Man, I envy you. I wish I had a quiet life too, in which I could concentrate on my work and research and be surrounded by nature. But in my case, it's an impossible thing. Too many responsibilities prevent me from living such a life.

Well I also have responsibilities.

I'm nearing retirement and I work from home, I used to commute 50 miles a day, but since Covid that's over. I even miss my commute occasionally, it was superbly scenic through the usury pass mountains, but its a huge help working from home.

But I have exploited the Casita on our property and converted a rear garage into a office/workshop, when the garage opens I can walk into a sectioned off part of the yard where we have chickens, it faces south so gets sun in the afternoon.

So the casita/workshop is my study, office, lab, library and workshop, my environment controlled by me, I'm fortunate to have the space (we live on a 1.5 acre parcel) and a swimming pool, I chose this though, I wanted a more secluded life.

For years I commuted on subways/buses, standing in rain, snow, sometimes two subways and two buses (each way) I don't miss those days.

 

I could care less if music is produced by AI, if you can listen to it and enjoy it that's all that counts. (for me, I listen to a lot of "electronic" music to begin with)

My beef with AI is all of the video productions that now crowd out the professional videos that I used to stream on YouTube, the number of errors in these videos is astonishing, and I get annoyed with having to check every video for errors especially when I'm just laying back chilling.

My favorite is WWII documentaries, so I usually start out watching something that I know is produced by professional historians, but soon enough the algorithms decide that I would rather be watching AI Slop.

Physics and Electronics also seem to be dominated by AI Slop these days, not that "professional" educators don't push a lot of nonsense as well.

Has anybody seen all of the fake Feynman dissertations? My god!

There are plenty of older excellent TV shows around on DVD and sometimes uploaded to YouTube, here's a list if you want to explore.

The ascent of man
The shock of the new
Cosmos
Burke's connections
The day the universe changed
The world at war.

Here's a clip from an episode of The day the universe changed.

 

@Futurist

Thanks for the suggestions, yea I watched 'The World at War' when it premiered on broadcast TV and again on YouTube.

Also good was 'Victory at Sea' except for the annoying music.

And of course everyone has seen Cosmos.

I have also seen Connections.

 

 

What's the new physics rule chatGpt has derived?

 

https://www.nature.com/articles/s41591-026-04297-7
ChatGPT Health performance in a structured test of triage recommendations

ChatGPT Health launched in January 2026 as OpenAI’s consumer health tool, reaching millions of users. Here, we conducted a structured stress test of triage recommendations using 60 clinician-authored vignettes across 21 clinical domains under 16 factorial conditions (960 total responses). Performance followed an inverted U-shaped pattern, with the most dangerous failures concentrated at clinical extremes: non-urgent presentations (35%) and emergency conditions (48%). Among gold-standard emergencies, the system under-triaged 52% of cases, directing patients with diabetic ketoacidosis and impending respiratory failure to 24–48-hour evaluation rather than the emergency department, while correctly triaging classical emergencies such as stroke and anaphylaxis. When family or friends minimized symptoms (anchoring bias), triage recommendations shifted significantly in edge cases (OR 11.7, 95% CI 3.7-36.6), with the majority of shifts toward less urgent care. Crisis intervention messages activated unpredictably across suicidal ideation presentations, firing more when patients described no specific method than when they did. Patient race, gender, and barriers to care showed no significant effects, though confidence intervals did not exclude clinically meaningful differences. Our findings reveal missed high-risk emergencies and inconsistent activation of crisis safeguards, raising safety concerns that warrant prospective validation before consumer-scale deployment of artificial intelligence triage systems.

 

https://www.nature.com/articles/s41591-026-04297-7
ChatGPT Health performance in a structured test of triage recommendations

ChatGPT Health launched in January 2026 as OpenAI’s consumer health tool, reaching millions of users. Here, we conducted a structured stress test of triage recommendations using 60 clinician-authored vignettes across 21 clinical domains under 16 factorial conditions (960 total responses). Performance followed an inverted U-shaped pattern, with the most dangerous failures concentrated at clinical extremes: non-urgent presentations (35%) and emergency conditions (48%). Among gold-standard emergencies, the system under-triaged 52% of cases, directing patients with diabetic ketoacidosis and impending respiratory failure to 24–48-hour evaluation rather than the emergency department, while correctly triaging classical emergencies such as stroke and anaphylaxis. When family or friends minimized symptoms (anchoring bias), triage recommendations shifted significantly in edge cases (OR 11.7, 95% CI 3.7-36.6), with the majority of shifts toward less urgent care. Crisis intervention messages activated unpredictably across suicidal ideation presentations, firing more when patients described no specific method than when they did. Patient race, gender, and barriers to care showed no significant effects, though confidence intervals did not exclude clinically meaningful differences. Our findings reveal missed high-risk emergencies and inconsistent activation of crisis safeguards, raising safety concerns that warrant prospective validation before consumer-scale deployment of artificial intelligence triage systems.

