Mordechai Rorvig
@mordecwhy.bsky.social
Science journalist and writer at Foom Magazine, a new, ad-free, grant-funded website for original reporting on developments in AI safety and ethics.
Reposted by Mordechai Rorvig
Great to see our "From Language to Cognition" work featured in @mordecwhy.bsky.social's latest piece on language models and the brain. Glad to contribute to the conversation!
www.foommagazine.org/language-mod...
www.foommagazine.org/language-mod...
Language models resemble more than just language cortex, show neuroscientists
In a paper presented in November 2025 at the Empirical Methods in Natural Language Processing (EMNLP) conference, researchers at the Swiss Federal Institute of Technology (EPFL), the Massachusetts Ins...
www.foommagazine.org
January 16, 2026 at 7:03 PM
Great to see our "From Language to Cognition" work featured in @mordecwhy.bsky.social's latest piece on language models and the brain. Glad to contribute to the conversation!
www.foommagazine.org/language-mod...
www.foommagazine.org/language-mod...
FOOM / NEW STORY OUT: "The results lend clarity to the surprising picture that has been emerging from the last decade of neuroscience research: That AI programs can show strong resemblances to large-scale brain regions."
www.foommagazine.org/language-mod...
www.foommagazine.org/language-mod...
Language models resemble more than just language cortex, show neuroscientists
In a paper presented in November 2025 at the Empirical Methods in Natural Language Processing (EMNLP) conference, researchers at the Swiss Federal Institute of Technology (EPFL), the Massachusetts Ins...
www.foommagazine.org
January 13, 2026 at 6:19 PM
FOOM / NEW STORY OUT: "The results lend clarity to the surprising picture that has been emerging from the last decade of neuroscience research: That AI programs can show strong resemblances to large-scale brain regions."
www.foommagazine.org/language-mod...
www.foommagazine.org/language-mod...
It often feels like, in a mental health or depression context, whatever it is that is wrong with me is so deeply entrenched that neither I nor anyone else would ever be able to figure out what it is.
December 30, 2025 at 5:34 PM
It often feels like, in a mental health or depression context, whatever it is that is wrong with me is so deeply entrenched that neither I nor anyone else would ever be able to figure out what it is.
In my latest for Foom, where I'm trying to provide free, high-quality, independent reporting on AI safety, I wanted to interview someone who could help me understand the challenging internal dynamics of the community. This was @ilex-ulmus.bsky.social.
www.foommagazine.org/the-moral-cr...
www.foommagazine.org/the-moral-cr...
The moral critic of the AI industry—a Q&A with Holly Elmore
An interview with Holly Elmore, a leading proponent of a non-technical solution to the existential risks of AI technology.
www.foommagazine.org
December 26, 2025 at 5:42 PM
In my latest for Foom, where I'm trying to provide free, high-quality, independent reporting on AI safety, I wanted to interview someone who could help me understand the challenging internal dynamics of the community. This was @ilex-ulmus.bsky.social.
www.foommagazine.org/the-moral-cr...
www.foommagazine.org/the-moral-cr...
"When dealing with the Big Cats, who are literally killing machines, there is always a distinct energy or electricity when you are in their presence." -Leif Cocks.
Probably also the best description of humans
Probably also the best description of humans
December 23, 2025 at 8:26 PM
"When dealing with the Big Cats, who are literally killing machines, there is always a distinct energy or electricity when you are in their presence." -Leif Cocks.
Probably also the best description of humans
Probably also the best description of humans
This kind of statement triggers me and it's the reason why science journalists and neuroscientists need to speak out, loudly, about the analogies discovered between DNNs and cortexes. People need to know we are not just dealing with an MS Word technology here
www.theguardian.com/lifeandstyle...
www.theguardian.com/lifeandstyle...
