zone411

23 ms·

271.
by zone411 3y ago
My generative melody models have done it for over a year but that's with human curation, so it's not self-improving. It's typically easier to curate than to generate, and it's especially true for music and images. It&#x2
272.
by zone411 4y ago
What's your source for "math outcomes are way better in every aspect"? Figuring this out is trickier than it seems. Not only do students vary in demographics, socio-economic factors, and English-learners, but there are also r
273.
by zone411 4y ago
https://hechingerreport.org/despite-popularity-with-parents-...
274.
by zone411 4y ago
Poker would've been a better choice because of the imperfect information and bluffing.
275.
by zone411 4y ago
When Amazon shut down IMDb's film forums, I contacted them to see if they were interested in selling them to at least archive the content. They were not interested. It's sad to see them kill off another Internet fixture.
276.
by zone411 4y ago
What? We were talking about making a language model. I mentioned the plugins in relation to the question of commercializing. I'm very clear about what my project was and was not doing. I get that you're bitter because Alexa became
277.
by zone411 4y ago
No, I have no plans to openly publish any of it. Some of my researcher employees have published their own stuff. I've previously written about how it was a huge commercial mistake for Google and others to openly publish their research,
278.
by zone411 4y ago
You're making incorrect assumptions. This project wasn't about scaling any published approaches. It was original neural net research that produced excellent results with a new architecture without self-attention, using a new optim
279.
by zone411 4y ago
I'm quite surprised at how little progress FAANG companies have made in recent years, as I believe much of what's happening now with ChatGPT was predictable. Here's a slide from a deck for a product I was developing in 2017:
280.
by zone411 4y ago
The previous post was flagged: https://news.ycombinator.com/item?id=35448899
281.
by zone411 4y ago
Please stop linking to Ted Chiang's article. I like him as an imaginative writer, but his article is just wrong and gives readers incorrect intuitions. His claim that GPT models are not able to learn decimal addition has been known to
282.
by zone411 4y ago
Right, it's switched around compared to the rest of the post.
283.
by zone411 4y ago
It really shouldn't act as a sparse model. I would bet on something being off.
284.
by zone411 4y ago
Not endorsing the arguments either way but let's say DNA printing ( https://time.com/6266923/ai-eliezer-yudkowsky-open-letter-no... ) or something like Stuxnet or crashing a nuclear-power country's stock market
285.
by zone411 4y ago
Here's a good comment from Marginal Revolution about this: https://marginalrevolution.com/marginalrevolution/2023/03/en...
286.
by zone411 4y ago
Why wouldn't we be able to evaluate their performance and compare them to humans? The purpose of test datasets is to do just that, and new ones are created every day. By combining several of them, we can create a decent benchmark. We c
287.
by zone411 4y ago
Oh nice, 65B! I was planning to try it out sometime but have been waiting for various repos to get their issues sorted out and I'm much less interested in smaller models. Are you using GPUs or CPU? Any tips on what to use? What's
288.
by zone411 4y ago
However, if you ask ChatGPT (GPT-4) about Wordle, it knows about it. > tell me about wordle Wordle is a popular online word game that gained significant traction in the early 2020s. The game's objective is to guess a secret five-let
289.
by zone411 4y ago
>Am I reading this correctly that the assumption here is that programming and writing skills aren't reliant on critical thinking? No, they're just listing some skills that have both negative and positive associations with expos
290.
by zone411 4y ago
The Pillars of the Earth and its follow-up by Ken Follet are fun (but long) fiction novels set in Medieval England, which help you envision living in that world. I'm not sure how accurate they are when it comes to people's general
291.
by zone411 4y ago
It underlines the idea that the entropy-time connection is pretty shaky though.
292.
by zone411 4y ago
"there is 10 million times more entropy in that radiation [cosmic microwave radiation] than there is in all of the mass of the universe" https://news.berkeley.edu/2016/09/20/new-book-links-flow-of-..
293.
by zone411 4y ago
This idea of time's relation to entropy is pretty iffy. I recommend this book https://news.berkeley.edu/2016/09/20/new-book-links-flow-of-... .
294.
by zone411 4y ago
Life reduces local entropy.
295.
by zone411 4y ago
I believe that this type of video captioning will be able to fill the gaps in the common-sense knowledge of and understanding of the world and its physics for LLMs. It should also be useful for robots.
296.
by zone411 4y ago
Human curation of AI-generated content is the true future.
297.
by zone411 4y ago
Right, not sending out your text to third parties is absolutely huge for certain industries. ChatGPT has been already banned by some companies, so the issue is whether or not the language model can be used at all. I'll add the possibil
298.
by zone411 4y ago
Yeah, there is no such "general consensus." I don't know where the OP got this idea.
299.
by zone411 4y ago
I've been reading many machine learning papers to keep up with the field in the last year and only implemented some. Are you saying that I completely wasted my time reading by them and learned nothing?
300.
by zone411 4y ago
You're splitting hairs and being too literal. We're comparing learning from a lecture, a book or a video to learning from an AI. The amount of thinking is the same (or more if the AI is personalized).

More ›