@glyph Did you quote post something?
-
@glyph it is curious, to me, that this kind of money doesn’t chase massive infrastructure investments, such as pumped storage, grid infrastructure, or expanding companies in the supply chain. One would imagine predictable and reliable returns with a need for capital to make it happen.
-
@glyph it is curious, to me, that this kind of money doesn’t chase massive infrastructure investments, such as pumped storage, grid infrastructure, or expanding companies in the supply chain. One would imagine predictable and reliable returns with a need for capital to make it happen.
@benjohn those companies are profitable, but vulnerable to competitive pressures (i.e.: capitalism) which will reduce margins over time, rather than having a “moat” (illegal but unprosecutable monopoly due to intellectual property law) to defend their margins permanently. nobody hates capitalism more than a capitalist
-
@glyph I know his real name is Henry, but I learned that after watching him for years. Never gonna overwrite Joel in my brain
-
it's truly amazing what LLMs can achieve. we now know it's possible to produce an html5 parsing library with nothing but the full source code of an existing html5 parsing library, all the source code of all other open source libraries ever, a meticulously maintained and extremely comprehensive test suite written by somebody else, 5 different models, a megawatt-hour of energy, a swimming pool full of water, and a month of spare time of an extremely senior engineer
@glyph The troubling aspect to me is the question about labor power and the way the models are produced. If it is truly possible to reimplement a complex parser in a new language for 1MW/h and one developer month that seems like a good trade. The amount of security related churn and heartache that could be avoided by rewriting popular libraries away from memory-unsafe languages, that's a substantial benefit. I doubt it's quite that easy or it would be happening in more cases though.
-
@GroupNebula563 @AlsoCrowie my point was that only any drained or evaporated water is actually used up. Anything in a cooling pipe is constant and stable for years
@RichiH @GroupNebula563 @AlsoCrowie it seems they are not using closed circuits. And the water may not be suitable for consumption after they used it.
-
@riverpunk @rupert @stuartl @glyph This is truly the kind of discussion I'm here for.
-
it's truly amazing what LLMs can achieve. we now know it's possible to produce an html5 parsing library with nothing but the full source code of an existing html5 parsing library, all the source code of all other open source libraries ever, a meticulously maintained and extremely comprehensive test suite written by somebody else, 5 different models, a megawatt-hour of energy, a swimming pool full of water, and a month of spare time of an extremely senior engineer
@glyph seems like copy/paste with extra steps...
-
@glyph I never want to cast shade on addicts for being addicted, addictions are fucking awful.
I absolutely will cast shade on addicts *or anyone else* for insisting I should be addicted too, and for transforming all of society around the idea that my being addicted is a good thing.
-
@glyph I never want to cast shade on addicts for being addicted, addictions are fucking awful.
I absolutely will cast shade on addicts *or anyone else* for insisting I should be addicted too, and for transforming all of society around the idea that my being addicted is a good thing.
-
@glyph I just keep coming back to that LLMs put our brains directly on the inner loop of an optimization algorithm — that's already true to some degree with advertising and social media engagement algorithms, but LLMs tighten that loop even more, and we don't know what that does to brains!
-
@glyph This feels about right, and with this, the next trick is building a culture of healing and support for when those who have fallen prey to this addiction are ready for change.
-
@glyph This feels about right, and with this, the next trick is building a culture of healing and support for when those who have fallen prey to this addiction are ready for change.
@mttaggart unfortunately I have no concept of what "rock bottom" for an LLM user looks like. Step 0 is going to have to be that society stops massively rewarding them with prestige and huge piles of cash first
-
@glyph I just keep coming back to that LLMs put our brains directly on the inner loop of an optimization algorithm — that's already true to some degree with advertising and social media engagement algorithms, but LLMs tighten that loop even more, and we don't know what that does to brains!
@xgranade looks like we're gonna find out
-
@xgranade looks like we're gonna find out
@glyph Maybe! Or maybe it'll be like leaded gasoline where we never really are able to trace back which awful things were due to that source of lead and which things weren't because it all gets mixed up in the "heavy metal in head, things bad now" bucket.
-
@xgranade to this point I have had kratom users — mostly indirectly, I am not particularly close with any — suggest that I try it because it is "more effective" and "easier to get" than prescription ADHD meds. And I'd definitely be lying if I didn't say I feel a *strong* pull towards believing that. It would be very nice to solve all my problems with a pill or a prompt
-
@glyph I never want to cast shade on addicts for being addicted, addictions are fucking awful.
I absolutely will cast shade on addicts *or anyone else* for insisting I should be addicted too, and for transforming all of society around the idea that my being addicted is a good thing.
-
@mttaggart unfortunately I have no concept of what "rock bottom" for an LLM user looks like. Step 0 is going to have to be that society stops massively rewarding them with prestige and huge piles of cash first
@glyph Absolutely, but even in that we can make out the general shape of the thing. We know this current model is not economically sustainable by any party. For users, the result will be inability to access models, or paying cripplingly high prices to do so. Option 1 will elucidate their inability to function without the models, and option 2 will impose the kinds of costs that look like rock bottom in other addictions.
-
@glyph Absolutely, but even in that we can make out the general shape of the thing. We know this current model is not economically sustainable by any party. For users, the result will be inability to access models, or paying cripplingly high prices to do so. Option 1 will elucidate their inability to function without the models, and option 2 will impose the kinds of costs that look like rock bottom in other addictions.
@mttaggart my suspicion is that when this happens we are going to find out that there is a huge predisposition component. there will be people who say "ah well. guess I'm a little rusty writing unit tests now, but time to get back at it" and we will have people who will go to sleep crying tears of frustration for the rest of their lives as they struggle to reclaim the feeling of being predictably productive again
-
@glyph Maybe! Or maybe it'll be like leaded gasoline where we never really are able to trace back which awful things were due to that source of lead and which things weren't because it all gets mixed up in the "heavy metal in head, things bad now" bucket.
@xgranade definitely going to make a killing when I put all this low-linear-algebra steel back on the market