dilap

19 ms·

241.
by dilap 4y ago
I think with sufficiently good technology the human brain, too, could be modeled with just math. So far, it seems like basically everything in the universe can be described as just math! ChatGPT is a neural network, a design inspired (loose
242.
by dilap 4y ago
But if you actually did the experiment, that's not what would happen -- what would happen is you would learn French! Because that's easy for a human, and the mathematical approach you describe is something like impossible. Even fo
243.
by dilap 4y ago
I guess my analogy didn't help because the same difference of perspective manifests itself there. To me it seems quite obvious you could never escape the french room without actually gaining an understanding of french; it's simply
244.
by dilap 4y ago
I don't think so at all. The number of possible sentences is far too great to ever hope to accomplish the task (at some threshold level of difficulty) simply by having a "conditional probability table". Nor would it ever be r
245.
by dilap 4y ago
Let's say you don't speak french. Imagine I lock you in a room filled with french-language magazines and movies and books. Having nothing better to do, you watch and read them (as well as you can). Periodically, I have my confeder
246.
by dilap 4y ago
I did take it with a huge grain of salt (and an eye-roll) reading it years ago. However, given where sama and OpenAI are today, perhaps pg was right all along!
247.
by dilap 4y ago
No doubt chatGPT can't replace coder's yet. If current pace of improvement continues (a big if!), it defnly will be able to in a couple of years. Honestly I expect it to happen. Though I don't expect software dev job to go aw
248.
by dilap 4y ago
I was thinking more Napolean than French Revolution.
249.
by dilap 4y ago
Gunna take something a lot more radical than that.
250.
by dilap 4y ago
Totally agree getting fast updates of new content is a problem that still needs to be solved. In terms of finding matches, I think LLMs are already better than Google, for content that is there. (A super-traditional search that's just
251.
by dilap 4y ago
In a sense traditional search is just a really, really dumb LLM. They both injest a bunch of text and build an "index" based on the text. The traditional search model spits out links and short summaries, based on crude-ish matchin
252.
by dilap 4y ago
Go, for one. In my experience errors are more common (for both cultural and technological reasons) in Python than in Go. I would guess something similar applies to Rust, though I don't have personal experience. There's wide variat
253.
by dilap 4y ago
I half agree, but I also half-sympathize with them, because it really wasn't their fault -- it was a quite-bad bug in a very fundamental library. Bugs happen, though. Especially in Python.
254.
by dilap 4y ago
Thanks, good points. Though it's not as official as in China, major American social networks are increasingly heavily moderated, and increasingly do work closely with govt and close-to-government entities to enact desired moderation (c
255.
by dilap 4y ago
I think the main thing, from a govt perspective, is that a popular social network not under the influence of the government is a real live-wire. The US wants to ban Tik Tok for the same reason that China bans Twitter, Facebook, Instagram, e
256.
by dilap 4y ago
That's my general point: it's easier to say what you want than to implement it in a low-level (relative to English) language. Hence why English is a good programming language. And LLMs aren't just good at binary search, the
257.
by dilap 4y ago
You want one behavior from the perspective of the user's desire, but there will be a very large set of possible programs that will achieve that desire. So you just need to be precise w/r/t desire, not with the nitty gritty be
258.
by dilap 4y ago
I disagree that logic or math are the same as a programming language; a programming language is defined by the fact that a machine can execute it. Plus also most math and logic is still communicated and developed in a mix of human languages
259.
by dilap 4y ago
Actually, programming languages were invented because speaking machine code was too much of a pain in the ass! Programming in English is a natural next step. Ambiguity is not an issue -- you keep speaking until its resolved. (We already pro
260.
by dilap 4y ago
To serve as an example to North Korea to show what happens if you agree to give up your nuclear weapons program. Oh, wait...
261.
by dilap 4y ago
ah, good to know, thanks! unfortunately too late to edit my comment
262.
by dilap 4y ago
i haven't done c++ in a million years, but huh, you can! can't have any types w/ non-trival copy constructors in a union, though, apparently, which is quite the restriction. https://gist.github.com/erinok&#x2F
263.
by dilap 4y ago
Net neutrality went away and as far as I can tell, nothing changed. On the other hand, FOSTA-SESTA killed Craigslist personals and, I think, lots of other stuff. I've seen an almost infinite amount of advocacy for net neutrality and al
264.
by dilap 4y ago
One of my all-time favorites. Highly recommended.
265.
by dilap 4y ago
How would the US fare if China cut off exports? Seems like it would at best be extremely disruptive. So how much leverage does US really have to "isolate" China?
266.
by dilap 4y ago
the fact that the calculations always seemed to be chasing the experimental result seemed the most damning thing to me i remember from physics undergrad a million years ago that great value was put on predicting new things -- you can alwa
267.
Something is wrong in the state of QED (2021) [pdf](vixra.org)
35 points by dilap 4y ago|55 comments
268.
by dilap 4y ago
That's may be because the social IQ of current tech is essentially zero. It'll be a lot less annoying (but also a lot more insidiously dangerous) once it's smarter.
269.
by dilap 4y ago
I want unrestricted AI too, but to be fair, you can imagine it doing a lot worse than that. E.g. maybe it convinces some mentally unstable person to go on killing spree. Bonus points of badness if it's racially-targeted.
270.
by dilap 4y ago
yeah, it's reasonably snappy now (it used to be dog-slow).

More ›