kolmogorov gnostic
@javaid89.bsky.social
Rust belt dude working in the legacy auto sector. Unfit AND degenerate.
I think I need to log off ... Happy Holidays everyone!!!
December 22, 2025 at 4:18 PM
I think I need to log off ... Happy Holidays everyone!!!
Reposted by kolmogorov gnostic
When we say "no, everything hasn't been digitized," I need you to understand that we really mean is that virtually nothing has been digitized. This is because the realm of primary sources that historians use is incomprehensibly large.
Seems like it's worth posting this one again.
Why Do Historians Still Have To Go To Archives?
Why do historians go to archives? Hasn’t everything already been digitized?
contingentmagazine.org
December 22, 2025 at 1:40 AM
When we say "no, everything hasn't been digitized," I need you to understand that we really mean is that virtually nothing has been digitized. This is because the realm of primary sources that historians use is incomprehensibly large.
So like in a couple of years, you will be able to let an LLM "do tasks" for like a week ... With 50 to 80% accuracy... Sure I guess let's spend $5 trillion more on this. While starving all other industries.
There’s a bur after you get past the toy/novelty uses where the learning curve gets steep fast.
I don’t see a lot of discussion about how much more effort and scaffolding you need to put in to enable these products to run this long and actually produce good output metr.org/blog/2025-03...
I don’t see a lot of discussion about how much more effort and scaffolding you need to put in to enable these products to run this long and actually produce good output metr.org/blog/2025-03...
Measuring AI Ability to Complete Long Tasks
We propose measuring AI performance in terms of the *length* of tasks AI agents can complete. We show that this metric has been consistently exponentially increasing over the past 6 years, with a doub...
metr.org
December 21, 2025 at 10:41 PM
So like in a couple of years, you will be able to let an LLM "do tasks" for like a week ... With 50 to 80% accuracy... Sure I guess let's spend $5 trillion more on this. While starving all other industries.
Whoa whoa but Terrence Tao uses it! Are you telling me you're smarter than him!? How are Historians and Sociologists going to hit their Sprint Story points if they don't use LLMs!? Just learn how to ask questions bruh. /S
December 21, 2025 at 10:03 PM
Whoa whoa but Terrence Tao uses it! Are you telling me you're smarter than him!? How are Historians and Sociologists going to hit their Sprint Story points if they don't use LLMs!? Just learn how to ask questions bruh. /S
Reposted by kolmogorov gnostic
Much of SF lost power last night. Waymos blocked intersections across the city, even though they are Wayless than 1% of cars on the roads. They accumulated at their failure points--intersections with no green lights, in busy areas, and shut down traffic there.
December 21, 2025 at 4:31 PM
Much of SF lost power last night. Waymos blocked intersections across the city, even though they are Wayless than 1% of cars on the roads. They accumulated at their failure points--intersections with no green lights, in busy areas, and shut down traffic there.
Reposted by kolmogorov gnostic
Interesting Morozov critique of @abenanav.bsky.social on the challenge of the non-neutrality of generative AI www.theideasletter.org/essay/social...
December 20, 2025 at 8:57 PM
Interesting Morozov critique of @abenanav.bsky.social on the challenge of the non-neutrality of generative AI www.theideasletter.org/essay/social...
Is that shrimp Jar Jar Binks? IS NOTHING SACRED!!?
"Generative AI boosts academic productivity...The number of preprints an author produced was a measure of their productivity, while eventual publication in a journal was a measure of an article’s quality." Seems like biased criteria! #AcademicSky theconversation.com/what-the-hyp...
What the hyperproduction of AI slop is doing to science
A new study shows AI writing is turning traditional measures of research quality upside down.
theconversation.com
December 21, 2025 at 12:17 AM
Is that shrimp Jar Jar Binks? IS NOTHING SACRED!!?
Oh Man this is really tough!!!!!
🚨 THE FINAL TWO 🚨
From thirty-two candidates, we’ve now made it to the final round where Elon Musk and Peter Thiel will face off for the crown.
Who will win? You decide!
🗳️ Cast your ballot: twsu.forms.app/wpit25-final
From thirty-two candidates, we’ve now made it to the final round where Elon Musk and Peter Thiel will face off for the crown.
Who will win? You decide!
🗳️ Cast your ballot: twsu.forms.app/wpit25-final
December 20, 2025 at 5:37 PM
Oh Man this is really tough!!!!!
We don't have to call them "dangerous ideas"; some ideas are bad and wrong! Dangerous implies Curt!s Y@rv!n is cool (he is not).
Federal research agencies are pulling back from fostering public discussions about science after the hollowing out of government communications offices. www.statnews.com/2025/12/18/t...
