Q McCallum
@qethanm.bsky.social
NEW BOOK: Twin Wolves: Balancing risk and reward to make the most of AI
https://twinwolvesai.com/
research: AI, risk, complexity, finance history
newsletter: https://complex-machinery.com
(some posts in 🇫🇷 🇩🇪 🇷🇺)
https://twinwolvesai.com/
research: AI, risk, complexity, finance history
newsletter: https://complex-machinery.com
(some posts in 🇫🇷 🇩🇪 🇷🇺)
Pinned
Twin Wolves AI
Balancing risk and reward to make the most of AI
TwinWolvesAI.com
It's here! I've just released my latest book, "Twin Wolves: Balancing risk and reward to make the most of AI."
This is a tight, executive-level read on how to approach AI (both ML/AI and genAI) in your company.
twinwolvesai.com
#dataBS
This is a tight, executive-level read on how to approach AI (both ML/AI and genAI) in your company.
twinwolvesai.com
#dataBS
Insurance is a game of risk transfer, and insurers do tons of homework on their side of the proposed bet. It's how they make their money.
So when they've decided a given risk is uninsurable, that is a sign.
So when they've decided a given risk is uninsurable, that is a sign.
Insurers retreat from AI cover as risk of multibillion-dollar claims mounts on.ft.com/3Xxth8F
Insurers retreat from AI cover as risk of multibillion-dollar claims mounts
AIG, Great American and WR Berkley seek permission to limit liability from AI agents and chatbots
on.ft.com
November 26, 2025 at 12:25 AM
Insurance is a game of risk transfer, and insurers do tons of homework on their side of the proposed bet. It's how they make their money.
So when they've decided a given risk is uninsurable, that is a sign.
So when they've decided a given risk is uninsurable, that is a sign.
The opening line is a right hook. And then the hits keep coming.
"OpenAI is a money pit with a website on top. That much we know already, but [...] there's a lot of guesswork required when estimating the depth of the pit."
"OpenAI is a money pit with a website on top. That much we know already, but [...] there's a lot of guesswork required when estimating the depth of the pit."
November 26, 2025 at 12:04 AM
The opening line is a right hook. And then the hits keep coming.
"OpenAI is a money pit with a website on top. That much we know already, but [...] there's a lot of guesswork required when estimating the depth of the pit."
"OpenAI is a money pit with a website on top. That much we know already, but [...] there's a lot of guesswork required when estimating the depth of the pit."
Reposted by Q McCallum
NEW: AI “recipe slop” is overrunning search and social. Food creators say Google’s AI Overviews and glossy fake food pics are drowning out real, tested recipes — collapsing traffic and setting home cooks up for disaster, especially this Thanksgiving.
Gift link: www.bloomberg.com/news/article...
Gift link: www.bloomberg.com/news/article...
AI Slop Recipes Are Taking Over the Internet -- And Thanksgiving Dinner
Food bloggers see traffic dip as home cooks turn to AI, inspired by impossible pictures
www.bloomberg.com
November 25, 2025 at 3:16 PM
NEW: AI “recipe slop” is overrunning search and social. Food creators say Google’s AI Overviews and glossy fake food pics are drowning out real, tested recipes — collapsing traffic and setting home cooks up for disaster, especially this Thanksgiving.
Gift link: www.bloomberg.com/news/article...
Gift link: www.bloomberg.com/news/article...
While the American tech sector focuses on "kinda / maybe / someday" AI hype, companies in China are getting results _today_ by focusing on practical, meaningful use cases.
Remember: AI-the-technology is fine. AI-the-far-off-dream-sold-as-reality is the problem.
Remember: AI-the-technology is fine. AI-the-far-off-dream-sold-as-reality is the problem.
Sam Altman wants artificial intelligence to cure cancer. China is focused on something more prosaic: making better washing machines.
Robots and AI Are Already Remaking the Chinese Economy
China’s factories and ports are learning to make and export more goods faster, cheaper and with fewer workers.
on.wsj.com
November 25, 2025 at 4:31 PM
While the American tech sector focuses on "kinda / maybe / someday" AI hype, companies in China are getting results _today_ by focusing on practical, meaningful use cases.
Remember: AI-the-technology is fine. AI-the-far-off-dream-sold-as-reality is the problem.
Remember: AI-the-technology is fine. AI-the-far-off-dream-sold-as-reality is the problem.
Hats off to @404media.co for their stance on data privacy.
Since they don't engage in invasive data collection, they're doing the honest thing of simply _asking_ people for information.
So why not give their survey a whirl?
