joe_the_user

5 ms·

61.
by joe_the_user 3mo ago
In a sense, what eventually gets legalized through settlement or what doesn't provoke a lawsuit isn't that relevant. The process of creating an LLM involves taking and processing a massive amount of human generated data, roughly a
62.
by joe_the_user 3mo ago
All LLMs are based on distillation broadly defined. Western models began distilling texts. If Chinese models are distilling Western models, they are taking information that Western models don't own anyway - but that doesn't mean t
63.
by joe_the_user 3mo ago
Overall, all the different linear attentions out there are approximations the original (quadratic) attention and this is important for the whole "AI" enterprise[2]. Original attention involves (very crudely) an approach of scannin
64.
by joe_the_user 3mo ago
I'd assume the gp was just inflating their rhetoric but cost of model is an interesting topic. I'd guess Anthropic and OpenAI's models are expensive relative to the cost of running the models but that the revenue still doesn&
65.
by joe_the_user 3mo ago
Employers who expect to know things about employee's personal lives are despicable. I expect to accomplish my tasks. I can be friendly, polite and humorous but you don't need to know anything about how I live. The main factor is a
66.
by joe_the_user 3mo ago
So, you have "Thesis #3: The real threat is the war against the infrastructures and institutions that make all forms of literacy possible". So they're saying things are bad, deeply bad. But they toss darts at everyone with sh
67.
by joe_the_user 3mo ago
The way through is to change the way you relate to your work, and to appreciate the fact that you can now produce far more output. I currently, emphasis on currently, feel like I can leverage LLMs in a very satisfying way to a deeper over
68.
by joe_the_user 3mo ago
OK, I should have put have put things as "when you are primarily interested in some benefit outside of the task at hand". Some of those benefits might be great, sure. But it's still not necessarily going to improve the enjoym
69.
by joe_the_user 3mo ago
I begrudgingly agree, but with the caveat that tending to a vegetable garden in your backyard is also a spiritual pursuit. Mathematics isn’t some special discipline that elevates you beyond other people, as much as it is useful and interes
70.
by joe_the_user 3mo ago
The only "fields" where a genie appearing and doing all your work for you are fields (jobs) where you only about the money at the end. There are some fields where a serious practitioner might be happy to have a genie do all the &q
71.
by joe_the_user 3mo ago
Dogpile was only a good idea while Search Engines were mostly trash. Well, search engines are trash again. Perhaps it should come back
72.
by joe_the_user 3mo ago
It's a matter of degree. If open source means "everything you need to reproduce this" then maybe an open source application should include a detailed architecture document and course in the c language. Oppositely, you can cer
73.
by joe_the_user 3mo ago
If AI were to become a super weapon why should I trust a private company to own it? If AI becomes a super weapon, it's possessor will no longer care whether you trust it. It would be nice if whether "we" trusted OpenAI et a
74.
by joe_the_user 3mo ago
I agree with your overall argument. But I think there's a challenge here - if the "ultimate AI" is now being built in the West, it is being built with private investment from many sources. I think even China depends on invest
75.
by joe_the_user 3mo ago
This is assuming a situation where, A) models "unaligned" by default and B) alignment can be added (though prompts and related things). The point, which isn't very surprising but still notable, is that the models are "am
76.
Can You 'Catch' Dementia from Your Spouse? By StudyFinds Analysis •(studyfinds.com)
10 points by joe_the_user 3mo ago|1 comments
77.
by joe_the_user 3mo ago
I'd say that AIs occasionally "going crazy" and calling for death to human is evidence that these things might "mis-align" on occasion. And I say that knowing that most of these events are just these thing parroting
78.
by joe_the_user 3mo ago
I think you're making a false dictomy. The these models can be actually dangerous - in reality and the people in charge of their development can believe this is true (on various levels) but still not take it super seriously and instead
79.
by joe_the_user 3mo ago
Well, predictions that I'll go to work and things will be as usual aren't as uncertain as Bitcoin pricing or as certain as planetary motion. However, the aim of math in these situations is often to give explanations for intuitive
80.
by joe_the_user 3mo ago
I remember in 2017, it was said that "regular" machine learning techniques were better for time series than neural networks. It seems like this blog post is indirectly references that belief and gives. It seems like the idea is si
81.
by joe_the_user 3mo ago
I think we're saying roughly the same thing. It's natural starting with a statistical model to say it generates data (as the article sometimes does) but in reality, you often/generated have data generated by a possibly determ
82.
by joe_the_user 3mo ago
The question "what probability distribution generated this" is very hard to answer if you're only getting short window before the distribution changes. A lot of distributions could generate relatively "short" set di
83.
by joe_the_user 3mo ago
Unfortunately, the votes for a regime of permanent bailouts were already in when Ben Barnacke made his famous speech proposing helicopters over cities if deflation reached a crisis level (though the Internet says Milton Friedman thought of
84.
by joe_the_user 3mo ago
I don't think you can say the Turing test has been passed in a computer versus determined humans setting. IE humans making a strategy effort to sort humans versus computers as well as humans motivated to distinguish themselves as human
85.
by joe_the_user 3mo ago
People enjoy saying "AI can replace middle management" as a comeuppance but it's not what people think. In many ways, AI is primarily serving as middle management, yes. But what happens is that instead of one manager engaging
86.
by joe_the_user 3mo ago
Indeed, Thinking about, if I had a lab, I'd be trying to get training data from a combination of the open web, piracy and all of my revivals. I wonder how labs are doing that.
87.
by joe_the_user 3mo ago
Now to be "that asshole" why unhoused instead of homeless? @dang, flagging. HN standards are basic civility, starting a comment with "I'll be an asshole" is clearly creating a pretext for violating HN standards. I
88.
by joe_the_user 3mo ago
I would say "get (professional) help!" is advice that always sounds good and can be good for X percent of people and not so good for the rest. I am part of project that offers food and resources to unhoused people outside the usua
89.
by joe_the_user 3mo ago
This thread is filled with strong arguments that ambulence operators do make money with medicare. But even more, it's completely false that the reason an ongoing, working business charges a huge price to some people is that some other
90.
by joe_the_user 3mo ago
Health care in general has been solved by state sponsored medicine in most industrialized nations. And when it isn't solved by directly state sponsored medicine, it's solved by formally and informally regulated monopolies. In fact

More ›