Chris Chapman
@cchapman.bsky.social
UX researcher, psychologist. Author "Quantitative User Experience Research" (w/Rodden), "R | Python for Marketing Research and Analytics" (w/Feit & Schwarz). Previously 24 yrs @ Google, Amazon, Microsoft. Personal account.
Blog at https://quantuxblog.com
Blog at https://quantuxblog.com
Reposted by Chris Chapman
almost every novelist you read who's not from the airport fiction or bestseller sections is working a full on day job, including multi-award winners and absolute legends, or they have a spouse or trust that supports them. virtually across the board. just normalize in your mind that's how it works.
I know we know this, but authors need to get better at talking about how they actually make their money. Stop pretending it's from the books.
November 25, 2025 at 10:35 PM
almost every novelist you read who's not from the airport fiction or bestseller sections is working a full on day job, including multi-award winners and absolute legends, or they have a spouse or trust that supports them. virtually across the board. just normalize in your mind that's how it works.
Reposted by Chris Chapman
“AI workers said they distrust the models they work on because of a consistent emphasis on rapid turnaround time at the expense of quality.”
Meet the AI workers who tell their friends and family to stay away from AI
When the people making AI seem trustworthy are the ones who trust it the least, it shows that incentives for speed are overtaking safety, experts say
www.theguardian.com
November 22, 2025 at 5:12 PM
“AI workers said they distrust the models they work on because of a consistent emphasis on rapid turnaround time at the expense of quality.”
Whenever you see research from industry, consider not only that the research was likely built with self-serving plans & analyses ... but also that negative results are cut off, defunded, or buried as soon as possible.
November 24, 2025 at 4:03 PM
Whenever you see research from industry, consider not only that the research was likely built with self-serving plans & analyses ... but also that negative results are cut off, defunded, or buried as soon as possible.
Product folks, don't just "add AI to it".
Amazon example, where GenAI is now part of search. It failed the first time I saw it, on an extremely simple brand search for "fluent pet".
One might argue that it doesn't matter; results are stull there. That is, unless you care about customer trust 🤷
Amazon example, where GenAI is now part of search. It failed the first time I saw it, on an extremely simple brand search for "fluent pet".
One might argue that it doesn't matter; results are stull there. That is, unless you care about customer trust 🤷
November 24, 2025 at 3:52 PM
Product folks, don't just "add AI to it".
Amazon example, where GenAI is now part of search. It failed the first time I saw it, on an extremely simple brand search for "fluent pet".
One might argue that it doesn't matter; results are stull there. That is, unless you care about customer trust 🤷
Amazon example, where GenAI is now part of search. It failed the first time I saw it, on an extremely simple brand search for "fluent pet".
One might argue that it doesn't matter; results are stull there. That is, unless you care about customer trust 🤷
Reposted by Chris Chapman
🔥
I once knew a Federal bank examiner, and one time someone asked him why we had to KEEP inspecting banks over and over. He basically said every new batch of business school grads invents bank fraud from first principles.
AirBnB CEO calling it “vibe revenue” just 👨🍳 😘
The underlying cause of every bubble - debt masquerading as financial innovation - depends on not just short financial memory & speculative neophytism, but reinventing jargon of finance, like how each generation of kids has new ways to say same things.
The underlying cause of every bubble - debt masquerading as financial innovation - depends on not just short financial memory & speculative neophytism, but reinventing jargon of finance, like how each generation of kids has new ways to say same things.
November 23, 2025 at 12:35 AM
🔥
Reposted by Chris Chapman
"The test of our progress is not whether we add more to the abundance of those who have much; it is whether we provide enough for those who have too little."
-FDR, 1937
-FDR, 1937
November 22, 2025 at 6:31 PM
"The test of our progress is not whether we add more to the abundance of those who have much; it is whether we provide enough for those who have too little."
-FDR, 1937
-FDR, 1937
Reposted by Chris Chapman
The malwarebytes Google opting-your-emails-into-ai-training thing is not true. The post is based on one bad tweet.
I know anti-tech lamenting is a Bluesky core principle, but please use the same judgement you would for any other clickbait.
www.theverge.com/news/826902/...
