Gregory's ServerThe planet was already on a course for death before we figured out how to make useful brain analogues on silicon hardware. Also, if and when we shift completely to carbon neutral power, the amount of electricity that AI consumes (which will only go down as processing becomes increasingly more power efficient) won't matter
|
17 comments
 @intransitivelie @jlsigman @martin_piper @ben I'd say being able to predict protein folding is pretty fucking useful. @atatassault @intransitivelie @jlsigman @martin_piper @ben This isn't an LLM using data without attribution lol @bananarama @intransitivelie @jlsigman @martin_piper @ben Deep Learning, Machine Learning, LLM, etc all mean the same thing: Neural Network AI @atatassault @intransitivelie @jlsigman @martin_piper @ben Additionally, each of those terms have distinct technical meanings. They're not the same.  @bananarama And neither of these techniques has a mode that provides attribution. It's an open research question that some but not many are discussing. @atatassault @bananarama @jlsigman @martin_piper @ben  @intransitivelie @atatassault @bananarama@mstdn.social @jlsigman @ben@m.benui.ca your denial is not factually correct. You've not provided any valid argument. Your denial without any reliable claims can be safely ignored.  @atatassault @bananarama @intransitivelie @jlsigman @martin_piper @ben > "Deep Learning, Machine Learning, LLM, etc all mean the same thing: Neural Network AI" Chatgpt, is that you? @bananarama @atatassault @intransitivelie @jlsigman @ben that is transformative which is in the agreement @martin_piper @atatassault @intransitivelie @jlsigman @ben Oops, still requires attribution. Try again. EDIT: Also notice, it is share-alike. @bananarama @atatassault @intransitivelie @jlsigman @ben and how do you know there won't be attribution? You're just assuming that. @martin_piper @atatassault @intransitivelie @jlsigman @ben I can't prove a negative, try again. To anyone reading this after the fact: the evidence is that LLMs are notoriously bad for hallucinating attribution. There would need to be some pretty major changes to get attribution to work accurately and reliably, and this doesn't even cover the share-alike licensing issues. @bananarama @atatassault @intransitivelie @jlsigman @ben you're the one claiming there isn't attribution...  @martin_piper @bananarama @atatassault @intransitivelie @jlsigman @ben The number of commercial LLMs or generative AIs in general that do attribution of their sources as licenced is currently 0. The entire industry have been able to get away with it for now several years. Do you expect stackexchange to be radically different in a positive way and not communicate about it? @atatassault It's useful, but it's a stretch to call it a brain analogue. AI that is useful, usually is useful because it is different from how our brains work, even if the algorithm fell out of an AI lab while thinking about brains. |
@atatassault @jlsigman @martin_piper @ben
"useful brain analogues" is doing a heck of a lot of heavy lifting there. I'd almost say it's a load-bearing error.