Most German firms adopt AI
5 months ago
A majority of German companies had adopted AI in the workplace and about one-third of Germans used it weekly, while employers provided relatively little staff training, raising dependency concerns.
Write your reply
by Lamar Pierce
The @orgscience.bsky.social AI taskforce released an editorial analyzing 6 yrs of AI use in submission/review data: AI-written papers are lower quality, nearly all rejected, and explain 42%⬆️ submits. We're seeing more rather than better research. Humans are reviewing machines & we're getting tired.
pubsonline.informs.org
April 28, 2026 at 12:49 AM
The @orgscience.bsky.social AI taskforce released an editorial analyzing 6 yrs of AI use in submission/review data: AI-written papers are lower quality, nearly all rejected, and explain 42%⬆️ submits. We're seeing more rather than better research. Humans are reviewing machines & we're getting tired.
Reposted by Martin Paul Eve, Andreas Wagner
Here is an AI trained just using text from 1931 or earlier, which leads to a lot of interesting experiments: can the model independently develop later inventions? Can it learn to code from examples alone?
You can talk to the model here: talkie-lm.com/chat
Details here: talkie-lm.com/introducing-...
You can talk to the model here: talkie-lm.com/chat
Details here: talkie-lm.com/introducing-...
April 28, 2026 at 12:45 AM
Here is an AI trained just using text from 1931 or earlier, which leads to a lot of interesting experiments: can the model independently develop later inventions? Can it learn to code from examples alone?
You can talk to the model here: talkie-lm.com/chat
Details here: talkie-lm.com/introducing-...
You can talk to the model here: talkie-lm.com/chat
Details here: talkie-lm.com/introducing-...
Reposted by Greg Linden
We'll find out exactly how "useful" AI is once companies and users are forced to pay for their actual token burn. When AI costs 10% of your headcount, and may reach 100%, flimsy claims of boosted productivity spouted by boosters and CTOs go quiet.
www.wheresyoured.at/ais-economics-dont-make-sense/
www.wheresyoured.at/ais-economics-dont-make-sense/
April 28, 2026 at 4:59 PM
We'll find out exactly how "useful" AI is once companies and users are forced to pay for their actual token burn. When AI costs 10% of your headcount, and may reach 100%, flimsy claims of boosted productivity spouted by boosters and CTOs go quiet.
www.wheresyoured.at/ais-economics-dont-make-sense/
www.wheresyoured.at/ais-economics-dont-make-sense/
'AI will be at the core of some exercises. For example, politics students in a course on voting behaviour will be told to choose an election and ask ChatGPT to explain the results.' 1/3
Surrey redesigns all degrees to embed ‘discipline-specific’ AI
Students expected to use AI for information before evaluating technology’s output as institution seeks to prepare learners for changing labour market
www.timeshighereducation.com
April 28, 2026 at 6:11 AM
'AI will be at the core of some exercises. For example, politics students in a course on voting behaviour will be told to choose an election and ask ChatGPT to explain the results.' 1/3
by Lamar Pierce
We hope this provides data and evidence to help us all understand how the process of creating, evaluating, and promoting great research must evolve. We are AI enthusiasts and heavy users, so believe that it can help us achieve better research. But incentives must change to reward better, not more.
April 28, 2026 at 12:49 AM
We hope this provides data and evidence to help us all understand how the process of creating, evaluating, and promoting great research must evolve. We are AI enthusiasts and heavy users, so believe that it can help us achieve better research. But incentives must change to reward better, not more.
I wonder if that feeling will wear off as AI-bots become more common.
This raises a lot of questions for me about tutoring using AI-bots.
Curious about other folks' experiences and views.
3/3
This raises a lot of questions for me about tutoring using AI-bots.
Curious about other folks' experiences and views.
3/3
April 27, 2026 at 4:40 PM
I wonder if that feeling will wear off as AI-bots become more common.
This raises a lot of questions for me about tutoring using AI-bots.
