For those unaware, Linus Torvalds this year has deeply cast his lot with the hyperscale LLMs for development of Linux code, dismissing arguments that this is an abdication of responsibility over that code base.
For those unaware, Linus Torvalds this year has deeply cast his lot with the hyperscale LLMs for development of Linux code, dismissing arguments that this is an abdication of responsibility over that code base.
Again: LLM’s are tools. If used right, tools are great, and LLM’s CAN be great for certain tasks. As far as I can tell, that is Torvald’s position as well; use tools as tools.
hate this reductionist phrase trying to frame llms as neutral tech and the premise that you just have to learn to “use it right”.
this tech sucks up power like crazy, devalues human knowledge and art, deskills workers on a massive scale, aids in global surveillance, demands data centers that displace people, spits out code that violates the licensing of a huge number of projects, drives hardware prices through the roof. then it statistically generates some plausible-looking code and gives you a little productivity boost followed by a technical debt that we’ll just kick down the line and deal with later.
in the current version they’re pretty unethical on every grounds. free software devs, espicially those with gpl’d projects, should know it better than anyone.
any other “tools” that do that?
I agree with his position, but he completely skips the moral issues using current LLMs, which is the big problem here.
LLMd are not tools. A hammer is a tool. A stochastic appliance can not be a tool.