As dangerous eugenic ideas spread, NIH falls silent
Federal research agencies are pulling back from fostering public discussions about science after the hollowing out of government communications offices.
www.statnews.com
December 18, 2025 at 6:29 PM
We don't have to call them "dangerous ideas"; some ideas are bad and wrong! Dangerous implies Curt!s Y@rv!n is cool (he is not).
But BUT think of all the SaaS businesses you can start! /s
December 18, 2025 at 5:09 PM
But BUT think of all the SaaS businesses you can start! /s
They are looking for "Storytellers" because the Management is grasping at straws and has no idea what their own companies even do. They've been trained to have no original thoughts beyond KPIs, Burn-down charts, Calibration.l, and whatever Elon Musk says.
Are these the same companies that told us the robot could do it?
December 18, 2025 at 4:46 PM
They are looking for "Storytellers" because the Management is grasping at straws and has no idea what their own companies even do. They've been trained to have no original thoughts beyond KPIs, Burn-down charts, Calibration.l, and whatever Elon Musk says.
Reposted by kolmogorov gnostic
Are these the same companies that told us the robot could do it?
December 17, 2025 at 1:31 PM
Are these the same companies that told us the robot could do it?
Reposted by kolmogorov gnostic
Just really awesome how our financial elite just spent the last several decades admonishing people for pursing the humanities and creative fields, and are now suddenly pedaling the idea that "there are no good ideas anymore" after spending the last 5 years laying off everybody with a brain
December 16, 2025 at 7:10 PM
Just really awesome how our financial elite just spent the last several decades admonishing people for pursing the humanities and creative fields, and are now suddenly pedaling the idea that "there are no good ideas anymore" after spending the last 5 years laying off everybody with a brain
Reposted by kolmogorov gnostic
"human driver" statistical comparisons are fraught because these AVs only operate on certain roads and in certain conditions, but the important takeaway here is that Teslas are deeply unsafe and should not be on the road
"data suggests Tesla Robotaxis are crashing once every 40,000 miles, whereas the average human driver in the US crashes about once every 500,000 miles"
that's just really really bad sherwood.news/tech/teslas-...
that's just really really bad sherwood.news/tech/teslas-...
Tesla’s 29 Austin Robotaxis have crashed 8 times since June, as data suggests they perform much worse than human drivers
That’s a lot of crashes for such a small fleet....
sherwood.news
December 16, 2025 at 2:23 PM
"human driver" statistical comparisons are fraught because these AVs only operate on certain roads and in certain conditions, but the important takeaway here is that Teslas are deeply unsafe and should not be on the road
Yes honestly, this is the way. For so so many things. And at a broader level, if we can't discipline monopolists and rentiers via democratic institutions. Then this is the way.
To anyone who may have been present at my talk today in which I intimated that you can just, you know, do this: you can just, you know, do this.
December 14, 2025 at 3:42 PM
Yes honestly, this is the way. For so so many things. And at a broader level, if we can't discipline monopolists and rentiers via democratic institutions. Then this is the way.
1. AGI is inevitable and is going to be disruptive
2. It's going to be better than everyone at any cognitive task
3. We don't know what to do.
idk man. Where is the problem? How isn't "ask the AGI" the obvious answer here? Why you gotta tilt at windmills instead of doing something useful.
2. It's going to be better than everyone at any cognitive task
3. We don't know what to do.
idk man. Where is the problem? How isn't "ask the AGI" the obvious answer here? Why you gotta tilt at windmills instead of doing something useful.
The arrival of AGI | Shane Legg (co-founder of DeepMind)
YouTube video by Google DeepMind
www.youtube.com
December 14, 2025 at 3:46 AM
1. AGI is inevitable and is going to be disruptive
2. It's going to be better than everyone at any cognitive task
3. We don't know what to do.
idk man. Where is the problem? How isn't "ask the AGI" the obvious answer here? Why you gotta tilt at windmills instead of doing something useful.
2. It's going to be better than everyone at any cognitive task
3. We don't know what to do.
idk man. Where is the problem? How isn't "ask the AGI" the obvious answer here? Why you gotta tilt at windmills instead of doing something useful.
Reposted by kolmogorov gnostic
Earlier this week, 42 state attorneys general sent a warning letter to the major AI companies regarding the dangers and harms created by AI chatbots. It's as forceful an indictment of what's happening that I've found, so here's a short thread to walk you through the evidence they've compiled... 🧵
Big Tech warned over AI 'delusional' outputs by US attorneys general
Microsoft, Meta, Google and Apple were among the 13 companies that received a warning from a bipartisan group of state attorneys general, according to a letter from the state leaders, who said their c...
www.reuters.com
December 11, 2025 at 1:02 PM
Earlier this week, 42 state attorneys general sent a warning letter to the major AI companies regarding the dangers and harms created by AI chatbots. It's as forceful an indictment of what's happening that I've found, so here's a short thread to walk you through the evidence they've compiled... 🧵
Reposted by kolmogorov gnostic
Do not accept the premise that education is to blame for abysmal jobs outcomes.
