Elon: "A major part of real-world AI has to be solved to make unsupervised, generalized full self-driving work, as the entire road system is designed for biological neural nets with optical imagers"
Seems like he's inching closer to reality by the day.
Yeah CTRL labs has been flying under the radar, but they have a ton of top talent and already demonstrated insane results. I will go further to say they are easily ahead of neuralink in terms of bringing a faster-bandwidth interface; of course, for more ambitious projects that Musk has tweeted about like mental health disorders, you must go through the brain, but there's no indication the neuralink approach will scale to this either.
Elon Musk has been pretty successful at getting top talent to push the limits of engineering, but with neuroscience, it seems the research isn't there yet. What I'm really interested to see is how long Karpathy will stay at Tesla as other self driving projects drop like flies (ATG, Zoox, and most recently L5)
The OP is claiming that. It is the purpose of their most recent tweets.
What I'm saying is there shouldn't be outrage until we figure out that the OP isn't lying, which there seems like there is reasonable doubt for. Because for very evil circumstances I wouldn't expect there to be transparency either for legal reasons.
That's my bad. Because the gmail link later in the thread is very likely to be mobile, I assumed as the first screenshot was so centered that it was also mobile. But it could be on desktop, so I edited my response
The entire thread is full of people jumping on the hate train, did anyone even stop to consider that Google is right on this?
https://support.google.com/accounts/answer/40695?hl=en There are two points here: first, they claim to provide a reason for closing your account when you log in to your google account (not google service). The OP provided what looks like google service notification on his mobile, but could be on desktop (thanks to reply for noticing this). Second, after escalating the issue internally, decided to email (internally) people saying this is an attack on his family. He is also publicly looking for a job. This seems like a severe overreaction for a google account ban... I think there is reasonable doubt here to suspect something is wrong here more than "Google decided to arbitrarily ban a specific account among billions in order to target one of their employees."
It's not illegal or collusion to remove their apps and products from Apple phones. I disagree with you - if you couldn't use instagram / fb messenger(s) / tiktok / YouTube, wayyyy less people would buy Apple.
@ the other comments, email greg brockman directly if you have a good idea and he'll speed up your access. I applied only a few weeks ago and I'm not famous (no twitter account), I just mentioned i'm a MS student and I had an idea of what I wanted to do
I have access — it's actually quite easy to use? People have also built tools say, in python, that allow you to quickly test prompting (it's somewhere on tech/vc twitter...) Like you said there isn't fundamental difficulty, there are just far more interesting things to do
I think you hit the nail on the head, with the salient point here being that in the near future "creative" things will be automated first (see Image GPT, Jukebox, etc. Google has 100 billion dollars cash and countless TPUs, best engineers, infra, etc - they could probably replicate results far better than each of these OpenAI projects within a few years). One of the things that got me into ML research was the notion that we could automate a lot of the hard work humans do every day (agriculture, cooking, desk jobs, etc) so that humans could do things that were uniquely theirs & interesting, that were human, that were beautiful... Unfortunately it turns out that classical music and waxing poetic are easily generative in an enjoyable way. In the most ironic fashion possible, it turns out that the very thing we do when we conduct ML research, what you call the "logical domain", is one of the only things that stays human-only in the foreseeable future.
GPT-3 and other projects seem to drive hype cycles in the tech community and convince people like Elon Musk that the AGI revolution is near. But I think recent progress is just another example of machine learning models being able to generalize on super large datasets, even if it's the biggest model so far. It's not clear to me that larger models will solve this in the limit; take the way GPT3 fails on addition past a certain number, and the fundamental inability for transformers to learn certain algorithms. It is certainly still possible for this type of large dataset, large model style of ML to make human life better in many ways - like Tesla is trying to do with self driving cars, or Covariant with automating Amazon-like jobs. But I think when it comes to tackling the hard problems of true intelligence, we're missing a dimension somewhere.
Read slide 17. The key assumptions are that: the confounder present at bunching point is representative of the confounder present everywhere, and the confounder affects outcome linearly. The second assumption is what strikes me as too strong, and you can read the paper Appendix C for how they address it. The reasoning is quite weak and they even phrase it as such: "Our main empirical findings
therefore do not seem to be an artifact of this linearity assumption". Overall I would take this paper with a grain of salt, from a cognitive science point of view it just makes no sense that doing more reading would have no cognitive benefits (in fact, we know the opposite to be true.)
Also, the paper hasn't even been reviewed yet. We should revisit it after it has been peer reviewed.
Your comment seems to imply that it didn't help. Would you recommend picking lifting/running up? I went on meds recently for the same issues but had a bad (lasting) experience, so I'm looking towards therapy and healthy eating / exercising for the future now.
I consider myself pretty strong at math (in university right now) and I was stumped by the vector space question. I never considered, actually, what domain scalars should be drawn from.
Wikipedia says "the scalars can be taken from any field, including the rational, algebraic, real, and complex numbers, as well as finite fields."
he made many contributions to other parts of science (wireless, communications), though, that I doubt bezos would be able to so the point still stands.