I know anti-tech lamenting is a Bluesky core principle, but please use the same judgement you would for any other clickbait.
www.theverge.com/news/826902/...
Google denies ‘misleading’ reports of Gmail using your emails to train AI
Google says “we do not use your Gmail content for training our Gemini AI model.”
www.theverge.com
November 22, 2025 at 8:58 AM
The malwarebytes Google opting-your-emails-into-ai-training thing is not true. The post is based on one bad tweet.
I know anti-tech lamenting is a Bluesky core principle, but please use the same judgement you would for any other clickbait.
www.theverge.com/news/826902/...
I know anti-tech lamenting is a Bluesky core principle, but please use the same judgement you would for any other clickbait.
www.theverge.com/news/826902/...
Reposted by Chris Chapman
*states at Arizona*
climate change comes for us all at some point. this is wild.
climate change comes for us all at some point. this is wild.
'Iran’s capital must be moved because the country “no longer has a choice,” President Masoud Pezeshkian said on Thursday in remarks carried by state media, warning that severe ecological strain has made Tehran impossible to sustain'
#Iran 🇮🇷
#Iran 🇮🇷
Iran president says capital move now a necessity as water crisis deepens
Iran’s capital must be moved because the country “no longer has a choice,” President Masoud Pezeshkian said on Thursday in remarks carried by state media, warning that severe ecological strain has mad...
www.iranintl.com
November 21, 2025 at 11:22 AM
*states at Arizona*
climate change comes for us all at some point. this is wild.
climate change comes for us all at some point. this is wild.
Reposted by Chris Chapman
I am dreading how all this will spread - because if Amazon is already this bad, every other business will be even worse
Amazon's #1 leadership principle used to be "customer obsession." So they still make it easier to reach a human - but already harder than it was before AI
Amazon's #1 leadership principle used to be "customer obsession." So they still make it easier to reach a human - but already harder than it was before AI
November 21, 2025 at 12:57 PM
I am dreading how all this will spread - because if Amazon is already this bad, every other business will be even worse
Amazon's #1 leadership principle used to be "customer obsession." So they still make it easier to reach a human - but already harder than it was before AI
Amazon's #1 leadership principle used to be "customer obsession." So they still make it easier to reach a human - but already harder than it was before AI
Reposted by Chris Chapman
In honor of Larry Summers, I asked the female members of my college class (1987) to say if they had any experiences with sexual harassment by teachers/professors, back when this was tolerated as, I dunno, the cost of attending college, and OH MY GOD.
November 21, 2025 at 12:13 AM
In honor of Larry Summers, I asked the female members of my college class (1987) to say if they had any experiences with sexual harassment by teachers/professors, back when this was tolerated as, I dunno, the cost of attending college, and OH MY GOD.
"Adversarial poetry" is the most delightful phrase I've heard in a long time
(and what a great use for it)
(and what a great use for it)
Looks like LLMs are *very* vulnerable to attack via poetic allusion: "curated poetic prompts yielded high attack-success rates (ASR), with some providers exceeding 90% ..."
https://arxiv.org/html/2511.15304v1
https://arxiv.org/html/2511.15304v1
November 20, 2025 at 11:40 PM
"Adversarial poetry" is the most delightful phrase I've heard in a long time
(and what a great use for it)
(and what a great use for it)
Reposted by Chris Chapman
I keep explaining that tech layoffs are orgs buying permission from Wall Street to do things that include building out AI data centres; even with the R&D credits gone the salaries don't add up to the capex, they're just paying the toll
November 20, 2025 at 9:54 PM
I keep explaining that tech layoffs are orgs buying permission from Wall Street to do things that include building out AI data centres; even with the R&D credits gone the salaries don't add up to the capex, they're just paying the toll
Reposted by Chris Chapman
AI companies need $2 trillion in annual revenue by 2030. It’s not going to happen. Big tech firms are borrowing more than ever and they don’t have a way to pay their creditors. @ddayen.bsky.social reports: trib.al/Y77U9j9
The AI Bubble Is Bigger Than You Think - The American Prospect
It’s not just OpenAI that looks overhyped. There’s a whole mountain of sketchy financial engineering underneath.
