This is disgraceful trash. Shame on the AHA and any historian who goes along with this.
Generative “A.i.” is controlled by billionaires. If you use that tech, you let billionaires control the information you absorb.
We saw Grok go Nazi. What prevents a billionaire from corrupting any other LLM?
Generative “A.i.” is controlled by billionaires. If you use that tech, you let billionaires control the information you absorb.
We saw Grok go Nazi. What prevents a billionaire from corrupting any other LLM?
The AHA has published Guiding Principles for Artificial Intelligence in History Education, offering a disciplinary approach to AI that focuses on the specific needs and challenges of history educators. 🗃️
Guiding Principles for Artificial Intelligence in History Education
These 14 foundational principles are meant to assist educators and administrators in crafting AI policies suited to local circumstances and the specific needs of students.
www.historians.org
August 5, 2025 at 9:04 PM
This is disgraceful trash. Shame on the AHA and any historian who goes along with this.
Generative “A.i.” is controlled by billionaires. If you use that tech, you let billionaires control the information you absorb.
We saw Grok go Nazi. What prevents a billionaire from corrupting any other LLM?
Generative “A.i.” is controlled by billionaires. If you use that tech, you let billionaires control the information you absorb.
We saw Grok go Nazi. What prevents a billionaire from corrupting any other LLM?
If you are an IRB staffer and you approve studies which will put participant data into chatGPT or any other LLM that is not completely internally controlled and disconnected from cloud services. WHAT THE FUCK.
April 13, 2024 at 11:48 PM
If you are an IRB staffer and you approve studies which will put participant data into chatGPT or any other LLM that is not completely internally controlled and disconnected from cloud services. WHAT THE FUCK.
My custom game engine written in my custom programming language, CURLEE, running on my custom operating system, JOE, on a real desktop being controlled via an llm
#screenshotsaturday #indiedev #indie #opensource #software
#screenshotsaturday #indiedev #indie #opensource #software
September 27, 2026 at 9:53 AM
My custom game engine written in my custom programming language, CURLEE, running on my custom operating system, JOE, on a real desktop being controlled via an llm
#screenshotsaturday #indiedev #indie #opensource #software
#screenshotsaturday #indiedev #indie #opensource #software
i hate how LLMs has collected vasts amounts of pictures of everyone's faces, from LLM-operated government agencies and LLM-controlled businesses, in order to do mass surveillance, and then the LLMs looked at all the pictures and the LLM made a false positive and then the LLM-police arrested her
If only a college professor could condescendingly yell at her that AI is useful, I bet she’d feel so much better about all this
Grandma jailed after AI error linked her to crime in state she never visited: lawyer
The charges were later dismissed after bank records showed she was in Tennessee at the time
www.the-independent.com
March 14, 2026 at 10:31 AM
i hate how LLMs has collected vasts amounts of pictures of everyone's faces, from LLM-operated government agencies and LLM-controlled businesses, in order to do mass surveillance, and then the LLMs looked at all the pictures and the LLM made a false positive and then the LLM-police arrested her
You know since Labor are controlled opposition this is going to end with a “settlement” where the Australian Government pays Altman and his cronies a huge chunk of tax revenue to add an LLM chatbot to a public service which will be worsened and possibly outright compromised by it
September 24, 2026 at 12:46 AM
You know since Labor are controlled opposition this is going to end with a “settlement” where the Australian Government pays Altman and his cronies a huge chunk of tax revenue to add an LLM chatbot to a public service which will be worsened and possibly outright compromised by it
But we also need to be wary of creating situations where entities that are owned and controlled by the wealthy, like corporations, are treated better than real living human beings.
We cannot create a scenario where a person will be punished for damaging an LLM, but not visa-versa...
We cannot create a scenario where a person will be punished for damaging an LLM, but not visa-versa...
December 28, 2025 at 7:30 PM
But we also need to be wary of creating situations where entities that are owned and controlled by the wealthy, like corporations, are treated better than real living human beings.
We cannot create a scenario where a person will be punished for damaging an LLM, but not visa-versa...
We cannot create a scenario where a person will be punished for damaging an LLM, but not visa-versa...
thanks to Bruce Sterling
“When LLMs operate in the real world through LLM-controlled robots, they can pose a serious, tangible threat.”
“When LLMs operate in the real world through LLM-controlled robots, they can pose a serious, tangible threat.”
It's Surprisingly Easy to Jailbreak LLM-Driven Robots
Researchers induced bots to ignore their safeguards without exception
spectrum.ieee.org
November 20, 2024 at 5:49 PM
thanks to Bruce Sterling
“When LLMs operate in the real world through LLM-controlled robots, they can pose a serious, tangible threat.”
“When LLMs operate in the real world through LLM-controlled robots, they can pose a serious, tangible threat.”