Curious about other folks' experiences and views.
3/3
This raises a lot of questions for me about tutoring using AI-bots.
Curious about other folks' experiences and views.
3/3
Reposted by Meredith Broussard
“AI systems cannot reliably explain their own reasoning. While they produce plausible legal arguments, it’s not clear that the reasons are the motivations for their actual decisions”
news.uchicago.edu/story/how-do...
news.uchicago.edu/story/how-do...
How do AI 'judges' compare to human ones? It's complicated, says a UChicago scholar
In annual Ryerson Lecture, legal scholar Eric Posner examines AI's growing role in legal decision-making—and why human judgment still matters
news.uchicago.edu
April 28, 2026 at 2:24 AM
“AI systems cannot reliably explain their own reasoning. While they produce plausible legal arguments, it’s not clear that the reasons are the motivations for their actual decisions”
news.uchicago.edu/story/how-do...
news.uchicago.edu/story/how-do...
I think this post is spot on and I worry about the current students/recent grads who aren't getting a chance to build up the skills they'll need because critical parts of their development have coincided with the AI bubble.
Everyone keeps saying "AI isn't going away" and no, "AI" as you've come to know it is in fact going to go away.
It's unprofitable and the user experience to date has been subsidized. It will never scale, and it's not getting any cheaper.
And localized fragments of that tech aren't the same.
It's unprofitable and the user experience to date has been subsidized. It will never scale, and it's not getting any cheaper.
And localized fragments of that tech aren't the same.
April 28, 2026 at 8:29 PM
I think this post is spot on and I worry about the current students/recent grads who aren't getting a chance to build up the skills they'll need because critical parts of their development have coincided with the AI bubble.
Did you see the AI-first masters degree being run out of a cupboard with all AI generated course materials which OfStud has just OKed? Report is pure fantasy
April 28, 2026 at 9:45 AM
Did you see the AI-first masters degree being run out of a cupboard with all AI generated course materials which OfStud has just OKed? Report is pure fantasy
Reposted by Joanna Bryson, Hilary J. Allen
Everyone keeps saying "AI isn't going away" and no, "AI" as you've come to know it is in fact going to go away.
It's unprofitable and the user experience to date has been subsidized. It will never scale, and it's not getting any cheaper.
And localized fragments of that tech aren't the same.
It's unprofitable and the user experience to date has been subsidized. It will never scale, and it's not getting any cheaper.
And localized fragments of that tech aren't the same.
April 28, 2026 at 5:23 AM
Everyone keeps saying "AI isn't going away" and no, "AI" as you've come to know it is in fact going to go away.
It's unprofitable and the user experience to date has been subsidized. It will never scale, and it's not getting any cheaper.
And localized fragments of that tech aren't the same.
It's unprofitable and the user experience to date has been subsidized. It will never scale, and it's not getting any cheaper.
And localized fragments of that tech aren't the same.
'If graduates are asked to supervise AI before having learnt the standards of the field, the result may be fluency without competence.'
'This is why the way universities deal with the disruptive force of AI is so essential.'
'This is why the way universities deal with the disruptive force of AI is so essential.'
AI literacy is not enough – universities must teach through disciplinary standards
The University of Surrey will have embedded AI use into all of its degree programmes by September. Annika Bautz explains why doing so in a discipline-specific way was so crucial The University of Surr...
wonkhe.com
April 28, 2026 at 6:20 AM
'If graduates are asked to supervise AI before having learnt the standards of the field, the result may be fluency without competence.'
'This is why the way universities deal with the disruptive force of AI is so essential.'
'This is why the way universities deal with the disruptive force of AI is so essential.'
AI is getting real good.
April 28, 2026 at 3:47 AM
AI is getting real good.
by Ken Caldeira
I am never quite sure what is AI slop and what is the result of someone's work.
April 27, 2026 at 6:22 PM
I am never quite sure what is AI slop and what is the result of someone's work.