‘My photos are warm and full of imagination – that’s something AI could never achieve’: Yuan Li’s best phone picture
This spectacular image taken in Sakrisøy, Norway, triggered accusations that it was simply too good to be true
www.theguardian.com
December 13, 2025 at 5:32 PM
This kind of statement triggers me and it's the reason why science journalists and neuroscientists need to speak out, loudly, about the analogies discovered between DNNs and cortexes. People need to know we are not just dealing with an MS Word technology here
www.theguardian.com/lifeandstyle...
www.theguardian.com/lifeandstyle...
Absolutely right. This is as big of a risk right now as anything else. It's flabbergasting how current government thinks anyone is being fooled about this.
Today much of the talk about AI is whether there’s a speculative bubble or whether OpenAI will beat its rivals in building new generations of AI
The vast majority of people couldn’t care less about that—They're worried about AI causing HUGE LAYOFFS www.theguardian.com/commentisfre...
@thefarce.org
The vast majority of people couldn’t care less about that—They're worried about AI causing HUGE LAYOFFS www.theguardian.com/commentisfre...
@thefarce.org
Most people aren’t fretting about an AI bubble. What they fear is mass layoffs | Steven Greenhouse
Artificial intelligence could make income inequality even worse and create a new underclass. Governments and society must take action
www.theguardian.com
December 13, 2025 at 1:00 AM
Absolutely right. This is as big of a risk right now as anything else. It's flabbergasting how current government thinks anyone is being fooled about this.
New article covering recent findings from October: Models that maximize business performance in realistic role-play scenarios are also more likely to inflict harms.
www.foommagazine.org/leading-mode...
www.foommagazine.org/leading-mode...
Leading models take chilling tradeoffs in realistic scenarios, new research finds
Models that maximize business performance in realistic role-play scenarios are also more likely to inflict harms.
www.foommagazine.org
December 12, 2025 at 6:20 PM
New article covering recent findings from October: Models that maximize business performance in realistic role-play scenarios are also more likely to inflict harms.
www.foommagazine.org/leading-mode...
www.foommagazine.org/leading-mode...
What does it mean when the study, the study's reviews, and all the other studies citing the study, which is actually a good and interesting study, all show clear signs of AI writing (without acknowledgement), lol
December 12, 2025 at 2:37 AM
What does it mean when the study, the study's reviews, and all the other studies citing the study, which is actually a good and interesting study, all show clear signs of AI writing (without acknowledgement), lol
Is any alignment research valid if it does not engage with the fact that we are surrounded by highly misaligned technologies, in highly misaligned societies, created by highly misaligned individuals?
December 7, 2025 at 11:44 PM
Is any alignment research valid if it does not engage with the fact that we are surrounded by highly misaligned technologies, in highly misaligned societies, created by highly misaligned individuals?
Yes, BCIs seem in need of stringent regulation probably more than any other technology ever invented. The regulation vacuum for AI does not lend optimism. We are going to need serious public interest advocacy from neuroscientists if this is going to be anything besides severely dystopian
The ethical frontiers of brain-computer interfaces medcitynews.com/2025/12/the-... via @ieeebrain.bsky.social; #BCI #NeuroTech
The Ethical Frontiers of Brain-Computer Interfaces - MedCity News
The challenge is not just to minimize harm, but to make choices that expand human flourishing rather than reduce us to more efficient workers or more predictable consumers.
medcitynews.com
December 4, 2025 at 9:00 PM
Yes, BCIs seem in need of stringent regulation probably more than any other technology ever invented. The regulation vacuum for AI does not lend optimism. We are going to need serious public interest advocacy from neuroscientists if this is going to be anything besides severely dystopian
It's really strange to me how so many researchers in fields I follow have their names on 20-30 papers a year. I realize much of this comes down to co-authorship, but even at that level, it's a mind boggling degree of networking or credit sharing. (Not questioning the work, just the social norms)
December 4, 2025 at 6:43 PM
It's really strange to me how so many researchers in fields I follow have their names on 20-30 papers a year. I realize much of this comes down to co-authorship, but even at that level, it's a mind boggling degree of networking or credit sharing. (Not questioning the work, just the social norms)
Reposted by Mordechai Rorvig
2/ This week, we need your help!
Take action at NewsNotSlop.org and follow the steps below, because holding media companies accountable on AI is going to take all of us.