“llms aren’t easily controlled! look at all the stories of them deleting prod or the user’s files”
“control” is not the same thing as “directing”
you can prevent an llm from deleting prod by never allowing it access to prod. that’s easy. you may not be able to direct it once you give it access
“control” is not the same thing as “directing”
you can prevent an llm from deleting prod by never allowing it access to prod. that’s easy. you may not be able to direct it once you give it access
September 20, 2026 at 4:42 PM
“llms aren’t easily controlled! look at all the stories of them deleting prod or the user’s files”
“control” is not the same thing as “directing”
you can prevent an llm from deleting prod by never allowing it access to prod. that’s easy. you may not be able to direct it once you give it access
“control” is not the same thing as “directing”
you can prevent an llm from deleting prod by never allowing it access to prod. that’s easy. you may not be able to direct it once you give it access
Evidence that a well prompted LLM can help learning from a (small, single-subject) randomized controlled trial at Harvard: “here we show that students learn more than twice as much in less time with an AI tutor compared to an active learning classroom, while also being more engaged and motivated.”
March 21, 2025 at 5:47 PM
Evidence that a well prompted LLM can help learning from a (small, single-subject) randomized controlled trial at Harvard: “here we show that students learn more than twice as much in less time with an AI tutor compared to an active learning classroom, while also being more engaged and motivated.”
LLMs good performance on medical exams does not translate to accurate performance in real-world settings (preregistered n~1,300 study). This can't be explained by current standard benchmarks for medical knowledge & simulated patient interactions.
www.nature.com/articles/s41...
www.nature.com/articles/s41...
Reliability of LLMs as medical assistants for the general public: a randomized preregistered study - Nature Medicine
In a randomized controlled study involving 1,298 participants from a general sample, performance of humans when assisted by a large language model (LLM) was sensibly inferior to that of the LLM alone ...
www.nature.com
February 11, 2026 at 10:30 AM
LLMs good performance on medical exams does not translate to accurate performance in real-world settings (preregistered n~1,300 study). This can't be explained by current standard benchmarks for medical knowledge & simulated patient interactions.
www.nature.com/articles/s41...
www.nature.com/articles/s41...
New paper that tries to empirically quantify how training-set contamination impacts LLM benchmark scores, by training a variety of 1B and 8B models with controlled types and levels of contamination.
Overestimation in LLM Evaluation: A Controlled Large-Scale Study on Data Contamination's Impact on Machine Translation
Data contamination -- the accidental consumption of evaluation examples within the pre-training data -- can undermine the validity of evaluation benchmarks. In this paper, we present a rigorous analys...
arxiv.org
February 7, 2025 at 4:02 PM
New paper that tries to empirically quantify how training-set contamination impacts LLM benchmark scores, by training a variety of 1B and 8B models with controlled types and levels of contamination.
the concept of "hallucinations" refers to the unknown in llm response, things that aren't deliberately controlled. when the person in charge of the model says he's going to "fix" the model and his fixes consistently make the model talk about race and eugenics that is a different thing
Elon's AI is now actively recommending a second Holocaust
July 9, 2025 at 8:48 AM
the concept of "hallucinations" refers to the unknown in llm response, things that aren't deliberately controlled. when the person in charge of the model says he's going to "fix" the model and his fixes consistently make the model talk about race and eugenics that is a different thing
THE HOSPITAL IV DRIP PUMP IZ CONTROLLED BY A CONSTANTLY HALLUCIN8ING LLM what the FUCK
January 16, 2026 at 3:02 AM
THE HOSPITAL IV DRIP PUMP IZ CONTROLLED BY A CONSTANTLY HALLUCIN8ING LLM what the FUCK
This is fucking stupid. The world doesn't need another arms race that'll lead us into war while we fuck with technology we don't understand. We need a total ban on AI weapons and international, democratically controlled, safe, slow research into AGI development which will not come from LLM.
June 19, 2025 at 11:38 PM
This is fucking stupid. The world doesn't need another arms race that'll lead us into war while we fuck with technology we don't understand. We need a total ban on AI weapons and international, democratically controlled, safe, slow research into AGI development which will not come from LLM.
Christys: “The Supreme Leader has been slotted.” A ‘news’ channel presented by a man with a brain controlled by an LLM trained on crap SAS novels.
February 28, 2026 at 8:19 PM
Christys: “The Supreme Leader has been slotted.” A ‘news’ channel presented by a man with a brain controlled by an LLM trained on crap SAS novels.
And, just as prisoners desperately desire and fight for their pennies-on-the-hour jobs because it provides them *something to do*. I can totally see how a tablet with an LLM you can nonstop interact with (controlled by the warden) would feel immensely good to many folks behind bars.
September 8, 2026 at 11:50 PM
And, just as prisoners desperately desire and fight for their pennies-on-the-hour jobs because it provides them *something to do*. I can totally see how a tablet with an LLM you can nonstop interact with (controlled by the warden) would feel immensely good to many folks behind bars.
"When civilians ask large language models questions about their health, the technology often gets it wrong."
