cs702

37 ms·

631.
by cs702 2y ago
Yes. I agree on both counts. But neither fact explains the disconnect between perceptions of personal economic situation vs. national economic situation. "Vibecession" is the conventional name for the disconnect. I'm making n
632.
by cs702 2y ago
When asked how they are doing, 72% of Americans report they are "at least doing okay financially." When asked how the nation's economy is doing, 58% rate national economic conditions as "poor."[a] Many observers
633.
by cs702 2y ago
Thank you. Like you, I also have the impression that folks at those AI safety orgs sincerely believe they're doing the right thing. But I would ask them the same questions I would ask the politicians: Which "experts" did they
634.
by cs702 2y ago
Jeremy -- thank you for sharing this on HN. And thank you also for everything else you've done for the community :-) I agree with you that one of the biggest issues -- maybe the biggest issue -- with the proposed legislation is that it
635.
by cs702 2y ago
Key passage from the executive summary: > To date, HPAI A(H5N1) viruses currently circulating most commonly in birds and poultry, with spillover to mammals and humans, do not have the ability to efficiently bind to receptors that predomi
636.
by cs702 2y ago
> still several of OOM difference I agree. Nothing I wrote above disagrees with your statement.
637.
by cs702 2y ago
The authors train a model with two contrastive objectives: (1) predict which video clip corresponds to a given audio file, and (2) predict which audio clip corresponds to a given video clip. Together, these two training objectives induce th
638.
Self-Supervised Visual Grounding of Sound and Language(mhamilton.net)
1 points by cs702 2y ago|1 comments
639.
by cs702 2y ago
Nice work. The authors train a model with two contrastive objectives: (1) predict which video clip corresponds to a given audio file, and (2) predict which audio clip corresponds to a given video clip. Together, these two training objective
640.
by cs702 2y ago
When the pay package was approved, virtually no one outside Tesla thought the company could hit the insane performance targets upon which the options were contingent. At the time Tesla was struggling to ramp up initial production of the M
641.
by cs702 2y ago
> The Nazis believed in war as the primary engine of human progress... Wow, their understanding of economic growth was... so naive, it was not even wrong .[a] --- [a] https://en.wikipedia.org/wiki/Not_even_wrong
642.
by cs702 2y ago
Very nice. Well-written. Feels "natural." Besides helping with interpretability, my immediate thought is that maybe we could pretrain models faster by adding regularization terms in the objective function that induce representatio
643.
by cs702 2y ago
Full quote: > It’s patently absurd that Apple isn’t smart enough to make their own AI, yet is somehow capable of ensuring that OpenAI will protect your security & privacy. Apple has no clue what’s actually going on once they hand yo
644.
by cs702 2y ago
Not coincidentally, the managers can charge fees only on the capital they're managing, i.e., the capital they haven't returned.
645.
by cs702 2y ago
Yes, the modification the authors propose is indeed "simple." It's "obvious" only in hindsight.
646.
by cs702 2y ago
Duplicate: https://news.ycombinator.com/item?id=40584237
647.
by cs702 2y ago
A fitting symbol for the importance of maintaining appearances in China. "How is the real estate market?" "It's all under control." "How are businesses doing?" "Our business sector is innovating and
648.
Flash-Linear-Attention(github.com)
1 points by cs702 2y ago|0 comments
649.
by cs702 2y ago
This short video clip quickly makes evident why SpaceX has become so dominant, now sending more than 90% of all global cargo to orbit every year. In short: Musk is operating from first principles, questioning everything from the ground up
650.
by cs702 2y ago
old-fashioned text files
651.
by cs702 2y ago
See this thread, also on this page: https://news.ycombinator.com/item?id=40609689
652.
by cs702 2y ago
> At some point we're going to go from tokens to embeddings for everything. Yes, I agree. Further down the road, I imagine we will end up finding interesting connections to the symbolic approaches of GOFAI, given that the embeddin
653.
by cs702 2y ago
*that has not yet been published
654.
by cs702 2y ago
Give the model the tokens "happily" and "I", and add to each input token its respective position embedding and the position embedding for the token to be predicted. You can do this in parallel for all tokens to be pred
655.
by cs702 2y ago
LeCun may or may not be right, but I'm not sure this is relevant to the discussion here. The OP's authors make no claims about how their work might help get us closer to AGI. They simply enable autoregressive LLMs to do new things
656.
by cs702 2y ago
One possibility is that mapping images to a token embedding consumes ~170x more compute+space than mapping a token id. Another possibility is that OpenAI is mapping each image to ~170 vectors in an embedding space that is shared with token
657.
by cs702 2y ago
This looks great . The authors randomly permute (i.e., shuffle) input tokens in training and add two positional encodings to each token: one with the token's position and another with the position of the token to be predicted. Otherwi
658.
by cs702 2y ago
It could be at least a couple of weeks before I can take a second look. At that point, I'll decide if I'm going to study the paper, i.e., work my way through it properly. If I do that, I'll post my thoughts here.
659.
by cs702 2y ago
Reposting my comment from https://news.ycombinator.com/item?id=40584237 : > AGI by 2027 is strikingly plausible. GPT-2 to GPT-4 took us from ~preschooler to ~smart high-schooler abilities in 4 years. Tracing trendlines
660.
by cs702 2y ago
Great post. Worthwhile reading. The industrial revolution was indeed a big deal. I have one suggestion for the author: Please also show plots with a logarithmic scale on the vertical axis, instead of only a linear scale. The logarithmic sca

More ›