Mike "looking for research fellowship" Caulfield
@mikecaulfield.bsky.social
Author: Verified: How to Think Straight, Get Duped Less, and Make Better Decisions about What to Believe Online (University of Chicago Press).
Researcher, infolit/misinfo/rhetoric/civic reasoning. Currently researching AI as tool for critical thinking.
Researcher, infolit/misinfo/rhetoric/civic reasoning. Currently researching AI as tool for critical thinking.
Pinned
What a simple question about My Cousin Vinny taught me about LLMs
If you want to see an LLM really malfunction, ask it to reconstruct a story from pieces. Also: Should this be a benchmark?
open.substack.com
I heard what everyone has been clamoring for is 6,000 words on what a simple question about the 1990s comedy My Cousin Vinny can teach us about the capabilities and surprising weaknesses of current LLMs. Well, I hate to disappoint people, so I went and wrote it. open.substack.com/pub/mikecaul...
Gift link to my piece on January 6 last year with Charlie Warzel arguing that social media was most dangerous not because it changed minds but because it allowed people to discard and neutralize the compelling evidence in front of them www.theatlantic.com/technology/a...
The Internet Is Worse Than a Brainwashing Machine
A rationale is always just a scroll or a click away.
www.theatlantic.com
January 6, 2026 at 2:09 PM
Gift link to my piece on January 6 last year with Charlie Warzel arguing that social media was most dangerous not because it changed minds but because it allowed people to discard and neutralize the compelling evidence in front of them www.theatlantic.com/technology/a...
Reposted by Mike "looking for research fellowship" Caulfield
that's a lot of power users gone because of dogpiling. christ
January 6, 2026 at 6:37 AM
that's a lot of power users gone because of dogpiling. christ
I just want people to remember that what this was was transparently obvious at the time and universally condemned. Hold on to that fact and don't let people change it under you.
January 6, 2026 at 7:17 AM
I just want people to remember that what this was was transparently obvious at the time and universally condemned. Hold on to that fact and don't let people change it under you.
One interesting thing I learned doing this is one reason my cousin Vinny feels unique but you can't put your finger on why is THERE IS NO VILLAIN. It's really weird!
I heard what everyone has been clamoring for is 6,000 words on what a simple question about the 1990s comedy My Cousin Vinny can teach us about the capabilities and surprising weaknesses of current LLMs. Well, I hate to disappoint people, so I went and wrote it. open.substack.com/pub/mikecaul...
What a simple question about My Cousin Vinny taught me about LLMs
If you want to see an LLM really malfunction, ask it to reconstruct a story from pieces. Also: Should this be a benchmark?
open.substack.com
January 6, 2026 at 7:04 AM
One interesting thing I learned doing this is one reason my cousin Vinny feels unique but you can't put your finger on why is THERE IS NO VILLAIN. It's really weird!
Saving this one
Bluesky should have a feature where if you get into an argument about whether an elected Democrat’s message was good or bad your phone delivers you an electric shock and makes you do a ten minute mindfulness lesson
January 6, 2026 at 6:52 AM
Saving this one
Reposted by Mike "looking for research fellowship" Caulfield
I will never be able to cover anything "normal" in this sport I swear
January 6, 2026 at 4:28 AM
I will never be able to cover anything "normal" in this sport I swear
this is correct
this is making me crazy because it doesn't mention the single best line delivery in the entire film!!! "They wa!" www.youtube.com/watch?v=z5Tq...
January 6, 2026 at 3:40 AM
this is correct
Reposted by Mike "looking for research fellowship" Caulfield
Someone get me a feinting chair this take on LLMs is nuanced.
I heard what everyone has been clamoring for is 6,000 words on what a simple question about the 1990s comedy My Cousin Vinny can teach us about the capabilities and surprising weaknesses of current LLMs. Well, I hate to disappoint people, so I went and wrote it. open.substack.com/pub/mikecaul...
What a simple question about My Cousin Vinny taught me about LLMs
If you want to see an LLM really malfunction, ask it to reconstruct a story from pieces. Also: Should this be a benchmark?
open.substack.com
January 5, 2026 at 4:29 AM
Someone get me a feinting chair this take on LLMs is nuanced.