trib.al
November 19, 2025 at 8:15 PM
AI companies need $2 trillion in annual revenue by 2030. It’s not going to happen. Big tech firms are borrowing more than ever and they don’t have a way to pay their creditors. @ddayen.bsky.social reports: trib.al/Y77U9j9
Reposted by Chris Chapman
just being out at night if you’re anywhere near roads is like actively painful at this point
I wish we, like, regulated shit nowadays. Between car bloat, blinding headlines, and neighbourhoods packed with rental suites but no on-property parking, driving is an absolute nightmare now.
Nearly all drivers say vehicles' lights are too bright in study
The study, commissioned by the Department for Transport, was completed by Berkshire's TRL.
www.bbc.com
November 19, 2025 at 3:06 AM
just being out at night if you’re anywhere near roads is like actively painful at this point
Reposted by Chris Chapman
Things that will get you kicked out of academia forever:
- taking maternity leave at the wrong time
- spending too much time with your kids
- reporting harassment
- not moving every 2-3 years
- taking a partner's job/preferences into account
- mouthing off before tenure
- taking maternity leave at the wrong time
- spending too much time with your kids
- reporting harassment
- not moving every 2-3 years
- taking a partner's job/preferences into account
- mouthing off before tenure
A guy makes ONE tiny mistake (has a years-long friendship with the world's worst sex trafficker; brags about sexually harassing colleagues; is racist; says women are stupid) and his whole LIFE is blown up (does slightly fewer speaking engagements; keeps teaching at #1 university)??!?!?!?!?!
So Harvard is keeping this guy, but Claudine Gay had to step down over ginned up plagiarism accusations and bad-faith accusations of anti-Semitism.
Got it.
Got it.
November 18, 2025 at 3:52 PM
Things that will get you kicked out of academia forever:
- taking maternity leave at the wrong time
- spending too much time with your kids
- reporting harassment
- not moving every 2-3 years
- taking a partner's job/preferences into account
- mouthing off before tenure
- taking maternity leave at the wrong time
- spending too much time with your kids
- reporting harassment
- not moving every 2-3 years
- taking a partner's job/preferences into account
- mouthing off before tenure
Reposted by Chris Chapman
new paper by Sean Westwood:
With current technology, it is impossible to tell whether survey respondents are real or bots. Among other things, makes it easy for bad actors to manipulate outcomes. No good news here for the future of online-based survey research
With current technology, it is impossible to tell whether survey respondents are real or bots. Among other things, makes it easy for bad actors to manipulate outcomes. No good news here for the future of online-based survey research
November 18, 2025 at 7:16 PM
new paper by Sean Westwood:
With current technology, it is impossible to tell whether survey respondents are real or bots. Among other things, makes it easy for bad actors to manipulate outcomes. No good news here for the future of online-based survey research
With current technology, it is impossible to tell whether survey respondents are real or bots. Among other things, makes it easy for bad actors to manipulate outcomes. No good news here for the future of online-based survey research
Reposted by Chris Chapman
If you email us proof of cancellation of your New York Times subscription (send to [email protected]), we’ll give you a FREE year-long digital subscription to Current Affairs.
NYT continues its long history of publishing op-eds pushing criminal wars of aggression
November 17, 2025 at 8:38 PM
If you email us proof of cancellation of your New York Times subscription (send to [email protected]), we’ll give you a FREE year-long digital subscription to Current Affairs.
Reposted by Chris Chapman
The problem with relying on pithy observations like “history tends to repeats itself or rhymes” is that history is long and has many stanzas. It’s impossible to know in advance with any certainty whether tomorrow is going to rhyme with 2008, 1929, or the fall of the Roman Empire.
November 18, 2025 at 10:08 AM
The problem with relying on pithy observations like “history tends to repeats itself or rhymes” is that history is long and has many stanzas. It’s impossible to know in advance with any certainty whether tomorrow is going to rhyme with 2008, 1929, or the fall of the Roman Empire.
