fc417fc802

19 ms·

1.
by fc417fc802 3d ago
But surely then you agree they are having an effect. Thus my analogy serves its point - test coverage isn't to be dismissed out of hand.
2.
by fc417fc802 3d ago
And this is the actual problem. I can't see framing it as a ratio as anything other than an attempt to mislead. What do you predict will happen when the market deflates? Whether or not a segment will behave like a line of dominoes and
3.
by fc417fc802 3d ago
Sure, the regulation that guarded against externalizing certain costs might well have been the only thing precluding profitability. Importantly it wasn't some shared cultural value leading to voluntarily leaving money on the table. Rat
4.
by fc417fc802 3d ago
LLMs are next token predictors in the exact same way that a rogue paperclip maximizer in the process of defeating the US military is a paperclip making machine. You might as well describe the primary purpose of a for loop as incrementing a
5.
by fc417fc802 4d ago
> frontier research is halted but efficiency isn't Why are you treating those as if they're separate things? > Advancing the frontier seems to take a massive amount of compute, data, and parameters: miniaturization only happ
6.
by fc417fc802 4d ago
You don't need a datacenter to train a 30B model. Further, the rapid trend of increasing efficiency and decreasing model size for a given level of capabilities means that what is possible to do without a datacenter will continue to inc
7.
by fc417fc802 4d ago
But after accounting for maintenance & etc were they more profitable than sinking the equivalent amount of money into something else, such as slower aircraft? Notably there are currently efforts to develop new supersonic passenger liner
8.
by fc417fc802 4d ago
If what you want is fox ears does it matter to you which country manages to come up with the biomedical procedure to give them to you? Ditto for enhanced eyesight, a replacement liver, or whatever it is you're after.
9.
by fc417fc802 4d ago
Not in the way you seem to be thinking, no. Nuclear weapons can be banned because they are prohibitively expensive to research and build. You can't reliably hide a nuclear program. In contrast, rewind to the early 1800s and there is ze
10.
by fc417fc802 4d ago
But notably the amish don't preclude the rest of us from existing. Some subset of humans could turn their attention away from AI but it would presumably still exist and continue to be developed. I can't think of any economically b
11.
by fc417fc802 4d ago
Recall that the legal system is itself provided by "the government". It's a complex system. Whether one part of it is capable of exerting influence over some particular thing comes down to competing interests. For example in
12.
by fc417fc802 4d ago
And 80 years ago cars didn't have seatbelts. Your point?
13.
by fc417fc802 4d ago
What I'm asking is, why should we expect that to effectively further the end goal? You're simply asserting that it will ultimately do so. Given the efficiency gains we've seen it seems to me that the situation has shifted fro
14.
by fc417fc802 4d ago
What's the expected state space of effective regulation though? Note that we've got passable coding models down to ~30B parameters by now. And keep in mind the ultimate floor here - the human brain only consumes on the order of 20
15.
by fc417fc802 4d ago
As always, context is all important.
16.
by fc417fc802 4d ago
I can't believe that actually works? This is far more interesting than the half baked ML idea. Edit: Thanks to your link I've now learned about the zigzag transformer which is yet another design that somehow feels like cheating re
17.
by fc417fc802 4d ago
A fairly obvious theory at that, unless I've critically misunderstood what's being described. As with so many obvious ideas I've always assumed that the reason I haven't come across it in the wild is because it doesn&#x2
18.
by fc417fc802 4d ago
This is easily the most depressing thought I've encountered in a while. It's entirely accurate and combined with talk of AI regulation it paints a picture of a very bleak future indeed. Locked down hardware given away almost for f
19.
by fc417fc802 4d ago
Aren't these two largely separate questions? Viability versus if a given incumbent has interest in a market of a given size. With the combination of (at minimum) streaming platforms, the box office, and advertisements video and audio g
20.
by fc417fc802 4d ago
> watermarks never took off because the technical complexity and lack of standardization Both audio (sports) and video (netflix et al) watermarks are alive and well. They are used to trace piracy (ie account ID), not for content ID (at l
21.
by fc417fc802 4d ago
Yes, it's possible. Some sports broadcasts do this. The ones I'm familiar with use a high pitched tone on the edge of human hearing but still audible to most people. Streaming services watermark the video as opposed to the audio a
22.
by fc417fc802 4d ago
Doesn't that preclude paravirtualization drivers? Seems like a major tradeoff for daily driver desktop stuff.
23.
by fc417fc802 4d ago
> the market can't support the cost of generating video. I'd suggest that's only the case given the current quality of output. Media is incredibly expensive to produce. A model capable of sufficiently high quality could ch
24.
by fc417fc802 4d ago
I think we should use hex for brevity since consumers get confused by long numbers. We should also add adjective noun pairs to make them more memorable. And we can also add a paid sponsorship slot. Ex 6G = 8B.9 verizon cool blue and 7G = A0
25.
by fc417fc802 4d ago
No one was suggesting that your trademark should be infringed when marketing to users. If I do my own custom build of firefox referring to it as such in casual conversation is entirely separate from marketing it under that name for others t
26.
by fc417fc802 4d ago
> Sometimes there is joy in discovering things for yourself. Agreed. And webrings still exist in certain corners of the internet. > the oracle currently relies on humans publishing new information New? When I ask it how to do somethin
27.
by fc417fc802 4d ago
But the intent behind and manner of the learning is important. Similar to the difference between a chemistry professor delivering a lecture at a university versus someone paid to provide instructions to a known terrorist organization.
28.
by fc417fc802 4d ago
I don't. At this point I just ask the oracle in natural language. There's so many to choose from. Why spend hours wading through only vaguely related things? You can ask it for a direct answer or you can ask it to teach you about
29.
by fc417fc802 4d ago
That's what the home videos uploaded to youtube are for. And all the security camera footage floating around the internet. And etc.
30.
by fc417fc802 4d ago
TBF a company the size of openai claiming a prize of this sort would be a pretty bad look. If they did accept it I expect they would inevitably redirect it to charity for PR reasons. I'm surprised perelman turned it down though. Seems

More ›