Carl T. Bergstrom
banner
carlbergstrom.com
Carl T. Bergstrom
@carlbergstrom.com
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
As I was saying…
November 7, 2025 at 11:28 PM
G. K. Chesterton raises his hand…
November 6, 2025 at 3:53 PM
November 6, 2025 at 5:12 AM
Mood: just manually renewed SSL certification on my AWS-hosted website.
November 6, 2025 at 4:16 AM
It’s a good night for it.
November 5, 2025 at 4:57 AM
Shutting the fuck up for the next two decades would a superb start, Bill.
November 5, 2025 at 3:49 AM
Excited to read the book based on Berkeley's Sense and Sensibility and Science course (sensibility.berkeley.edu/index.php?ti...).

If you enjoyed Calling Bullshit, you'll find this a real treat.

Thanks, Saul!
November 5, 2025 at 2:10 AM
OK, inspired by @banditelli.org I wanted to post at least something crow. 🪶

Here they are engaging in mass vandalism just for the fun of it.

OK actually they're digging out beetle larvae that harm the lawn—you can see one in the iphone photo below—but it's pretty funny to see the mess they make.
November 4, 2025 at 5:33 AM
When the suspects appear to be white guys, it’s not a bombing — it’s an intentional explosion.
November 3, 2025 at 7:00 AM
Don't know if any of you have ever been tempted to use Evernote, but any company that eliminates their free tier AND makes current users subscribe to extract their own data is little better than a protection racket in my book.
October 29, 2025 at 6:53 AM
Also totally unfair to 18th century philosopher John Rawls.
October 28, 2025 at 10:11 PM
Regarding "far more common than reported here":

Left: the Nature News story above, Sept 18th 2025

Right: Wired directly quoting OpenAI, Oct 28, 2025

www.wired.com/story/chatgp...
October 28, 2025 at 8:54 PM
Springsteen: Deliver Me From Nowhere

Short thread:
October 28, 2025 at 4:11 PM
Confidence goals
🪶
October 26, 2025 at 7:45 PM
Por ejemplo:
October 26, 2025 at 4:18 AM
Seriously, @nytpitchbot.bsky.social you need to just hang up the gloves. You can’t compete.
October 26, 2025 at 3:10 AM
To be fair, dribbling a soccer ball through fire in the desert while wearing Oakley’s is exactly the sort of thing I would’ve thought it was super badass in 1985.

But I was 14 and it was 1985.
October 26, 2025 at 2:42 AM
Ubadah Sabbagh is referring to closed-access for-profit publishing here, but I keep thinking about how perfectly it describes the general push to incorporate large language models into the scientific workflow in the interest of higher "productivity".

ubadah.substack.com/p/beware-giv...
October 25, 2025 at 9:32 PM
I’ve been casual student of physics for four decades. Why did no one ever explain to me so clearly what a conservation law is?

From Susskind and Hrabovsky (2014) The Theoretical Minimum.



This is why reading basic material from brilliant thinkers is so often worth the investment.
October 25, 2025 at 7:43 AM
How could I have forgotten the company that’ll help you rat out your your neighbors to ICE?

techcrunch.com/2025/10/16/a...
October 25, 2025 at 2:33 AM
When we’re not doing sports gambling ads, LLMs that tell you how not to seem like you like your date too much, google cloud data-dredging baseball superstitions, this is the participatory surveillance world series.

Can we go back to beer ads, oversized trucks, and blood pressure pills please?
October 25, 2025 at 2:19 AM
Everything that’s going down in the NBA and we’re still advertising for sports betting on screen during the actual World Series coverage with the announcers talking about what’s a good bet?

Fuck this.
October 25, 2025 at 12:16 AM
5. For example, research produced and funded by tech companies often either frames problems as user-driven, or explores solutions as the obligation of users (e.g. community notes). Seldom does it explore consequences of design, UX, or algorithmic implementation, let alone the business model.
October 24, 2025 at 1:01 AM
1. We ( @jbakcoleman.bsky.social, @cailinmeister.bsky.social, @jevinwest.bsky.social, and I) have a new preprint up on the arXiv.

There we explore how social media companies and other online information technology firms are able to manipulate scientific research about the effects of their products.
October 24, 2025 at 12:47 AM
Does your ecology journal even lift, bro?
October 23, 2025 at 6:23 AM