Gå til indhold

Ted Underwood

@tedunderwood.com did:plc:565ebob5f6hw33hjdkxty6qj

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

Tilmeldte sig Bluesky

Følgere
23.828
Følger
6.440
Opslag
25.701
Dage på ranglister
0

Oversigt

Ted Underwood (@tedunderwood.com) er en Bluesky-bruger, der tilmeldte sig 10 april 2023. Kontoen har 23.828 følgere, følger 6.440 konti og har udgivet 25.701 opslag.

Optrædener på ranglister

De bedste dage, hvor denne kontos opslag kom på Debyas daglige topliste.

Denne konto har endnu ikke optrådt på en daglig rangliste.

Nyligt likede opslag

Offentlige likes fra Bluesky-netværket.

  • LP@lloydphilip.bsky.social

    Agreed. But most hear about Frontier models and think it’s the easy button. Aside from coding, which I don’t use LLMs for, it’s a lot of work for me to get the messaging right when I need to develop White papers or PPTX for clients and executives. It takes a long time to get the messaging right.

    Vis opslag
  • Peter Bloem@pbloem.sigmoid.social.ap.brid.gy

    @tedunderwood.com A cynic might say that LLM use would not make the _average_ academic article any worse.

    Vis opslag
  • Paul Hünermund@p-hunermund.com

    But at least in college essays, there was a lot of text produced that wasn't worth our time before AI either. So LLMs do bring a real quality improvement here in terms of style and efficiency (unless you're teaching creative writing).

    Vis opslag
  • Paul Hünermund@p-hunermund.com

    What we want to discourage is AI slop that sounds good on the surface but meanders on without ever making a point. But that's not something tools like Pangram can detect, as Jessica points out in her post.

    Vis opslag

Log ind for at se mere

Konti, de blokerer

Blokeringer på Bluesky er offentlige poster.