Since they don't engage in invasive data collection, they're doing the honest thing of simply _asking_ people for information.
So why not give their survey a whirl?
We need your help. Because 404 Media uses Ghost (a privacy-forward stack) we don't know much about our readers. Do they find us through articles, or podcasts? Are readers on Bluesky, Mastodon, both? Please fill this out to help us grow sustainably, it takes 1 min! www.404media.co/please-pleas...
Please, please do our reader survey
It'll take just a minute and help 404 Media figure out how to grow sustainably.
www.404media.co
November 25, 2025 at 3:23 PM
Hats off to @404media.co for their stance on data privacy.
Since they don't engage in invasive data collection, they're doing the honest thing of simply _asking_ people for information.
So why not give their survey a whirl?
Since they don't engage in invasive data collection, they're doing the honest thing of simply _asking_ people for information.
So why not give their survey a whirl?
No clue why this post feels relevant again...
People say that the current state of genAI echoes 2001 (Dot-Com).
I see that, but I see even stronger parallels to 2008 (GFC).
In particular, it feels like the genAI space is aiming for "too big to fail" status.
Might be a good time to review some GFC-related reads ...
I see that, but I see even stronger parallels to 2008 (GFC).
In particular, it feels like the genAI space is aiming for "too big to fail" status.
Might be a good time to review some GFC-related reads ...
November 25, 2025 at 4:44 AM
No clue why this post feels relevant again...
Reposted by Q McCallum
Today marks one year since I left my job with my pals at Aquatic. I've not quite achieved all the things I wanted to do in this time but I definitely have had an amazing time. CE work, conference talks, Computerphile, emulators...it's been a blast.
One week til new job at Hudson River! Can't wait!!
One week til new job at Hudson River! Can't wait!!
November 24, 2025 at 12:14 PM
Today marks one year since I left my job with my pals at Aquatic. I've not quite achieved all the things I wanted to do in this time but I definitely have had an amazing time. CE work, conference talks, Computerphile, emulators...it's been a blast.
One week til new job at Hudson River! Can't wait!!
One week til new job at Hudson River! Can't wait!!
Reposted by Q McCallum
This is *amazing*. This goes straight to Deceptive Design Hall of Shame.
They made a "take a break" nudge that has no obvious "ok, I'll take a break" affordance. Its three affordances are:
1) Keep chatting (default, highlighted)
2) x out — keeps chatting
3) "This was helpful" — what is this?
🧵
They made a "take a break" nudge that has no obvious "ok, I'll take a break" affordance. Its three affordances are:
1) Keep chatting (default, highlighted)
2) x out — keeps chatting
3) "This was helpful" — what is this?
🧵
One of the changes that OpenAI has made to make ChatGPT safer is a "take a break" nudge. There's something quite interesting about the design here. Which thing does it make you want to click?
November 24, 2025 at 2:17 PM
This is *amazing*. This goes straight to Deceptive Design Hall of Shame.
They made a "take a break" nudge that has no obvious "ok, I'll take a break" affordance. Its three affordances are:
1) Keep chatting (default, highlighted)
2) x out — keeps chatting
3) "This was helpful" — what is this?
🧵
They made a "take a break" nudge that has no obvious "ok, I'll take a break" affordance. Its three affordances are:
1) Keep chatting (default, highlighted)
2) x out — keeps chatting
3) "This was helpful" — what is this?
🧵
I would argue that the genAI industry is built on a number of big assumptions, and "will the hardware last?" is at best a secondary concern.
The data center boom is built on the assumption that AI chips will last at least five years. But we don’t actually know how long chips are good for.
The AI Industry Is Built on a Big Unproven Assumption
In addition to the projections about reaching artificial general intelligence and replacing millions of human workers, there’s the mundane matter of how long AI chips will last.
bloom.bg
November 25, 2025 at 2:00 AM
I would argue that the genAI industry is built on a number of big assumptions, and "will the hardware last?" is at best a secondary concern.
If you've enjoyed any of my longer work, that's because of my editors and reviewers.
(The same holds for plenty of other writers, I'm sure.)
A thoughtful "have you considered [subtopic X or angle Y]" can send a piece in a new and interesting direction for writer and reader alike.
(The same holds for plenty of other writers, I'm sure.)
A thoughtful "have you considered [subtopic X or angle Y]" can send a piece in a new and interesting direction for writer and reader alike.
By the way, it's National Editor Appreciation Day.
Nobody ever reads a great article and thinks, "Dang, that was some tight editing." But editing is a big part of how it got to be great.
So writers, if you've ever worked with a good editor, shout 'em out.