Reposted by Chris Chapman
Make bad art. Make uncertain art. Make indifferent art. Make confused, and confusing, art. Make uninspired art. Making bad art is the essential and irreplaceable aspect of making great art, next time
make bad art because some of the worst people you know are making even worse art and literally nothing you make will be that bad.
make bad art!!! it's better than making no art!!!
November 17, 2025 at 4:06 PM
Make bad art. Make uncertain art. Make indifferent art. Make confused, and confusing, art. Make uninspired art. Making bad art is the essential and irreplaceable aspect of making great art, next time
Reposted by Chris Chapman
My #rstats cheat code for today is the binom.confint function in the binom package that will spit out *12* different ways of calculating a CI for a proportion.
Also, this is why you use R for statistics...
(and of course the correct CI method is bayes 😎)
Also, this is why you use R for statistics...
(and of course the correct CI method is bayes 😎)
November 16, 2025 at 2:52 AM
My #rstats cheat code for today is the binom.confint function in the binom package that will spit out *12* different ways of calculating a CI for a proportion.
Also, this is why you use R for statistics...
(and of course the correct CI method is bayes 😎)
Also, this is why you use R for statistics...
(and of course the correct CI method is bayes 😎)
Reposted by Chris Chapman
Data-centers make Page One in Washington State.
“.. they cost millions if not billions to build, consume vast amounts of water and power, and generally do not employ significant numbers of people beyond the construction phase ..”
“.. they cost millions if not billions to build, consume vast amounts of water and power, and generally do not employ significant numbers of people beyond the construction phase ..”
November 16, 2025 at 2:27 PM
Data-centers make Page One in Washington State.
“.. they cost millions if not billions to build, consume vast amounts of water and power, and generally do not employ significant numbers of people beyond the construction phase ..”
“.. they cost millions if not billions to build, consume vast amounts of water and power, and generally do not employ significant numbers of people beyond the construction phase ..”
Reposted by Chris Chapman
Language models are models of a corpus, not cognition
I do, of course, have a paragraph from "Why We Fear AI" for this
November 15, 2025 at 11:35 AM
Language models are models of a corpus, not cognition
Reposted by Chris Chapman
Oven makers pimping their “interactive color touch screen” and all I really want is a big freaking knob. Do you know how many computers I want in my oven? Zero. I want zero computers in my oven.
November 15, 2025 at 3:23 PM
Oven makers pimping their “interactive color touch screen” and all I really want is a big freaking knob. Do you know how many computers I want in my oven? Zero. I want zero computers in my oven.
Reposted by Chris Chapman
WSJ: “.. Can we even build all the necessary physical infrastructure? And if so, will the resulting AI-powered products generate enough revenue to pay back that investment?”
@wsj.com 🤡
www.wsj.com/tech/ai/when...
@wsj.com 🤡
www.wsj.com/tech/ai/when...
November 14, 2025 at 1:22 PM
WSJ: “.. Can we even build all the necessary physical infrastructure? And if so, will the resulting AI-powered products generate enough revenue to pay back that investment?”
@wsj.com 🤡
www.wsj.com/tech/ai/when...
@wsj.com 🤡
www.wsj.com/tech/ai/when...
Reposted by Chris Chapman
We didn't plagiarize, you made us plagiarize by asking questions to which we stole the answers.
"Because its output is generated by users of the chatbot via their prompts, OpenAI said, they were the ones who should be held legally liable for it – an argument rejected by the court."
"Because its output is generated by users of the chatbot via their prompts, OpenAI said, they were the ones who should be held legally liable for it – an argument rejected by the court."
ChatGPT violated copyright law by ‘learning’ from song lyrics, German court rules
OpenAI ordered to pay undisclosed damages for training its language models on artists’ work without permission
www.theguardian.com
November 14, 2025 at 2:47 AM
We didn't plagiarize, you made us plagiarize by asking questions to which we stole the answers.
"Because its output is generated by users of the chatbot via their prompts, OpenAI said, they were the ones who should be held legally liable for it – an argument rejected by the court."
"Because its output is generated by users of the chatbot via their prompts, OpenAI said, they were the ones who should be held legally liable for it – an argument rejected by the court."