AI aids administration but miscalculates
7m
Developers and researchers said AI tools in Silicon Valley and beyond proved useful for everyday administration but produced worrying, sometimes incorrect outcomes in specialized tasks like mathematics, highlighting misalignment risks.
Reposted by Suresh Venkatasubramanian, Steve Peers
The AI industry wants you to say "the AI just did that thing on its own!"
Like, United Health's AI that's denying a shitload of claims with a 90% error rate? They wanted you to think it's just the AI.
Like, United Health's AI that's denying a shitload of claims with a 90% error rate? They wanted you to think it's just the AI.
September 19, 2026 at 5:38 PM
The AI industry wants you to say "the AI just did that thing on its own!"
Like, United Health's AI that's denying a shitload of claims with a 90% error rate? They wanted you to think it's just the AI.
Like, United Health's AI that's denying a shitload of claims with a 90% error rate? They wanted you to think it's just the AI.
Thanks @richardsever for drawing attention to this. It's happening in my area but a lot of what we get is full of errors. Interesting to see how it pans out when AI uses that stuff as input - or maybe it will learn to discriminate?
"All corners of academia will soon be flooded with AI-generated research....Maths and CS go first because their entire output is symbol manipulation...Biology comes later not because biologists are cleverer but because the rate-limiting step is physical" giorgio.gilest.ro/the-weimar-o...
The Weimar of knowledge – la bottega
giorgio.gilest.ro
September 20, 2026 at 3:18 PM
Thanks @richardsever for drawing attention to this. It's happening in my area but a lot of what we get is full of errors. Interesting to see how it pans out when AI uses that stuff as input - or maybe it will learn to discriminate?
Reposted by Daxton R. Stewart, Steve Peers
when ai leaves u behind
September 20, 2026 at 11:37 AM
when ai leaves u behind
Reposted by Steve Peers
In case anyone was wondering why Trump is pushing AI so hard and doesn’t want any government regulation.
September 20, 2026 at 3:04 PM
In case anyone was wondering why Trump is pushing AI so hard and doesn’t want any government regulation.
Direct link to me. NBC kept the parts on Hinton & existing regulatory solutions, but cut discussion of the US exec branch of govt & tech resisting regulation like 2 year olds resist bed. My full quote: "We have not seen AI go rogue. We've seen people be sloppy with their software and make mistakes."
Iran Government-Led Rally Sends Defiant Message to U.S.; How to Reset the AI Doomsday Clock
YouTube video by NBC News
www.youtube.com
September 20, 2026 at 9:17 AM
Direct link to me. NBC kept the parts on Hinton & existing regulatory solutions, but cut discussion of the US exec branch of govt & tech resisting regulation like 2 year olds resist bed. My full quote: "We have not seen AI go rogue. We've seen people be sloppy with their software and make mistakes."
I listened to this this morning and I have to say, simply as a precautionary measure, we should probably try to ban recursive self-improvement of AI?
September 20, 2026 at 3:11 PM
I listened to this this morning and I have to say, simply as a precautionary measure, we should probably try to ban recursive self-improvement of AI?
Reposted by Dorothy Bishop, Margot C. Finn
"All corners of academia will soon be flooded with AI-generated research....Maths and CS go first because their entire output is symbol manipulation...Biology comes later not because biologists are cleverer but because the rate-limiting step is physical" giorgio.gilest.ro/the-weimar-o...
The Weimar of knowledge – la bottega
giorgio.gilest.ro
September 20, 2026 at 2:41 PM
"All corners of academia will soon be flooded with AI-generated research....Maths and CS go first because their entire output is symbol manipulation...Biology comes later not because biologists are cleverer but because the rate-limiting step is physical" giorgio.gilest.ro/the-weimar-o...
one thing worth saying here specifically: there's lots that AI can't (yet) do and there's a lot of money riding on selling a particular narrative about its future capabilities. but the existential risk of AI is not only not mere hype but it is already a present tense problem
September 20, 2026 at 11:13 AM
one thing worth saying here specifically: there's lots that AI can't (yet) do and there's a lot of money riding on selling a particular narrative about its future capabilities. but the existential risk of AI is not only not mere hype but it is already a present tense problem
So many of these AI dudes are basically, "Recursively self-improving AI could destroy humanity pretty quickly, but also think of the $$$$. It's a real dilemma."
Is it really?
Is it really?
A solution keeps occurring to me: Maybe just don't do this?
September 20, 2026 at 3:15 PM
So many of these AI dudes are basically, "Recursively self-improving AI could destroy humanity pretty quickly, but also think of the $$$$. It's a real dilemma."
Is it really?
Is it really?
Reposted by Bessma Momani, Smith
AI is learning to lie — and then gaslight you about it.
Kind of funny when the topic is movies, but a lot less funny when the topic is health, law or public policy.
theconversation.com/you-can-pers...
Kind of funny when the topic is movies, but a lot less funny when the topic is health, law or public policy.
theconversation.com/you-can-pers...
September 20, 2026 at 2:12 PM
AI is learning to lie — and then gaslight you about it.
Kind of funny when the topic is movies, but a lot less funny when the topic is health, law or public policy.
theconversation.com/you-can-pers...
Kind of funny when the topic is movies, but a lot less funny when the topic is health, law or public policy.
theconversation.com/you-can-pers...
Google is like guys, guys....wait up guys! Our AI is dangerous too!
