fenomas

9 ms·

121.
by fenomas 2y ago
Agreed, and it's such a shame! A kid goes to math class and learns, say, derivatives as this weird set of transformations that have to be memorized, and it's only later in in physics class that they start to see why the transforma
122.
by fenomas 2y ago
And to do so in bad faith, no less - the event seems to have had 2-3 sessions related to feminism (out of a couple dozen), but no connection beyond that. https://schedule.mozillafestival.org/plaza
123.
by fenomas 2y ago
I meant the fact that math works at all, as a symbolic framework with consistency and the power to prove some statements and disprove others, etc. Math only has those features because it examines symbols abstractly regardless of any connect
124.
by fenomas 2y ago
> The true power of math comes from the correspondence between those symbolic transformations and observation from the real world. Not at all, think it through further. Obviously it's true that mathematics is more practically useful
125.
by fenomas 2y ago
It was something like that the app must have been "originally written" in certain approved programming languages. It basically targeted cross-compilation, but it was written so vaguely that on paper it appeared to even disallow na
126.
by fenomas 2y ago
Fun fact nobody remembers: back in the day Adobe built a feature where you could build flash source to an iphone app - you hit "publish" and an .ipa came out. It worked fine, people made apps, they got accepted to the app store, u
127.
by fenomas 2y ago
Weirdly aggressive, but fair enough.
128.
by fenomas 2y ago
I'll edit out the part of my comment where I suggested they don't.
129.
by fenomas 2y ago
> when it is actually about relations I think of it more as, math is ultimately about symbols . Like, if a mathematician says that "2 = 2" is a true statement, a reasonable onlooker might ask "Does that mean that all twos
130.
by fenomas 2y ago
Rule of Goats.
131.
by fenomas 2y ago
> What I'm saying is that there are many "concrete instantiations" of drafting lettering style, that all look basically the same, because they all came from the same source, and Gorton is just one of them. Open a couple of
132.
by fenomas 2y ago
> but someone must figure out how to get "I don't know" information out of them. In Andrej Karpathy's most recent video he discusses in some detail how this is already done and how effective it is.
133.
by fenomas 2y ago
The author isn't saying he finds the style unfamiliar - the entire article is about how ubiquitous it is. He's a typographer, so I think he means "ugly" in the sense of having features that modern typographers consider t
134.
by fenomas 2y ago
> This style of writing didn't start as a font from some company. TFA doesn't suggest that, at all. A font is a concrete instantiation of a writing style, and TFA is about the history of one such concrete instance - not the gen
135.
by fenomas 2y ago
There's a footnote along these lines that links to the actual algorithm.
136.
by fenomas 2y ago
Hrrm, I'm reading back and I may have misinterpreted your first post too. If so apologies!
137.
by fenomas 2y ago
Sorry, I don't follow what you're disagreeing with. I was summarizing what Karpathy talks about in the vicinity of 1:31:00 - where he talks (I assume notionally) about a specific neuron lighting lighting up to indicate uncertainty
138.
by fenomas 2y ago
Note that he's inferring from a base model there, which are fairly capable of regurgitating their (highly-weighted) inputs since they do nothing but predict pre-training tokens. For instruct services like ChatGPT, if they regurgitate s
139.
by fenomas 2y ago
For point 1, he gets into that more later in the video - e.g. specifically on counting, and about how/when to have models invoke tools instead of doing math themselves, etc. Also for the second point, check later in the video when he t
140.
by fenomas 2y ago
Yes, IIUC those points are correct - you need relatively capable models, and well-crafted questions. The comparison with AlphaGo is that the processes are analogous, not identical - the key point being that in both cases the model is choosi
141.
by fenomas 2y ago
Karpathy's point in the video is that the models don't need to be exhaustively told what they don't know - they already have a good understanding of the extents of their knowledge. Older models just didn't use that under
142.
by fenomas 2y ago
As I understood that part, in RL for LLMs you take questions for which the model already sometimes emits correct answers, and then repeatedly infer while reinforcing the activations the model made during correct responses, which lets it evo
143.
by fenomas 2y ago
Strongly disagree. People jump to that when they see a healthy society that's relatively homogeneous, and ignore counterexamples like relatively homogeneous countries, or states within countries, that have poor education/healthcar
144.
by fenomas 2y ago
> This is rarely talked about but is so important > 98% ethically Japanese In the million of these discussions I've seen, this is usually the first/only explanation people jump to. (moreso for people only superficially famil
145.
by fenomas 2y ago
It's not common , but it's a notable widely publicized issue - so if you're writing a video game and you want a low-level antagonist for the hero to fight, it'd be one of the default options.
146.
by fenomas 2y ago
Please read things in context - my comment was about what people mean when they talk about a thing giving developers superpowers. I was not making a claim about how much you, personally, benefit from debuggers. Also: I don't use debugg
147.
by fenomas 2y ago
Devs who don't feel that way aren't talking about the stuff you're talking about. Look at it this way - a powerful debugger gives you superpowers. That doesn't mean it turns bad devs into good devs, or that devs with a p
148.
by fenomas 2y ago
I don't follow this take. ChatGPT outputted a bug subtle enough to be overlooked by you and your colleague and your test suite, and that means it's not ready for prime time? The day when generative AI might hope to completely ha
149.
by fenomas 2y ago
> What's going on is you're not able to see how alphas apply to human society. You're not able to jump the intellectual gap to identify, "hey if packs of animals have alphas, what's the human equivalent?" I
150.
by fenomas 2y ago
> Alphas are just leaders If you redefine alpha to just mean "leader", then the claim that CEOs are alphas is obviously true, but also meaningless. But the theory TFA is about was not just that wolf packs had a leader. It made

More ›