Bats think we're idiots because we don't hear ultrasonic sound, and thus can't echolocate. And we call the LLMs idiots because they consume tokenized inputs, and don't have access to the raw
274.
by sdenton4 1y ago
I have been using one model to generate piles of random-adjective tables, then use real live randomness to select a bunch of traits to feed to the model doing character generation. It's instructed that it can take or leave individual t
275.
The Bittern Lesson for Bioacoustics(arxiv.org)
4 points by sdenton4 1y ago|0 comments
276.
by sdenton4 1y ago
Yeah, this is in 'flies like a plane, not like a bird' territory. But I think it's closer than you think. The systems do learn and have improved rapidly over the last year. Humans have two learning modes - short-term in-conte
277.
by sdenton4 1y ago
Huh, this post is not what I thought it would be! Even after the first two paragraphs! There's a line of thought which states that intelligence rhymes with compression: Identifying patterns allows better prediction, enables better comp
278.
by sdenton4 1y ago
I think getting out of the binary is good for the long run. We have something which is artificial, intelligent, and general in scope. We're there. Is it perfect? No. Is it even good? Sometimes! Do airplanes flap their wings? Also no, b
279.
by sdenton4 1y ago
"If it is still possible to put a goalpost somewhere - and we don't care where - then it's not AGI."
280.
by sdenton4 1y ago
Here's an alternative hypothesis... People thrive under regularity, and young people (especially) tend not to understand that. Similarly, being able to focus on a single thing is a kind of super-power, while multi-tasking generally hur
281.
by sdenton4 1y ago
Any button a human can touch is a vector for abuse...
282.
by sdenton4 1y ago
Oh yeah - the tracker itself wasn't the sampler, but one would often import sampled waveforms (or borrrrrow 'em from other tracks). I seem to remember fasttracker ][ (which i spent a ridiculous amount of time with as a teenager) h
283.
by sdenton4 1y ago
The tracker space is weird and full of twisty passages... Sampling was a common starting point for the basic sounds. The cool thing about tracker files is that they would include all the pcm samples, so people would build up their library o
284.
by sdenton4 1y ago
SharpPlus Sqlite Developer v3.2.8.354 goes hard.
285.
by sdenton4 1y ago
It creates conditions for more direct accountability. There's a pretty standard pattern of getting elected, doing the more extreme things, and then giving the voters time to cool off before the election happens.
286.
by sdenton4 1y ago
I dunno - having some social proof that a person can actually live with other people successfully is also a helpful kind of indicator.
287.
by sdenton4 1y ago
I lived in communal houses for a very long time. The best was something like this - we had communal meals four nights a week, and everyone had one cooking night. The other nights, you know you just show up and there's food. It's w
288.
by sdenton4 1y ago
But I don't know that from the proposed re-titling of the article. The importance of the conjecture is more salient for the general reader than the (uninformative) names of the people who conjectured in the first place.
289.
by sdenton4 1y ago
Is it an important conjecture, or just something someone came up with last week?
290.
by sdenton4 1y ago
To add slightly more detail: Most research math programs in the US are 5(ish) year 'combined' masters+phd programs. The first couple years are basically course work, seminars, finding your area and advisor, and then the rest is th
291.
by sdenton4 1y ago
All completely correct. It has been extremely difficult to reintroduce functionally extinct species (ie, regenrate populations in zoos, and then reintroduce them to their original habitat). The animal culture piece is really big. But probab
292.
by sdenton4 1y ago
It's worth puzzling about the mechanisms that would lead a fast reproducing species to slow down... My best guess is that internal competition for scarce resources leads to territoriality, which in turn might select for better resource
293.
by sdenton4 1y ago
Doesn't Leibniz's argument assume a single all-powerful god, and an external notion of good and evil? I would call it a 'not-even-wrong' argument - perhaps it's perfectly correct given its assumptions, but the assum
294.
by sdenton4 1y ago
Instead of thinking of this as a benefit of wolves (specifically), you can view it as an advantage of having something in their ecological niche. In this case, the niche is 'apex predator.' It's a niche which is repeated acro
295.
by sdenton4 1y ago
Get the cars out of the streets and the (possibly modified) ebikes onto them... They have a literally smaller footprint, and are a huge improvement over cars.
296.
by sdenton4 1y ago
Otoh, the degradation of democracy into oligarchy and then tyranny was called by Socrates...
297.
by sdenton4 1y ago
Yah, it's a great point that the whole scheme is predicated on very questionable land use policy...
298.
by sdenton4 1y ago
Turns out that people bombing your infrastructure and your people actually makes life worse, regardless of your ranking in the world economic stack ranking...
299.
by sdenton4 1y ago
There's two answers to that.... The first is, how do you know the subjective optimization your making is actually any good? You're just moving the problem back one layer of abstraction. The second is, we did that, eventually, by t
300.
by sdenton4 1y ago
When I was working in audio compression, evaluation was very painful because we had no programmatic way to measure how good some reconstructed audio sounds to a human. Any metric you could come up with was gameable, and direct optimization

More ›