Take action at NewsNotSlop.org and follow the steps below, because holding media companies accountable on AI is going to take all of us.
December 1, 2025 at 5:44 PM
2/ This week, we need your help!
Take action at NewsNotSlop.org and follow the steps below, because holding media companies accountable on AI is going to take all of us.
Take action at NewsNotSlop.org and follow the steps below, because holding media companies accountable on AI is going to take all of us.
WSJ editorial board giving absolutely disgusting support for undemocratic regime change; cancel your subscriptions. "If Maduro refuses to leave, and Trump shrinks from acting to depose him, Trump and the credibility of the US will be the losers.”
www.theguardian.com/world/2025/d...
www.theguardian.com/world/2025/d...
Trump reportedly gave Maduro ultimatum to relinquish power in Venezuela
US president sent a ‘blunt message’ to his South American counterpart, sources say
www.theguardian.com
December 1, 2025 at 6:52 PM
WSJ editorial board giving absolutely disgusting support for undemocratic regime change; cancel your subscriptions. "If Maduro refuses to leave, and Trump shrinks from acting to depose him, Trump and the credibility of the US will be the losers.”
www.theguardian.com/world/2025/d...
www.theguardian.com/world/2025/d...
What did it mean, in 2021, when researchers began to see strange shapes on the insides of their models? We now have a much more sophisticated understanding—models may be thinking in terms of shapes precisely because some concepts are intrinsically geometric.
www.foommagazine.org/scientists-m...
www.foommagazine.org/scientists-m...
Scientists make sense of shapes in the minds of the models
It was at least since 2021, according to the authors of a preprint from March, that researchers began to see something interesting on the insides of their models.
Also known as an AI program, created...
www.foommagazine.org
November 29, 2025 at 4:21 PM
What did it mean, in 2021, when researchers began to see strange shapes on the insides of their models? We now have a much more sophisticated understanding—models may be thinking in terms of shapes precisely because some concepts are intrinsically geometric.
www.foommagazine.org/scientists-m...
www.foommagazine.org/scientists-m...
A major issue I find with engagement-optimized feeds is the way they lull you into a sense of complacency. 'Here's some nice things to make you feel better,' they seem to say; but when you're not ruthlessly intentional with your time, you find yourself hours later gasping for air; emotions exploited
November 20, 2025 at 2:42 PM
A major issue I find with engagement-optimized feeds is the way they lull you into a sense of complacency. 'Here's some nice things to make you feel better,' they seem to say; but when you're not ruthlessly intentional with your time, you find yourself hours later gasping for air; emotions exploited
This may seem like an obvious observation, but I went to the dentist and wow, my teeth are looking a lot less yellow now. Too much coffee and tea for me, I suppose.
November 18, 2025 at 1:59 AM
This may seem like an obvious observation, but I went to the dentist and wow, my teeth are looking a lot less yellow now. Too much coffee and tea for me, I suppose.
I pledge to be transparent about the use of AI—Will you join me? A brief but important essay I've written for my new website.
www.foommagazine.org/i-pledge-to-...
www.foommagazine.org/i-pledge-to-...
I pledge to be transparent about the use of AI—Will you join me?
I work as a science journalist. For an article of mine to have any value, readers need to know they can trust it. They need to know it was written by me and not by some other person or for some other ...
www.foommagazine.org
November 14, 2025 at 4:05 PM
I pledge to be transparent about the use of AI—Will you join me? A brief but important essay I've written for my new website.
www.foommagazine.org/i-pledge-to-...
www.foommagazine.org/i-pledge-to-...
I've been writing with a lot of typos, lately, because I turned off AI on all my devices; I don't want corporate machines reading and training on everything I write. This has made for some sloppy mistakes, but it also shows humanness.
Which makes me think, maybe we should just .. bring typos back.
Which makes me think, maybe we should just .. bring typos back.
November 12, 2025 at 7:27 PM
I've been writing with a lot of typos, lately, because I turned off AI on all my devices; I don't want corporate machines reading and training on everything I write. This has made for some sloppy mistakes, but it also shows humanness.
