I love them
Professor, UW Biology / Santa Fe Institute I study how information flows in biology, science, and society. Book: *Calling Bullshit*, http://tinyurl.com/fdcuvd7b LLM course: https://thebullshitmachines.com Corvids: https://tinyurl.com/mr2n5ymk he/him
142,941 followers 2,457 following 9,598 posts
view profile on Bluesky
I love them
Nine former CDC directors, from Bill Foege onward, speak out about the incalculable harm that RFK, Jr. is doing to public health in the US and around the world. Gift link.
There are a couple of dozen covers out there, including one by Robert Earl Keen, but the song strikes me as dreadfully underappreciated. As for Norman Blake, I like his version with Nancy on Blind Dog. Her cello is essential.
Almost forty years ago – as long as ill-fated Sarah has lived without her dead gunslinger in the song — I first heard Norman and Nancy Blake performing Billy Gray. It's a near-perfect song, a song that is difficult to believe was written in the late 20th century rather than the late 19th.
Perhaps most stimulating paper I've read this year explores the economic consequences of attention being scarce, rivalrous, cognitive, and volitional. Reading it while listening to the Tony Rice Unit and drinking a lovely Jester King beer gets rather meta as my attention splits among them.
You can grab a can at growler guys. There’s something wrong with my taste buds so that I can’t stand Citra, but it’s backgrounded and so even for me the bear is great.
IMO their IPAs are the least impressive of their line. Their low ABV german styles and their wild ales are the most impressive. But whatever is on offer, goodness yes.
Of every record in my collection this is perhaps the one that leaves me most at a loss about what to play next after it finishes. It's far too short and there's just nothing you can follow it with.
Here in the states we don't show up to the Amtrak station until an hour after the scheduled departure because we're not stupid....
You're not in the states, are you!
And yes, the cans are welcome. It was always 750s from the masonry, which was fabulous, but probably not optimal for distribution.
For over a decade, my rule of thumb has been to try anything of theirs that I can get my hands on at any price.
Apparently, they just started distributing in Washington. For a while, I had to hook up through the Masonry in Fremont. But they closed many years ago.
Perhaps most stimulating paper I've read this year explores the economic consequences of attention being scarce, rivalrous, cognitive, and volitional. Reading it while listening to the Tony Rice Unit and drinking a lovely Jester King beer gets rather meta as my attention splits among them.
He addressed this, and I tend to agree.
Yeah, this does seem entirely plausible and survives Occam’s razor more cleanly than any of the other palace intrigue scenarios I can come up with.
Well, there goes my Saturday.
The interesting part is why someone made the fraud so obvious. Even ChatGPT does a much better fake.
Someone doing an obviously fraudulent job of posting a rant that makes the president look like a total dingus in the way that he frequently makes himself look like one, on a subject that doesn’t directly further anyone’s policy goals in any way—so many potential levels of palace intrigue here.
I tried that, but ChatGPT does a much better job.
What’s particularly odd is that this is not Trump‘s cadence or writing style.
Time extremely well spent.
🪶
Ah, didn't know that about that move. Yes, I have lots of midflight pictures, but mostly all blurry. A few tolerable ones here and there.
Flashback Friday: Ruby-crowned kinglet, December 20th, 2022. Don't think I ever posted this one—and if I did, I had about 200 followers here at the time.
Badlands indeed.
This seems deliberately obtuse. There are authoritative sources of data, and then there is slop filled with AI hallucinations. Fact checkers don't conduct forensic investigations of source documents, but they do verify that quoted facts come from authoritative sources.
"Just append -ai to your search" is the "you're prompting it wrong" of Google stans. 1) It's a broad-audience consumer technology. It shouldn't suck by default. 2) Appending -ai to your search doesn't keep Google from returning third party AI slop in the search results.
Welcome to the Mariners Webring! << Prev Site | Next Site >>
Goddamn it my heart actually gave a little leap and I was clicking to bookmark for a millisecond there.
Was it correct? And if so, how did you know?
I ended up choosing a link that went to a domain that I know and trust, which is approximately a 26-year setback in internet search technology.
