Busy weeks, and not nearly enough time to complete any of the drafts for the blog. In the meantime, at least a couple of links to interesting things I’ve read lately.
-
Arvind Narayanan and Sayash Kapoor with two excellent essays over at AI Snake Oil: AI as Normal Technology and AGI is not a milestone. Particularly the former, where they argue that AI is a normal technology is worth your time. Their argument is centered around how AI will have an effect on society, not the technology itself. It’s in that sense, the adoption, that AI is normal.
-
Related is The Misunderstood Role of AI Skeptics which divide “AI Skeptics” into two broad categories.
-
Nathan Lambert, at Allen AI (and with that also perhaps somewhat biased in his opinions given their work on the Olmo model), has published a short post on LLMs with open weights from an implementation perspective.
-
Human Learning in the Age of Machine Learning explores why why educate ourselves, and where education is heading.
-
Moderation.txt explains why “there is no such thing as ‘no politics’.
-
My tips on giving technical talks have a few really good suggestions. Not the least about what language one choose to use on stage.