Which makes me think, maybe we should just .. bring typos back.
Which makes me think, maybe we should just .. bring typos back.
Amazing story by Pamela Gordon in the @theguardian.com, who documented the life of a young man living on the streets of England from the age of 13. I fear that in the US, Craig would have had an even much harder time, if that could be believed.
www.theguardian.com/news/2025/no...
www.theguardian.com/news/2025/no...
When I met Craig he was 13 and homeless. I still thought his life might turn around. I was tragically wrong
The long read: I knew he was running away from something. It wasn’t until many years later that I discovered the truth
www.theguardian.com
November 11, 2025 at 4:23 PM
Amazing story by Pamela Gordon in the @theguardian.com, who documented the life of a young man living on the streets of England from the age of 13. I fear that in the US, Craig would have had an even much harder time, if that could be believed.
www.theguardian.com/news/2025/no...
www.theguardian.com/news/2025/no...
Check out my friend's podcast if you like detective games! Listened to an episode yesterday and thoroughly enjoyed it. Although, mainly playing Rimworld myself at the moment.
Hey all,
We are launching a podcast about detective games. We are going to touch on anything we think is detective themed, or has significant detective mechanics. We plan to release every two weeks, and hope you will enjoy! #detective #podcast
open.spotify.com/show/4e9MkSi...
We are launching a podcast about detective games. We are going to touch on anything we think is detective themed, or has significant detective mechanics. We plan to release every two weeks, and hope you will enjoy! #detective #podcast
open.spotify.com/show/4e9MkSi...
Introduction
open.spotify.com
November 9, 2025 at 7:53 PM
Check out my friend's podcast if you like detective games! Listened to an episode yesterday and thoroughly enjoyed it. Although, mainly playing Rimworld myself at the moment.
In my second article for Foom, I wrote about a recent study looking at the status of AGI safety engineering.
www.foommagazine.org/plans-to-bui...
www.foommagazine.org/plans-to-bui...
Plans to build AGI with nuclear reactor-like safety lack 'systematic thinking,' say researchers
Aspirations to adapt the principle of defense in depth from nuclear engineering to AI engineering appear to fall short on key requirements.
In a preprint from October 13, two researchers from the Ruh...
www.foommagazine.org
November 7, 2025 at 4:21 PM
In my second article for Foom, I wrote about a recent study looking at the status of AGI safety engineering.
www.foommagazine.org/plans-to-bui...
www.foommagazine.org/plans-to-bui...
Incredibly comprehensive reporting on what sounds like a massive sleeping giant oil pollution problem going on in Oklahoma right now.
Inspiring how so many people in such a defunct regulatory agency nevertheless still kept trying to affect change.
I feel sorry for Danny Ray.
Inspiring how so many people in such a defunct regulatory agency nevertheless still kept trying to affect change.
I feel sorry for Danny Ray.
Every oil state disposes of toxic wastewater underground. But in Oklahoma, it hasn't been staying there, creating large, ongoing pollution events known as "purges."
An investigation a year in the making for @readfrontier.bsky.social and @propublica.org
www.readfrontier.org/stories/toxi...
An investigation a year in the making for @readfrontier.bsky.social and @propublica.org
www.readfrontier.org/stories/toxi...
Toxic wastewater from oil fields keeps pouring out of the ground. Oklahoma regulators failed to stop it.
Salt water laced with cancer-causing chemicals, a byproduct of oil and gas drilling, is spewing from old wells. Experts warn of a pollution crisis spreading underground and threatening Oklahoma’s drin...
www.readfrontier.org
November 5, 2025 at 6:40 PM
Incredibly comprehensive reporting on what sounds like a massive sleeping giant oil pollution problem going on in Oklahoma right now.
Inspiring how so many people in such a defunct regulatory agency nevertheless still kept trying to affect change.
I feel sorry for Danny Ray.
Inspiring how so many people in such a defunct regulatory agency nevertheless still kept trying to affect change.
I feel sorry for Danny Ray.