Nobody ever reads a great article and thinks, "Dang, that was some tight editing." But editing is a big part of how it got to be great.
So writers, if you've ever worked with a good editor, shout 'em out.
November 25, 2025 at 1:47 AM
If you've enjoyed any of my longer work, that's because of my editors and reviewers.
(The same holds for plenty of other writers, I'm sure.)
A thoughtful "have you considered [subtopic X or angle Y]" can send a piece in a new and interesting direction for writer and reader alike.
(The same holds for plenty of other writers, I'm sure.)
A thoughtful "have you considered [subtopic X or angle Y]" can send a piece in a new and interesting direction for writer and reader alike.
Reposted by Q McCallum
In the memo, “Nvidia also responds to claims that the ‘current situation is analogous to historical accounting frauds (Enron, WorldCom, Lucent) that featured vendor financing and SPVs [special purpose vehicles.)’”
@barrons.com $NVDA
@firstadopter.bsky.social
www.barrons.com/articles/nvi...
@barrons.com $NVDA
@firstadopter.bsky.social
www.barrons.com/articles/nvi...
November 24, 2025 at 7:08 PM
In the memo, “Nvidia also responds to claims that the ‘current situation is analogous to historical accounting frauds (Enron, WorldCom, Lucent) that featured vendor financing and SPVs [special purpose vehicles.)’”
@barrons.com $NVDA
@firstadopter.bsky.social
www.barrons.com/articles/nvi...
@barrons.com $NVDA
@firstadopter.bsky.social
www.barrons.com/articles/nvi...
Every tech shop claims that they run tests before a release. But how many of them _really_ test?
Do they test (and, importantly, correct) as thoroughly as this cruise line?
Do they test (and, importantly, correct) as thoroughly as this cruise line?
Before a ship’s debut, employees kick the tires in what’s known as a “shakedown.”
WSJ’s Jacob Passy boarded the Celebrity Xcel to watch the final preparations. 🔗 on.wsj.com/480LmC2
WSJ’s Jacob Passy boarded the Celebrity Xcel to watch the final preparations. 🔗 on.wsj.com/480LmC2
November 25, 2025 at 1:10 AM
Every tech shop claims that they run tests before a release. But how many of them _really_ test?
Do they test (and, importantly, correct) as thoroughly as this cruise line?
Do they test (and, importantly, correct) as thoroughly as this cruise line?
Huge move to narrow the information asymmetry in hiring.
Woah!
Ontario passed a law starting Jan 2026 that it is illegal for companies to ghost you.
If you have interviewed for a job they are legally required to inform you if they hire someone else or close the position.
The law also requires salary transparency on listings and disclosing use of AI
Ontario passed a law starting Jan 2026 that it is illegal for companies to ghost you.
If you have interviewed for a job they are legally required to inform you if they hire someone else or close the position.
The law also requires salary transparency on listings and disclosing use of AI
November 24, 2025 at 4:53 PM
Huge move to narrow the information asymmetry in hiring.
Reposted by Q McCallum
January 2025: I note that genAI's key friction stems from sellers living in the future and buyers living in the present.
(newsletter.complex-machinery.com/archive/027-...)
November 2025: investors see genAI hype as a vastly-overpriced lottery ticket that echoes of a Ponzi scheme.
(newsletter.complex-machinery.com/archive/027-...)
November 2025: investors see genAI hype as a vastly-overpriced lottery ticket that echoes of a Ponzi scheme.
The artificial-intelligence industry took Silicon Valley’s hustle mentality to an extreme—and investors are catching on.
AI Investors Want More Making It and Less Faking It
The artificial-intelligence industry took Silicon Valley’s hustle mentality to an extreme—and investors are catching on.
on.wsj.com
November 23, 2025 at 1:57 PM
January 2025: I note that genAI's key friction stems from sellers living in the future and buyers living in the present.
(newsletter.complex-machinery.com/archive/027-...)
November 2025: investors see genAI hype as a vastly-overpriced lottery ticket that echoes of a Ponzi scheme.
(newsletter.complex-machinery.com/archive/027-...)
November 2025: investors see genAI hype as a vastly-overpriced lottery ticket that echoes of a Ponzi scheme.
Great post by @webology.bsky.social that drives home key lessons about complex systems:
1/ Large incidents occur when several small, innocuous issues collide in an unfortunate manner.
2/ You don't need a large group of issues for this to happen. Just a few will do.
1/ Large incidents occur when several small, innocuous issues collide in an unfortunate manner.
2/ You don't need a large group of issues for this to happen. Just a few will do.