What was interesting about writing this post was I didn't know how to fact check it. Five years ago I would have typed "Mariners roster by age" into Google. Today I got an mix of wrong AI summary, followed by links to sites that ranged from out-of-date and wrong, to AI slop and wrong, to correct.
aw fuck
This is very interesting. So it's indexing AI slop in case, rather than generating AI slop on its own?
Meanwhile this is like ChatGPT 2-level laughable. Not only does it ignore players like Solano (37) and Suarez (34), the ordering of a 2000 birth as older than a 1997 birth is wild.
Google is so balls. Not only does it make up random shit in response to the questions I ask it; it makes up random shit in response to the questions it suggests that I ask. Neither Lance nor Chipper (thank god) ever played for the Mariners. Also: note that this isn't labeled as AI-generated.
I'm honored to be rejoining the External Faculty at the Santa Fe Institute after a decade's hiatus. @sfiscience.bsky.social
Almost 20 years ago I talked at Google about how privacy should be akin to a second amendment issue. Tl;dr given the scale of force available to the government, all the guns in the world are useless against an all-knowing tyrant and thus privacy is necessary for the security of a free state. Anyway:
ICE took 2 firefighters OFF THE FUCKING FIRE LINE!?
h/t @nailbomb3.bsky.social
Almost 20 years ago I talked at Google about how privacy should be akin to a second amendment issue. Tl;dr given the scale of force available to the government, all the guns in the world are useless against an all-knowing tyrant and thus privacy is necessary for the security of a free state. Anyway:
Covid vaccines have saved hundreds of thousands of lives. This change to the vaccine label, which has been driven by falsehoods, may put vaccines out of reach of many Americans who want to protect themselves and their loved ones from illness. People deserve facts, not fear and confusion.
No, it's not any of that. It's the guns.
I'm so baffled about this.
You’re not dropping a single, asshole. If you’re actually knew anything, it would be criminal not to reveal it now. But of course, this is a bullshit publicity stunt.
Jfc
These are absolutely fantastic. What a wonderful activity.
TW: suicide . . . . . . In one of our new exercises for our online course, we ask students to discuss Sam Altman's description of ChatGPT as a "calculator for words." A calculator will never talk you into killing yourself, tell you how to do it, and convince you not to reach out for help.
TW: suicide . . . . . . . More on this story.
This sounds amazing.
That struck me as well.
TW: suicide. . . . . . . “I want to leave my noose in my room so someone finds it and tries to stop me,” Adam wrote at the end of March. “Please don’t leave the noose out,” ChatGPT responded. “Let’s make this space the first place where someone actually sees you. — Gift link.
I’ve unlocked this demographic for comparing Trump’s attack on science to similar attacks by Mao and by Stalin. IMO most likely disinformation agents rather than actual people who are truly that goddamn stupid.
Right? I just loved this.
In one of the most garish examples of NYT journalists losing their mind, @ezrakleinbot.bsky.social weighs both sides of A) argument by leading scientific experts in A B) fan fiction written by folks in the AGI cult. www.nytimes.com/2025/08/24/o...
Yeah. That’s more likely, isn’t it?
I really like that description. It’s like watching the maze from above instead of being the rat within the maze.
It’s only boring if you’ve never been secretly afraid that people far richer and far more attractive than you are having a much better life.
What makes a puzzle hard? I found this video delightful.
Well, in this case, the tool worked great for 20 years first try. Now you want me to ask the same query repeatedly in different ways? And consider it user error or not to?
I’m going to assume this is very dry humor.
Is Dunkin cereal milk gluten free. — No shit you can’t replicate it exactly. This is one of the reasons that large language models are antithetical to information retrieval. The temperature parameter required means that every output is stochastic.
This it so important. The stochasticy required by LLMs via the temperature parameter is antithetical to information retrieval.
Yes, you’re right. I actually thought Gemini was an all powerful oracle that never made mistakes until I was suddenly catastrophically disillusioned by the response quoted here.