Reposted by Mike "looking for research fellowship" Caulfield
A very good read, unsurprisingly.
I heard what everyone has been clamoring for is 6,000 words on what a simple question about the 1990s comedy My Cousin Vinny can teach us about the capabilities and surprising weaknesses of current LLMs. Well, I hate to disappoint people, so I went and wrote it. open.substack.com/pub/mikecaul...
What a simple question about My Cousin Vinny taught me about LLMs
If you want to see an LLM really malfunction, ask it to reconstruct a story from pieces. Also: Should this be a benchmark?
open.substack.com
January 5, 2026 at 1:43 PM
A very good read, unsurprisingly.
Reposted by Mike "looking for research fellowship" Caulfield
This is delightful
I heard what everyone has been clamoring for is 6,000 words on what a simple question about the 1990s comedy My Cousin Vinny can teach us about the capabilities and surprising weaknesses of current LLMs. Well, I hate to disappoint people, so I went and wrote it. open.substack.com/pub/mikecaul...
What a simple question about My Cousin Vinny taught me about LLMs
If you want to see an LLM really malfunction, ask it to reconstruct a story from pieces. Also: Should this be a benchmark?
open.substack.com
January 5, 2026 at 4:08 AM
This is delightful
Reposted by Mike "looking for research fellowship" Caulfield
I heard what everyone has been clamoring for is 6,000 words on what a simple question about the 1990s comedy My Cousin Vinny can teach us about the capabilities and surprising weaknesses of current LLMs. Well, I hate to disappoint people, so I went and wrote it. open.substack.com/pub/mikecaul...
What a simple question about My Cousin Vinny taught me about LLMs
If you want to see an LLM really malfunction, ask it to reconstruct a story from pieces. Also: Should this be a benchmark?
open.substack.com
January 5, 2026 at 1:48 AM
I heard what everyone has been clamoring for is 6,000 words on what a simple question about the 1990s comedy My Cousin Vinny can teach us about the capabilities and surprising weaknesses of current LLMs. Well, I hate to disappoint people, so I went and wrote it. open.substack.com/pub/mikecaul...
Reposted by Mike "looking for research fellowship" Caulfield
ZOMG @mikecaulfield.bsky.social knocked this one out of the park. A detailed demonstration of LLM’s weaknesses using a pop culture fave most of us have seen.
I heard what everyone has been clamoring for is 6,000 words on what a simple question about the 1990s comedy My Cousin Vinny can teach us about the capabilities and surprising weaknesses of current LLMs. Well, I hate to disappoint people, so I went and wrote it. open.substack.com/pub/mikecaul...
What a simple question about My Cousin Vinny taught me about LLMs
If you want to see an LLM really malfunction, ask it to reconstruct a story from pieces. Also: Should this be a benchmark?
open.substack.com
January 5, 2026 at 2:08 AM
ZOMG @mikecaulfield.bsky.social knocked this one out of the park. A detailed demonstration of LLM’s weaknesses using a pop culture fave most of us have seen.
I heard what everyone has been clamoring for is 6,000 words on what a simple question about the 1990s comedy My Cousin Vinny can teach us about the capabilities and surprising weaknesses of current LLMs. Well, I hate to disappoint people, so I went and wrote it. open.substack.com/pub/mikecaul...
What a simple question about My Cousin Vinny taught me about LLMs
If you want to see an LLM really malfunction, ask it to reconstruct a story from pieces. Also: Should this be a benchmark?
open.substack.com
January 5, 2026 at 1:48 AM
I heard what everyone has been clamoring for is 6,000 words on what a simple question about the 1990s comedy My Cousin Vinny can teach us about the capabilities and surprising weaknesses of current LLMs. Well, I hate to disappoint people, so I went and wrote it. open.substack.com/pub/mikecaul...
Reposted by Mike "looking for research fellowship" Caulfield
For my enemies, straight-talking realpolitik about the consolidation of power, control of the judiciary, and rampant corruption. For my friends, a complex host of extraordinary legal issues at the intersection of international law and the presidency.