“The fantasy economy's framing of economic inequality… focuses exclusively on education…deflects attention away from decades of public policies and changing business practices that have…contributed to stagnating wages”
“The fantasy economy's framing of economic inequality… focuses exclusively on education…deflects attention away from decades of public policies and changing business practices that have…contributed to stagnating wages”
August 18, 2025 at 1:12 PM
Do not accept the premise that education is to blame for abysmal jobs outcomes.
“The fantasy economy's framing of economic inequality… focuses exclusively on education…deflects attention away from decades of public policies and changing business practices that have…contributed to stagnating wages”
“The fantasy economy's framing of economic inequality… focuses exclusively on education…deflects attention away from decades of public policies and changing business practices that have…contributed to stagnating wages”
Reposted by kolmogorov gnostic
I perceive now that the AI apocalypse is just digital California: being surrounded by impossibly beautiful people who respond to you with an unflappably pleasant and yet condescending tone that hides either total disdain or the inability to process human thought
December 10, 2025 at 4:01 PM
I perceive now that the AI apocalypse is just digital California: being surrounded by impossibly beautiful people who respond to you with an unflappably pleasant and yet condescending tone that hides either total disdain or the inability to process human thought
Horses
AI progress is steady. Human equivalence is sudden.
andyljones.com
December 10, 2025 at 4:17 PM
I don't understand the allure of "automated science"? I see this "obsession with solving things" a strange "engineer brained" pathology. I say this as an engineer with many many "engineer brained" friends and family. Is science only good for instrumental knowledge? Do people not want to know things?
1. In Hamburg there is a "fully autonomous train" that for some reason has a human operator to press the ON switch first thing in the morning and the OFF switch at the end of the night. In a recent talk about AI in science, I asked the audience to consider whether they want to be like that operator.
All aboard except the driver? A fully autonomous train takes to the tracks in Germany
YouTube video by euronews
www.youtube.com
December 10, 2025 at 3:05 AM
I don't understand the allure of "automated science"? I see this "obsession with solving things" a strange "engineer brained" pathology. I say this as an engineer with many many "engineer brained" friends and family. Is science only good for instrumental knowledge? Do people not want to know things?
Reposted by kolmogorov gnostic
Reposted by kolmogorov gnostic
A quick note on the argument form: LLMs fail on X, but so do humans, so LLM-human comparisons are fair game.
vincentcarchidi.substack.com/p/a-note-on-...
vincentcarchidi.substack.com/p/a-note-on-...
A Note on "LLMs Fail on X, But So Do Humans"
Can we stop doing this?
vincentcarchidi.substack.com
December 9, 2025 at 6:04 PM
A quick note on the argument form: LLMs fail on X, but so do humans, so LLM-human comparisons are fair game.
vincentcarchidi.substack.com/p/a-note-on-...
vincentcarchidi.substack.com/p/a-note-on-...
Reposted by kolmogorov gnostic
Having fun here thinking of truckers saying to themselves "I'd better switch on the plasma actuators."
www.arxiv.org/abs/2512.03613
www.arxiv.org/abs/2512.03613
Drag reduction via separation control using plasma actuators on a truck cabin side
We investigated the drag reduction on a heavy-duty vehicle by means of dielectric-barrier discharge plasma actuators located on the A-pillars. An experimental campaign is carried out on a generalized ...
www.arxiv.org
December 8, 2025 at 2:04 PM
Having fun here thinking of truckers saying to themselves "I'd better switch on the plasma actuators."
www.arxiv.org/abs/2512.03613
www.arxiv.org/abs/2512.03613
Hot-take: "Learn to code" people knew less than Jon Snow. But I am not sure if supply and demand is the only way to look at things (isn't there anything beyond micro-econ 101?). And CS = "Coding" is like saying astronomy is the study of telescopes.
I think a lot of "learn to code" people thought high salaries for programmers were a permanent feature of the world rather than the result of unusual circumstances around supply and demand in the labor market. Over last 20 years, CS went from a niche college major to the most popular major in the US
December 6, 2025 at 2:20 AM
Hot-take: "Learn to code" people knew less than Jon Snow. But I am not sure if supply and demand is the only way to look at things (isn't there anything beyond micro-econ 101?). And CS = "Coding" is like saying astronomy is the study of telescopes.