"The advancements could inadvertently expose more people to medical misinformation." h/t POLITICO Pulse
www.nature.com/articles/s41...
"The advancements could inadvertently expose more people to medical misinformation." h/t POLITICO Pulse
www.nature.com/articles/s41...
Reliability of LLMs as medical assistants for the general public: a randomized preregistered study - Nature Medicine
In a randomized controlled study involving 1,298 participants from a general sample, performance of humans when assisted by a large language model (LLM) was sensibly inferior to that of the LLM alone ...
www.nature.com
February 10, 2026 at 11:34 AM
"When civilians ask large language models questions about their health, the technology often gets it wrong."
"The advancements could inadvertently expose more people to medical misinformation." h/t POLITICO Pulse
www.nature.com/articles/s41...
"The advancements could inadvertently expose more people to medical misinformation." h/t POLITICO Pulse
www.nature.com/articles/s41...
So-called “AI” handwringing over safety for bioweapons are LLM-advocates preying on ignorance. It’s never been especially hard or secret to make a bioweapon and indeed crude ricin is fairly common, but the actual barrier is personal art and controlled equipment like [REDACTED]
June 10, 2026 at 10:22 PM
So-called “AI” handwringing over safety for bioweapons are LLM-advocates preying on ignorance. It’s never been especially hard or secret to make a bioweapon and indeed crude ricin is fairly common, but the actual barrier is personal art and controlled equipment like [REDACTED]
For those outside the great empires, the real lesson of DeepSeek is that anyone can build an LLM. We need AI that isn't controlled by the US oligarchy or the CCP.
January 29, 2025 at 12:29 AM
For those outside the great empires, the real lesson of DeepSeek is that anyone can build an LLM. We need AI that isn't controlled by the US oligarchy or the CCP.
"In our work, we found that none of the tested language models were ready for deployment in direct patient care."
#medlibs
www.nature.com/articles/s41...
#medlibs
www.nature.com/articles/s41...
Reliability of LLMs as medical assistants for the general public: a randomized preregistered study - Nature Medicine
In a randomized controlled study involving 1,298 participants from a general sample, performance of humans when assisted by a large language model (LLM) was sensibly inferior to that of the LLM alone ...
www.nature.com
February 11, 2026 at 1:15 PM
"In our work, we found that none of the tested language models were ready for deployment in direct patient care."
#medlibs
www.nature.com/articles/s41...
#medlibs
www.nature.com/articles/s41...
Work led by Andrew Bean who's finishing his PhD with Adam Mahdi and I. Great collaboration Rebecca Payne, Guy Parsons, Hannah Rose Kirk, Juan Manuel Ciro Torres, Rafael Esteban Mosquera Gómez, Sara Hincapié Monsalve, Aruna S. Ekanayaka, Lionel Tarassenko!
Read at www.nature.com/articles/s41...
Read at www.nature.com/articles/s41...
Reliability of LLMs as medical assistants for the general public: a randomized preregistered study - Nature Medicine
In a randomized controlled study involving 1,298 participants from a general sample, performance of humans when assisted by a large language model (LLM) was sensibly inferior to that of the LLM alone ...
www.nature.com
February 9, 2026 at 5:08 PM
Work led by Andrew Bean who's finishing his PhD with Adam Mahdi and I. Great collaboration Rebecca Payne, Guy Parsons, Hannah Rose Kirk, Juan Manuel Ciro Torres, Rafael Esteban Mosquera Gómez, Sara Hincapié Monsalve, Aruna S. Ekanayaka, Lionel Tarassenko!
Read at www.nature.com/articles/s41...
Read at www.nature.com/articles/s41...
I do think the conversation around LLM usage in tech is somewhat different from creative disciplines, nonfiction writing, language translation, etc. But some things remain unchanged: Tools should be open source rather than APIs controlled by corporations and training data must be ethically sourced.
April 12, 2026 at 11:06 PM
I do think the conversation around LLM usage in tech is somewhat different from creative disciplines, nonfiction writing, language translation, etc. But some things remain unchanged: Tools should be open source rather than APIs controlled by corporations and training data must be ethically sourced.
Friend is making a game with an LLM controlled girl - Asa, with whom player (viewers) can interact on stream.
I was in charge of character design and we're slowly working on sprites!
Look forward to her development and next stream!
www.twitch.tv/originaldoku
#art #gamedev
I was in charge of character design and we're slowly working on sprites!
Look forward to her development and next stream!
www.twitch.tv/originaldoku
#art #gamedev
April 7, 2025 at 7:05 AM
Friend is making a game with an LLM controlled girl - Asa, with whom player (viewers) can interact on stream.
I was in charge of character design and we're slowly working on sprites!
Look forward to her development and next stream!
www.twitch.tv/originaldoku
#art #gamedev
I was in charge of character design and we're slowly working on sprites!
Look forward to her development and next stream!
www.twitch.tv/originaldoku
#art #gamedev