November 23, 2025 at 3:27 PM
Great post by @webology.bsky.social that drives home key lessons about complex systems:
1/ Large incidents occur when several small, innocuous issues collide in an unfortunate manner.
2/ You don't need a large group of issues for this to happen. Just a few will do.
1/ Large incidents occur when several small, innocuous issues collide in an unfortunate manner.
2/ You don't need a large group of issues for this to happen. Just a few will do.
Reposted by Q McCallum
Direct link to starter pack here for those want to skip the LinkedIn step: go.bsky.app/K7pijeB
November 22, 2025 at 7:15 PM
Direct link to starter pack here for those want to skip the LinkedIn step: go.bsky.app/K7pijeB
January 2025: I note that genAI's key friction stems from sellers living in the future and buyers living in the present.
(newsletter.complex-machinery.com/archive/027-...)
November 2025: investors see genAI hype as a vastly-overpriced lottery ticket that echoes of a Ponzi scheme.
(newsletter.complex-machinery.com/archive/027-...)
November 2025: investors see genAI hype as a vastly-overpriced lottery ticket that echoes of a Ponzi scheme.
The artificial-intelligence industry took Silicon Valley’s hustle mentality to an extreme—and investors are catching on.
AI Investors Want More Making It and Less Faking It
The artificial-intelligence industry took Silicon Valley’s hustle mentality to an extreme—and investors are catching on.
on.wsj.com
November 23, 2025 at 1:57 PM
January 2025: I note that genAI's key friction stems from sellers living in the future and buyers living in the present.
(newsletter.complex-machinery.com/archive/027-...)
November 2025: investors see genAI hype as a vastly-overpriced lottery ticket that echoes of a Ponzi scheme.
(newsletter.complex-machinery.com/archive/027-...)
November 2025: investors see genAI hype as a vastly-overpriced lottery ticket that echoes of a Ponzi scheme.
What, indeed.
(LBOs are for daytime reading.)
(LBOs are for daytime reading.)
November 22, 2025 at 6:08 PM
What, indeed.
(LBOs are for daytime reading.)
(LBOs are for daytime reading.)
Reposted by Q McCallum
Important read by @nickhunebrown.bsky.social . On the surface it's about genAI slop. Deeper, we have:
- trust
- how con artists tell us what we want to hear
- how even a little digging can shed light on the truth
- the extra scrutiny honest people will face because of scammers in the ranks
- trust
- how con artists tell us what we want to hear
- how even a little digging can shed light on the truth
- the extra scrutiny honest people will face because of scammers in the ranks
A few months ago @thelocal.to got a promising pitch from a writer with bylines in whole bunch of reputable publications—The Cut, The Guardian, Dwell, Architectural Digest, etc. Then I started investigating. Here's a story about fabulists in journalism's AI slop era. thelocal.to/investigatin...
Investigating a Possible Scammer in Journalism’s AI Era | The Local
A suspicious pitch from a freelancer led editor Nicholas Hune-Brown to dig into their past work. By the end, four publications, including The Guardian and Dwell, had removed articles from their sites.
thelocal.to
November 20, 2025 at 1:02 PM
Important read by @nickhunebrown.bsky.social . On the surface it's about genAI slop. Deeper, we have:
- trust
- how con artists tell us what we want to hear
- how even a little digging can shed light on the truth
- the extra scrutiny honest people will face because of scammers in the ranks
- trust
- how con artists tell us what we want to hear
- how even a little digging can shed light on the truth
- the extra scrutiny honest people will face because of scammers in the ranks
Reposted by Q McCallum
Is AI making job recruitment less meritocratic? We're getting some v interesting research studies on this question now, and the news is... not good. @jburnmurdoch.ft.com & I dive in, in the latest edition of our newsletter The AI Shift www.ft.com/content/e5b7...
November 14, 2025 at 10:13 AM
Is AI making job recruitment less meritocratic? We're getting some v interesting research studies on this question now, and the news is... not good. @jburnmurdoch.ft.com & I dive in, in the latest edition of our newsletter The AI Shift www.ft.com/content/e5b7...
Agreed. Given that a "strong forecast" is a key element of a bubble, this framing is not the big win it appears to be.
“Nvidia Gives Strong Forecast, Countering Fears of AI Bubble”
I very much disagree with this framing. People buying lots of Nvidia GPUs tells us nothing about the eventual ROI on those GPUs.
In fact, a bubble effectively <requires> “too many” of these to be bought.
I very much disagree with this framing. People buying lots of Nvidia GPUs tells us nothing about the eventual ROI on those GPUs.
