cs702

34 ms·

451.
by cs702 2y ago
Nowadays, whenever I browse Netflix, I feel like that Bruce Springsteen song, "57 Channels (And Nothin' On)."[a] Sure, there are lots of choices, but they all kinda suck . I find myself wondering, why? The OP weaves an insig
452.
by cs702 2y ago
I'm not implying anything or delving into metaphysical matters. All I'm saying above is that the number of neuron-neuron connections in current AI systems is still tiny, so as of right now, we have no way of knowing in advance wha
453.
by cs702 2y ago
I never claimed the machines would achieve "human level," however you define it. What I actually wrote at the root of this thread is that we have no way of knowing in advance what the future capabilities of these AI systems migh
454.
When will computer hardware match the human brain? (Hans Moravec, 1997) [pdf](jetpress.org)
1 points by cs702 2y ago|0 comments
455.
by cs702 2y ago
Actually, I'm comparing the "neuron-neuron connection count," while admitting that the comparison is not apples-to-apples. This kind of comparison isn't a new idea. I think Hans Moravec[a] was the first to start making t
456.
by cs702 2y ago
Thanks. Yeah, I've heard there are a bunch of efforts like that, but as far as I know, all are very early stage. I do wonder if the most energy-efficient way to scale up AI models is by implementing them in organic substrates.
457.
by cs702 2y ago
That's a hardware question, not a software question, but it is a fair question. I don't know if the hardware can be scaled up. That's why I wrote " if we're able to scale them" at the root of this thread.
458.
by cs702 2y ago
Imagine it were possible to take a rat brain, keep it alive with a permanent source of energy, wire its input and output connections to a computer, and then train the rat brain's output signals to predict the next token, given previous
459.
by cs702 2y ago
It's already better than real rats-as-a-service, certainly: https://news.ycombinator.com/item?id=42449424
460.
by cs702 2y ago
Yes, I agree, but energy efficiency is orthogonal to capabilities.
461.
by cs702 2y ago
Humankind has developed all sorts of systems and processes to cope with the unpredictability of human beings: legal systems, organizational structures, separate branches of government, courts of law, police and military forces, organized ma
462.
by cs702 2y ago
No, it doesn't "solve" the unpredictability problem. But we haven't solved it for human beings either. Human brains are unpredictable. Look around you.
463.
by cs702 2y ago
As of right now, we have no way of knowing in advance what the capabilities of current AI systems will be if we are able to scale them by 10x, 100x, 1000x, and more. The number of neuron-neuron connections in current AI systems is still ti
464.
by cs702 2y ago
In Australia and several Southeast Asian countries, Microsoft recently added Copilot to the consumer version of Microsoft 365, and raised its price. Users in those regions are now complaining that whenever they edit a document, Copilot’s ra
465.
by cs702 2y ago
A though-provoking essay that makes me think "yes, that's exactly right," again and again, whenever I re-read it. Highly recommended. Please consider using the original title: is "On the nature of Computing Science.&quot
466.
by cs702 2y ago
Yes, there's a lot of hype, much of it fueled by all those "ai influencers."
467.
by cs702 2y ago
Yes, there's a lot hype, and also quite a bit of drama, but there's also real progress. Every time I read anything by Gary Marcus, I feel like we're stuck in a silly loop: while True: progress = HardWordByMany_functio
468.
What a Christmas cash scheme from the Great Depression tells us about money(ft.com)
2 points by cs702 2y ago|1 comments
469.
by cs702 2y ago
Brilliant. The artist's website[a] described the program perfectly using the soul-crushing jargon of corporate bureaucracy: > Our program is a professional service to the financial industry; rats are being trained to become superi
470.
by cs702 2y ago
The intuition is that the relative frequency at which past tokens get attention from future tokens is a good proxy for their relative importance. The model the authors use, in fact, maps attention scores to features in the frequency domai
471.
by cs702 2y ago
Very clever, very meta, and it seems to work really well. The two big take-aways for me are: * It's possible to train a model to learn to summarize context from the attention matrix , based only on dot-product scores (k @ q.T * mask),
472.
by cs702 2y ago
Maybe. I'm not so sure. My take is, he's trying to be thorough. The lack the formatting sometimes makes his posts feel like a wall of uninterrupted text. When reading his stuff, I find it helpful to put my browser in "reader&
473.
by cs702 2y ago
This is a thought-provoking essay by Dan Luu, whose essays I always find thought-provoking. I'm surprised Dan didn't make the connection that the webs of mistrust between fiefdoms that form inside organizations as they grow are...
474.
by cs702 2y ago
Fabulous. A big round of applause for Andy Kowinski and the SWE-bench folks!
475.
by cs702 2y ago
Thanks for the quick reply! Interesting. I would have thought one of those "minimum viable" RNNs (like https://arxiv.org/abs/2410.01201 ) would have been ideal for this. I might tinker a bit with this :-)
476.
by cs702 2y ago
Nice work. Thank you for commenting on HN! Did you guys try using an RNN or some other kind of DNN to encode the patches?
477.
by cs702 2y ago
Not sure how applicable Idaho's situation is to other states, because Idaho is one of the poorest states in the country.[a] My quick take: Laws and regulations can be bad or even terrible... but sometimes they are good and necessary. F
478.
by cs702 2y ago
Yes, I see what you mean about the comments elsewhere on this thread/page. For what it's worth, I was only trying to avoid going off-topic ¯\_(ツ)_/¯ Regarding providers overcharging, the following article by Atul Gawande (fro
479.
by cs702 2y ago
They are an integral part of the mess we're in, and they have a vested interest in the status quo. Otherwise, I agree with you.
480.
by cs702 2y ago
Yes. I almost added a comment about that too. In my experience, all doctors, nurses, and other caregivers with whom I have interacted at hospitals have told me they feel treated like "factory workers" who must meet stringent &quot

More ›