256 shaares
21 results
tagged
code
"We are not here to discuss "open weight" models; these are still built on a foundation of companies destructively mining the web from data centers that wreck communities. Such models cannot exist without exploitation. When a model can be fully trained by end users using data that was collected with consent, then we can talk about that, but right now that is nothing but science fiction and speculation."
"Sure, LLMs are bad at some things, but you can't deny that they're useful for programming!" Sound familiar? In this week's episode, Emily and Alex break down the key myths around AI-boosted productivity in tech. Plus, Alex previews her work with DAIR's newly launched Luddite Lab, where workers are organizing against automation.
Les promoteurs des agents de codage IA réinventent le "Software Process Engineering" sans même le savoir.
"Nobody in history has ever described software development in machine-executable detail and had it actually work at scale. Sure, you might be the first. Anything’s possible."
Anthropic’s guide on context engineering formulates it quite clearly: the goal is to find the smallest possible set of high-signal tokens that maximizes the probability of obtaining the desired result.