In fact, a bubble effectively <requires> “too many” of these to be bought.
November 20, 2025 at 1:27 PM
Agreed. Given that a "strong forecast" is a key element of a bubble, this framing is not the big win it appears to be.
Important read by @nickhunebrown.bsky.social . On the surface it's about genAI slop. Deeper, we have:
- trust
- how con artists tell us what we want to hear
- how even a little digging can shed light on the truth
- the extra scrutiny honest people will face because of scammers in the ranks
- trust
- how con artists tell us what we want to hear
- how even a little digging can shed light on the truth
- the extra scrutiny honest people will face because of scammers in the ranks
A few months ago @thelocal.to got a promising pitch from a writer with bylines in whole bunch of reputable publications—The Cut, The Guardian, Dwell, Architectural Digest, etc. Then I started investigating. Here's a story about fabulists in journalism's AI slop era. thelocal.to/investigatin...
Investigating a Possible Scammer in Journalism’s AI Era | The Local
A suspicious pitch from a freelancer led editor Nicholas Hune-Brown to dig into their past work. By the end, four publications, including The Guardian and Dwell, had removed articles from their sites.
thelocal.to
November 20, 2025 at 1:02 PM
Important read by @nickhunebrown.bsky.social . On the surface it's about genAI slop. Deeper, we have:
- trust
- how con artists tell us what we want to hear
- how even a little digging can shed light on the truth
- the extra scrutiny honest people will face because of scammers in the ranks
- trust
- how con artists tell us what we want to hear
- how even a little digging can shed light on the truth
- the extra scrutiny honest people will face because of scammers in the ranks
Reposted by Q McCallum
In today's Complex Machinery:
The meltdown of auto conglomerate First Brands Group serves as a warning about genAI companies.
Also:
- my new book!
- Diet Bullshit, depicted
newsletter.complex-machinery.com/archive/049-...
#dataBS
The meltdown of auto conglomerate First Brands Group serves as a warning about genAI companies.
Also:
- my new book!
- Diet Bullshit, depicted
newsletter.complex-machinery.com/archive/049-...
#dataBS
#049 - Looking for a grenade in a haystack
The meltdown of auto parts company First Brands may tell us about lurking problems in the genAI space.
newsletter.complex-machinery.com
November 18, 2025 at 12:42 PM
In today's Complex Machinery:
The meltdown of auto conglomerate First Brands Group serves as a warning about genAI companies.
Also:
- my new book!
- Diet Bullshit, depicted
newsletter.complex-machinery.com/archive/049-...
#dataBS
The meltdown of auto conglomerate First Brands Group serves as a warning about genAI companies.
Also:
- my new book!
- Diet Bullshit, depicted
newsletter.complex-machinery.com/archive/049-...
#dataBS
The terms "AI" and "genAI" carry double meanings:
On one side you have the underlying technology. On the other side you have the hype, vapor, and failed promises.
I'm bullish on AI-the-technology. We need a new term to separate that from the AI-as-vapor hype wave.
Thoughts?
#dataBS
On one side you have the underlying technology. On the other side you have the hype, vapor, and failed promises.
I'm bullish on AI-the-technology. We need a new term to separate that from the AI-as-vapor hype wave.
Thoughts?
#dataBS
November 19, 2025 at 4:12 PM
The terms "AI" and "genAI" carry double meanings:
On one side you have the underlying technology. On the other side you have the hype, vapor, and failed promises.
I'm bullish on AI-the-technology. We need a new term to separate that from the AI-as-vapor hype wave.
Thoughts?
#dataBS
On one side you have the underlying technology. On the other side you have the hype, vapor, and failed promises.
I'm bullish on AI-the-technology. We need a new term to separate that from the AI-as-vapor hype wave.
Thoughts?
#dataBS
Reposted by Q McCallum
Irony is dead, but we can still talk to its genAI chatbot doppelgänger.
(It keeps saying: "too big to fail"...)
(It keeps saying: "too big to fail"...)
No firm is immune if AI bubble bursts, Google CEO tells BBC reut.rs/43D5Zl4
No firm is immune if AI bubble bursts, Google CEO tells BBC
Alphabet Chief Executive Sundar Pichai said no company would be unscathed if the artificial intelligence boom collapses, as soaring valuations and heavy investment in the sector fuel concerns of a bubble.
reut.rs
November 18, 2025 at 1:35 PM
Irony is dead, but we can still talk to its genAI chatbot doppelgänger.
(It keeps saying: "too big to fail"...)
(It keeps saying: "too big to fail"...)