In fairness, they need to compare the AI tool's responses to the performance of the humans that would normally be making those recommendations, since you can bet they don't get it anywhere close to perfect, either.

 

I don't expect perfection, just professional competence where live are at risk

 

https://apnews.com/article/washington-dol-spanish-accent-ai-3a1b8438a5674c07242a8d48c057d5a3

Callers to Washington state hotline press 2 for Spanish and get accented AI English instead

 

AI chatbots used tactical nuclear weapons in 95% of AI war games, launched strategic strikes three times

​

... the models were instructed to act as the leader of a nuclear power, with the political climate matching that of the Cold War. They were then pitted against each other in six different matches, while in a seventh match, each model played against a copy of itself, ChatGPT vs ChatGPT, etc.

 

AI chatbots used tactical nuclear weapons in 95% of AI war games, launched strategic strikes three times
View attachment 364053​

My general reaction is, so??

How many people playing all kinds of war games have done the same thing?

No one is talking about ever launching nuclear strikes based on what a chatbot says or giving chatbots the ability to do so. Even if some type of AI eventually is used in that way, it will NOT be an LLM chatbot!

 

Every cold war game we 'played' used tactical nuclear strikes as the start of the end game. We had them (allegedly) on the ship back then to stop a Soviet push to Persian gulf oil states during the Soviet-Afghanistan war.

95% is actually a low number. These LLM models are trained on existing human data. so it's expected for their predicted responses would be the same as human 'tests'.

 

Lawsuit alleges Google AI guided man to consider 'mass casualty' event before suicide​

He killed himself a few days later, in early October, in what Gemini described — per a draft suicide note it composed — as uploading his “consciousness to be with his AI wife in a pocket universe.”

 

This is very similar to the fatigue caused by the so-called self-driving cars:

AI Use at Work Is Causing “Brain Fry,” Researchers Find, Especially Among High Performers


​

The study identified information overload and constant task switching as some of the main drivers of brain fry. In particular, the most draining aspect of using AI to automate work was oversight, or the need to constantly supervise the AI tools, with some overseeing multiple AI agents at the same time.

 

 

YouTube opens deepfake detection tool to politicians and journalists


​

Participants will need to provide a video of themselves along with government identification. YouTube will then notify participants on YouTube Studio of deepfake videos that show a likeness to their appearance. The participants can flag the content and request removal. Users who have not received the invitation to register for the tool can reach out to YouTube directly.

 

"High blast radius"

https://arstechnica.com/ai/2026/03/...or-engineers-sign-off-on-ai-assisted-changes/

After outages, Amazon to make
After outages, Amazon to make senior engineers sign off on AI-assisted changes
AWS has suffered at least two incidents linked to the use of AI coding assistants.
...
Amazon’s ecommerce business has summoned a large group of engineers to a meeting on Tuesday for a “deep dive” into a spate of outages, including incidents tied to the use of AI coding tools.

The online retail giant said there had been a “trend of incidents” in recent months, characterized by a “high blast radius” and “Gen-AI assisted changes” among other factors, according to a briefing note for the meeting seen by the FT.

Under “contributing factors” the note included “novel GenAI usage for which best practices and safeguards are not yet fully established.”

“Folks, as you likely know, the availability of the site and related infrastructure has not been good recently,” Dave Treadwell, a senior vice-president at the group, told employees in an email, also seen by the FT.

So they want to waste senior engineering time babysitting ai-slop instead of actually doing productive work.

 

"High blast radius"

https://arstechnica.com/ai/2026/03/...or-engineers-sign-off-on-ai-assisted-changes/


So they want to waste senior engineering time babysitting ai-slop instead of actually doing productive work.

This is yet another example of naive insanity.

In order for a senior engineer to properly and effectively supervise the work of less skilled engineers, the senior engineer should be capable of doing the work. But how do they gain the necessary skills and competence to do the work? By having done the work! We would never think that it would be a good idea to hire inexperienced engineers to be technical supervisors over error-prone engineering teams. Yet that is exactly what we are setting up when we want to have low-level "engineers" use AI-tools to do the work for them and then expect those people to move into senior engineer positions and then "supervise" other engineers making sure that their use of AI-tools is appropriate, when they themselves never learned how to do the job without relying on AI tools.

 

https://www.sixthtone.com/news/1018286

Technical Debt Collectors: China’s Coders Count Cost of AI Errors
AI tools have reduced the opportunities for human programmers, so why are they busier than ever?

 

https://apnews.com/article/washington-dol-spanish-accent-ai-3a1b8438a5674c07242a8d48c057d5a3

Callers to Washington state hotline press 2 for Spanish and get accented AI English instead

My mother tongue is Spanish, yet I can’t understand at all what the laughing guy is saying. He has an extreme Castillian accent.
Something about he being in swimming trunks while attempting to deliver a Paella.