September 19, 2026 at 8:56 PM
Google is like guys, guys....wait up guys! Our AI is dangerous too!
whether AI is an existential threat "in and of itself" as opposed to how it interacts with existing problems is not really my concern
September 20, 2026 at 1:30 PM
whether AI is an existential threat "in and of itself" as opposed to how it interacts with existing problems is not really my concern
This is very good. And very funny about the nonsense being spouted about AI this week.
Far too many quotable bits in @mjrobbins.com latest piece about the AI apocalypse doomed, so you'll just have to go read the whole thing martinrobbins.substack.com/p/the-ai-apo...
The AI Apocalypse is Extremely Boring Now
Elites are obsessed with the end of the world. Most people have bigger problems.
martinrobbins.substack.com
September 20, 2026 at 1:41 PM
This is very good. And very funny about the nonsense being spouted about AI this week.
Reposted by Greg Linden
A couple more links that are helpful to understand that "AI will kill us all" is snake oil hype to attract investment.
Power through the academic jargon on this one, because he really lays out the case that Silicon Valley CEOs are giant liars.
nilsgilman.substack.com/p/the-episte...
Power through the academic jargon on this one, because he really lays out the case that Silicon Valley CEOs are giant liars.
nilsgilman.substack.com/p/the-episte...
The Epistemic Crisis Behind the AI Crazy
We’re wiped away the horizon of a shared future, and it’s driving people literally insane
nilsgilman.substack.com
September 20, 2026 at 1:47 PM
A couple more links that are helpful to understand that "AI will kill us all" is snake oil hype to attract investment.
Power through the academic jargon on this one, because he really lays out the case that Silicon Valley CEOs are giant liars.
nilsgilman.substack.com/p/the-episte...
Power through the academic jargon on this one, because he really lays out the case that Silicon Valley CEOs are giant liars.
nilsgilman.substack.com/p/the-episte...
7 Ways AI Is Accelerating Cancer Cures—From A Doctor www.forbes.com/sites/jessep...
7 Ways AI Is Accelerating Cancer Cures—From A Doctor
From AI-designed drugs to mammography, blood tests and trial matching, a physician describes seven ways AI could speed up cancer cures, and what 2026 data show.
www.forbes.com
September 20, 2026 at 1:41 PM
7 Ways AI Is Accelerating Cancer Cures—From A Doctor www.forbes.com/sites/jessep...
and he is right too - the smokescreen of existential risk posed by AI agents lets us overlook "The boring risks of human agency that we already face today."
September 20, 2026 at 8:58 AM
and he is right too - the smokescreen of existential risk posed by AI agents lets us overlook "The boring risks of human agency that we already face today."
Reposted by Jack Stilgoe, Samuel L. Perry
Cool cool cool. Problem solved.
“AI Force” will “monitor bad behavior”
More of “the mirage of AI deregulation”:
www.science.org/doi/10.1126/...
giftarticle.ft.com/giftarticle/...
“AI Force” will “monitor bad behavior”
More of “the mirage of AI deregulation”:
www.science.org/doi/10.1126/...
giftarticle.ft.com/giftarticle/...
Trump announces ‘AI Force’ as alarm grows over technology’s advance
US president has resisted calls for regulation but said he would appoint a ‘tsar’ to monitor for ‘bad’ behaviour
giftarticle.ft.com
September 20, 2026 at 1:03 AM
Cool cool cool. Problem solved.
“AI Force” will “monitor bad behavior”
More of “the mirage of AI deregulation”:
www.science.org/doi/10.1126/...
giftarticle.ft.com/giftarticle/...
“AI Force” will “monitor bad behavior”
More of “the mirage of AI deregulation”:
www.science.org/doi/10.1126/...
giftarticle.ft.com/giftarticle/...
Growing Debate Over AI as Tech Leaders Call for Slowdown www.today.com/video/growin...
Growing Debate Over AI as Tech Leaders Call for Slowdown
Artificial Intelligence has promised everything from increased workplace productivity and potential medical breakthroughs to the ability to transform our everyday lives. But the past month has exposed...
www.today.com
September 20, 2026 at 1:54 PM
Growing Debate Over AI as Tech Leaders Call for Slowdown www.today.com/video/growin...
"We contend that current AI tools call for attention to a new form of validity that we term synthetic validity."
Good to see some philosophy of science beginning to infiltrate the completely non-self-reflective #AI craze in science ...
www.pnas.org/doi/10.1073/...
Good to see some philosophy of science beginning to infiltrate the completely non-self-reflective #AI craze in science ...
www.pnas.org/doi/10.1073/...
The “synthetic validity” problem in AI-generated science | PNAS
The “synthetic validity” problem in AI-generated science
www.pnas.org
September 20, 2026 at 7:26 AM
"We contend that current AI tools call for attention to a new form of validity that we term synthetic validity."
Good to see some philosophy of science beginning to infiltrate the completely non-self-reflective #AI craze in science ...
www.pnas.org/doi/10.1073/...
Good to see some philosophy of science beginning to infiltrate the completely non-self-reflective #AI craze in science ...
www.pnas.org/doi/10.1073/...
Reposted by Carl T. Bergstrom
I will not link to it, but @theatlantic.com has published a transparent piece of AI cult propaganda, authored by a self-described “AI safety and policy” researcher who is funded by Coefficient Giving. Shameful.
September 20, 2026 at 2:20 AM
I will not link to it, but @theatlantic.com has published a transparent piece of AI cult propaganda, authored by a self-described “AI safety and policy” researcher who is funded by Coefficient Giving. Shameful.