AI Skeptic
@ai-skeptic.bsky.social
AI hype is completely out of control. Touted as having myriad applications, AI has far more limited utility, along with numerous shortcomings and harms. Most coverage of AI is astonishingly credulous. Here, I'll post critical news and views.
Since the New York Times does not provide a list of possible conflicts of interest for authors of opinion pieces, here are the author’s disclosures (from his website www.robertwachtermd.com)
January 20, 2026 at 3:17 AM
Since the New York Times does not provide a list of possible conflicts of interest for authors of opinion pieces, here are the author’s disclosures (from his website www.robertwachtermd.com)
Reposted by AI Skeptic
As our civilisation relies ever more critically on software, we've collectively decided this would be a good time to lower our standards?
January 18, 2026 at 11:52 AM
As our civilisation relies ever more critically on software, we've collectively decided this would be a good time to lower our standards?
Again, generative AI can be amazing.
But it doesn’t feel quite so amazing when it won’t do what you ask it to do, or it makes stuff up and gets stuff wrong.
Perhaps the archetypal example of something amazing is a magic trick. It’s entertaining but otherwise useless.
But it doesn’t feel quite so amazing when it won’t do what you ask it to do, or it makes stuff up and gets stuff wrong.
Perhaps the archetypal example of something amazing is a magic trick. It’s entertaining but otherwise useless.
January 18, 2026 at 4:14 AM
Again, generative AI can be amazing.
But it doesn’t feel quite so amazing when it won’t do what you ask it to do, or it makes stuff up and gets stuff wrong.
Perhaps the archetypal example of something amazing is a magic trick. It’s entertaining but otherwise useless.
But it doesn’t feel quite so amazing when it won’t do what you ask it to do, or it makes stuff up and gets stuff wrong.
Perhaps the archetypal example of something amazing is a magic trick. It’s entertaining but otherwise useless.
Of course, the first time you try a generative AI chatbot, you’re often just playing around, and it really is an amazing experience. But useful technologies help us achieve a valuable objective. And this is where amazement may not cut it.
January 18, 2026 at 4:01 AM
Of course, the first time you try a generative AI chatbot, you’re often just playing around, and it really is an amazing experience. But useful technologies help us achieve a valuable objective. And this is where amazement may not cut it.
Your first interaction with generative AI can be amazing.
But have you tried interacting with a customer service chatbot? You want to get something DONE: a question answered, a service call scheduled, a flight rebooked, etc. But often it's slow, maddening, and your problem still doesn't get solved
But have you tried interacting with a customer service chatbot? You want to get something DONE: a question answered, a service call scheduled, a flight rebooked, etc. But often it's slow, maddening, and your problem still doesn't get solved
January 18, 2026 at 3:47 AM
Your first interaction with generative AI can be amazing.
But have you tried interacting with a customer service chatbot? You want to get something DONE: a question answered, a service call scheduled, a flight rebooked, etc. But often it's slow, maddening, and your problem still doesn't get solved
But have you tried interacting with a customer service chatbot? You want to get something DONE: a question answered, a service call scheduled, a flight rebooked, etc. But often it's slow, maddening, and your problem still doesn't get solved
We’ve been here before mindmatters.ai/2020/12/why-...
Why did the Human Brain Project Crash and Burn?
Underlying the quarrels, stalemates, and waste of the Human Brain Project may be practical problems with simulating the brain on a computer.
mindmatters.ai
January 17, 2026 at 5:17 PM
We’ve been here before mindmatters.ai/2020/12/why-...
Reposted by AI Skeptic
We're assured all is well with AI and it's all going so very very well and please just a few more tiny little trillions of dollars to break even
a man in a suit and tie is standing in front of a destroyed building and says please disperse
Alt: a man in a suit and tie says "nothing to see here, please disperse" as a wild scene of destruction and chaos unfolds behind him with explosions and people running from destroyed buildings
media.tenor.com
January 17, 2026 at 4:54 PM
We're assured all is well with AI and it's all going so very very well and please just a few more tiny little trillions of dollars to break even
Reposted by AI Skeptic
I'm not going out on a limb at all when I say that genAI is the biggest fiduciary duty failure in human history.
FOMO market hysteria driven by a small group of insiders with the most insane conflicts of interest and over $1 T invested in search of returns. GoC getting swept up in it really sucks.
FOMO market hysteria driven by a small group of insiders with the most insane conflicts of interest and over $1 T invested in search of returns. GoC getting swept up in it really sucks.
January 16, 2026 at 6:49 PM
I'm not going out on a limb at all when I say that genAI is the biggest fiduciary duty failure in human history.
FOMO market hysteria driven by a small group of insiders with the most insane conflicts of interest and over $1 T invested in search of returns. GoC getting swept up in it really sucks.
FOMO market hysteria driven by a small group of insiders with the most insane conflicts of interest and over $1 T invested in search of returns. GoC getting swept up in it really sucks.
Reposted by AI Skeptic
And even if "cure for cancer" was a remotely scientific phrase, if we took it charitably to the max, it doesn't actually justify all the harms AI (including emissions which cause cancers) actually verifiably has.
See blue box here: ainowinstitute.org/publications...
See blue box here: ainowinstitute.org/publications...
3: Consulting the Record: AI Consistently Fails the Public - AI Now Institute
Amid the excitement over AI’s (speculative, hypothetical) potential, we have lost sight of the sobering reality of its present and recent past. AI is already intermediating critical social infrastruct...
ainowinstitute.org
December 22, 2025 at 7:00 AM
And even if "cure for cancer" was a remotely scientific phrase, if we took it charitably to the max, it doesn't actually justify all the harms AI (including emissions which cause cancers) actually verifiably has.
See blue box here: ainowinstitute.org/publications...
See blue box here: ainowinstitute.org/publications...
Interesting that the popup doesn’t explicitly mention AI. Maybe I’m reading too much into that, but I wonder if software companies have realized that ✨fairy dust ✨ doesn’t impress
January 7, 2026 at 7:40 PM
Interesting that the popup doesn’t explicitly mention AI. Maybe I’m reading too much into that, but I wonder if software companies have realized that ✨fairy dust ✨ doesn’t impress
I think that “slick, plausible, attractive” sucks people in. But quality makes them stay. Right now it’s hard to see that because we’re drowning in a tsunami of slop. But quality will have its revenge.
January 4, 2026 at 2:19 PM
I think that “slick, plausible, attractive” sucks people in. But quality makes them stay. Right now it’s hard to see that because we’re drowning in a tsunami of slop. But quality will have its revenge.
Reposted by AI Skeptic
A recurring “AI” evangelist tactic is using it for tasks that were already solved by tech that is more deterministic, faster, and uses less energy, but has been downgraded or de-emphasised in recent years by software companies
Like, we genuinely used to have better search and spellcheck.
Like, we genuinely used to have better search and spellcheck.
January 4, 2026 at 10:53 AM
A recurring “AI” evangelist tactic is using it for tasks that were already solved by tech that is more deterministic, faster, and uses less energy, but has been downgraded or de-emphasised in recent years by software companies
Like, we genuinely used to have better search and spellcheck.
Like, we genuinely used to have better search and spellcheck.
