Ted Underwood
banner
tedunderwood.com
Ted Underwood
@tedunderwood.com
Uses machine learning to study literary imagination, and vice-versa. Likely to share news about AI & computational social science / Sozialwissenschaft / 社会科学

Information Sciences and English, UIUC. Distant Horizons (Chicago, 2019). tedunderwood.com
I went to do this and discovered I had already done it
January 11, 2026 at 4:22 PM
Let’s hope you’re right. The sweet spot for me is only modestly larger, with most of the growth international / non-Anglophone
January 11, 2026 at 4:09 PM
I guess I’m saying “gentlemen, it has been a privilege playing with you tonight,” except in reverse: our tipped-deck quartet is over because the ship is righting itself while lifeboats converge on it and masses of people clamber up its sides
January 11, 2026 at 3:47 PM
Trident-and-net sounds cool, but I'm not sure you get a weapon when you're thrown to the lions
January 10, 2026 at 11:00 PM
Yeah. Honestly "10% increase" in productivity would be amazing and is usually an unrealistic promise.
January 10, 2026 at 10:06 PM
verily
Imagine you time travel back to 1847 and you find the left response to industrialization is a) machines will never be as good as human weavers or b) we need to copyright loom patterns or c) it’s a speculative bubble.

You’d say “Y’all. Not helping. What you need is obviously a labor movement.”
January 10, 2026 at 8:32 PM
Reposted by Ted Underwood
a thing i've been thinking about posting for a long time is "back in those days, the workers didn't say "eww factories were created by capitalists to enrich capitalists, we should abandon factories" they said "we should seize them and direct their output for the good of humanity"
January 10, 2026 at 7:07 PM
Yes, I was muttering about that yesterday, and I agree that’s the opportunity
January 10, 2026 at 1:26 PM
This is why I wrote the last thing about “transparency” actually meaning restriction.

This is the form it takes: less powerful than CC but absolutely self-contained:
Gemini integration is not there yet, but it's going to be possible to have a Claude-Code-level agentic assistant in your email. Systematically cross-referencing inquiries with replies, asking clarification questions, remembering preferences, generating to-do lists. Absolutely fatal app.
January 10, 2026 at 1:24 PM
Yes. It’s a “stone soup” situation. You don’t need to know how to code; this stone makes the whole soup. But sure, the soup will be a little better if you understand dependencies …
January 10, 2026 at 1:22 PM
(Where “more transparent” may actually mean “more restricted” — in the same way the point of the “chat” interface was to channel shoggothy language into a comprehensible subgenre that looked like conversation.)
January 10, 2026 at 1:14 PM
(At which point, their lunacy became a widespread problem.)

CC-like tools are going to be more impactful than GPT-3 of course. A whole industry is transforming. But it’s plausible to me that non-coders will not see the point or feel empowered until we get agentic tools that feel more transparent.
January 10, 2026 at 1:08 PM