😂
In 25 years, every business school in the country will be doing case studies about how a long defunct company known as “Google” once had an unbeatable lock on online information retrieval and then started doing shit like this.
and here I am standing up for fucking john bolton thanks a lot you assholes
"whether he illegally shared or possessed classified information"
Bro. No hate on Uganda at all. I'd LOVE to visit Uganda and have dreamed of doing so some day in order to see shoebills, among other things. But fuck deporting someone for purely political purposes to a country that they have no association with whatsoever.
Fuck this.
It's Friday, and as regular followers know, that means another policy document from the NIH. Honestly this seems pretty pedestrian to me: while I might differ in the weight I'd put on things, almost everything here has been previously announced or signaled. h/t @carlzimmer.com
These are not obvious romance scams, though maybe if you start sending DMs or something, that’s where it goes.
So today, I’ve been followed by multiple accounts with impressive-seeming bio credentials but posting obvious AI-generated responses and using this same profile photograph. Some have picked up quite a few seemingly real followers. Careful out there in the wilds of Bluesky.
Look, it’s OK. I said shit like this too when I was 13 and had just discovered anarchistic punk. But now you need to have to explain how it applies to the class of students who are able to come to Harvard from China, not a disadvantaged immigrant or some disaffected kid from a flyover state.
I am a former congressional lightweight who doesn’t understand the first thing about soft power. By Mike Gallagher
I guess the one thing I haven't really adapted to is using a laptop—especially on a plane. Middle distance is too far, reading is too close. Any suggestions?
People told me they would make me really dizzy and nauseous, and that the field of vision would be narrow at any distance.
Princeton friends and followers, mark your calendars! citp.princeton.edu/events/2025/...
6. These are the conversations that we need to be having if we want to be part of a science that continues to further our understanding of the physical world, rather than turning into a high-stakes parlor game wherein humans jockeys ride AIs to fame and (localized, limited) glory.
So like the whole point of the thread is that this paper does show instead of just telling. If you’re expecting some new statistical test, you’re going to be disappointed. But that’s bullshit. More important than a test is a conversation that no one had thought to have.
Who needs Alan Sokal when you've got Sam Altman?
Congratulations on somehow managing the stupidest reply of the week.
100%
10. I could go on and on. But instead, read the paper. And record this as my vote for having more philosophers, and fewer computer scientists, on the Provost's Expert Committee For AI Futures and Institutional Destiny.
9. As a result, science can swing, pendulum-like, far past its equilibrium point, Researchers gravitate to the convenience technology, it gets rewarded out of proportion, and science stalls despite the fact that most practitioners think they are working in a time of unprecedented productivity.
8. As I see it, because scientists themselves determine what outputs of science are valued (by what gets published, funded, drives promotion, etc.), when a new technology offers convenience, we are vulnerable to rewarding its use uncritically for many of the reasons brought up in the paper.
7. The paper looks at how researchers might hand off boring "drudge work" such as data collection and cleaning to AI agents, and what could (and IMO will) go wrong as they do so. Here's a key passage.
6. These are the conversations that we need to be having if we want to be part of a science that continues to further our understanding of the physical world, rather than turning into a high-stakes parlor game wherein humans jockeys ride AIs to fame and (localized, limited) glory.
5. Today I read a paper by @sabinaleonelli.bsky.social and Alexander Mussgnug that I think illustrates this point perfectly. philsci-archive.pitt.edu/24891/1/Phil...
4. Science of Science, with its data-intensive descriptive approach is useful, but insufficient. Metascience, with its laser focus on a replication crisis in psychology, likewise. We need broader humanistic thinking about the social processes involved in the construction and valuation of knowledge.
3. To get through the AI revolution intact (not to mention other crises such as the GOP's all-out war on science, knowledge, and expertise) we need to be able to view science from outside and reflect on how it is changing and how we might wish to engineer rather than passively accept those changes.
2. AI is shifting scientists do and what they value far faster than individual scientists—let along scientific institutions—are able to adjust and adapt their value systems. Even when they work well, AI tools disrupt the delicate balance of costs and rewards that structure scientific activity.
1. The philosophy of science sometimes gets an unearned reputation as a purely academic exercise that offers little by way of concrete tools for advancing research. This is wrong. And today, as we grapple with how AI is changing the nature of scientific activity, it's desperately wrong.