January 4, 2026 at 2:07 AM
For my enemies, straight-talking realpolitik about the consolidation of power, control of the judiciary, and rampant corruption. For my friends, a complex host of extraordinary legal issues at the intersection of international law and the presidency.
Reposted by Mike "looking for research fellowship" Caulfield
"one of the most precise attacks on sovereignty" what?!?
January 3, 2026 at 5:01 PM
"one of the most precise attacks on sovereignty" what?!?
"one of the most precise attacks on sovereignty" what?!?
January 3, 2026 at 5:01 PM
"one of the most precise attacks on sovereignty" what?!?
He abducted necessary cast for the insane reality TV show he wants to run.
January 3, 2026 at 3:53 PM
He abducted necessary cast for the insane reality TV show he wants to run.
Reposted by Mike "looking for research fellowship" Caulfield
Among many many other things I think one of the risks of imagining you're fighting terrorists is you might underestimate the risks of creating them.
Among the many reasons you don’t kidnap a foreign head of state at gunpoint even if you have the capability, is that it sparks consequences you can neither control nor anticipate.
January 3, 2026 at 3:10 PM
Among many many other things I think one of the risks of imagining you're fighting terrorists is you might underestimate the risks of creating them.
Reposted by Mike "looking for research fellowship" Caulfield
A deluge of work on regime change after the 2003 invasion of Iraq yielded strong (and rare) agreement in IR: it doesn't work & has terrible consequences. @profdownes.bsky.social's book Catastrophic Success lays it out, but it's right there in the title. www.cornellpress.cornell.edu/book/9781501...
January 3, 2026 at 1:41 PM
A deluge of work on regime change after the 2003 invasion of Iraq yielded strong (and rare) agreement in IR: it doesn't work & has terrible consequences. @profdownes.bsky.social's book Catastrophic Success lays it out, but it's right there in the title. www.cornellpress.cornell.edu/book/9781501...
Speechless, really
January 3, 2026 at 3:03 PM
Speechless, really
Anyone got a gift link to this? www.wsj.com/tech/persona...
Your Key Survival Skill for 2026: Critical Ignoring
In an age of endless low-quality information, it’s time to fight our instinct to seek out and absorb all we can. It takes practice.
www.wsj.com
January 2, 2026 at 5:16 PM
Anyone got a gift link to this? www.wsj.com/tech/persona...
Reposted by Mike "looking for research fellowship" Caulfield
Without getting too much into *take* territory maybe movies partially suck now because so many people treat them as an endless parade of mundane streaming novelty rather than rewatchable artifacts that transcend "content"
January 2, 2026 at 8:29 AM
Without getting too much into *take* territory maybe movies partially suck now because so many people treat them as an endless parade of mundane streaming novelty rather than rewatchable artifacts that transcend "content"
Reposted by Mike "looking for research fellowship" Caulfield
There is also a class of people who are just confused by others rewatching or owning movies for precisely this reason and I think it is sad. Rewatching films is a great joy.
without any implications about quality, some people watch movies and read books as if the whole point is to try to guess what’s going to happen
these people drive me nuts
these people drive me nuts
January 2, 2026 at 8:18 AM
There is also a class of people who are just confused by others rewatching or owning movies for precisely this reason and I think it is sad. Rewatching films is a great joy.
Reposted by Mike "looking for research fellowship" Caulfield
GROK WILL SAY ANYTHING IT IS PROMPTED TO SAY.
“Grok gave no further details”—what are you even doing here, @reuters.com?!
“Grok gave no further details”—what are you even doing here, @reuters.com?!
Grok says safeguard lapses led to images of 'minors in minimal clothing' on X reut.rs/498SoUk
Grok says safeguard lapses led to images of 'minors in minimal clothing' on X
Elon Musk's xAI artificial intelligence chatbot Grok said on Friday lapses in safeguards had resulted in "images depicting minors in minimal clothing" on social media platform X and that improvements were being made to prevent this.
reut.rs
January 2, 2026 at 2:58 PM
GROK WILL SAY ANYTHING IT IS PROMPTED TO SAY.
“Grok gave no further details”—what are you even doing here, @reuters.com?!
“Grok gave no further details”—what are you even doing here, @reuters.com?!