Lots of ideas in this proposed research agenda for economists thinking about the effects of #AI on the #economy, including implications for #innovation and #geoeconomics. Courtesy of @erikbryn.bsky.social, @akorinek.bsky.social and Ajay Agrawal. tinyurl.com/ycxjx4ah

Shameless self-promotion alert: Director @erikbryn.bsky.social spoke with Time about how AI is–and isn’t–transforming the workplace. 5/10 time.com/charter/7316...

💭When AI automates, firms may prefer to avoid hiring new entrants than to let existing staff go 🤔 My take on @erikbryn.bsky.social & @stanfordhai.bsky.social's finding that younger workers are hit hardest by AI #AIandWork #FutureOfWork 🔗 bit.ly/4nl6G9W

Trump officials will limit intl. students to 4 years (median to finish PhD: 5.7 years). The justification? DHS claims to protect national security it hopes a threat files an extension after 4 years so USCIS reviews their application. @mclem.org @erikbryn.bsky.social‬ www.forbes.com/sites/stuart...

✍️The Lab has released a NEW STUDY with some of the first large-scale evidence of AI’s impact on entry-level employment! Lab Director @erikbryn.bsky.social, Ruyu Chen & Bharat Chandar used @ADP data to reveal 6 key facts about AI's labor market impact: digitaleconomy.stanford.edu/publications...

Why there isn’t there more discussion about accelerating the adoption of autonomous cars? There were about 40,000 Americans killed in car crashes last year. A recent study found Waymo experienced 92% fewer injury crashes with pedestrians and 85% fewer suspected serious injury crashes overall.

“We are investing so much in driving the capabilities for hundreds of billions of dollars, and we're investing very little in thinking about how we make sure that leads to widely shared prosperity,” says HAI Senior Fellow @erikbryn.bsky.social in @npr.org Morning Edition: www.npr.org/2025/08/05/n...

Lab Director @erikbryn.bsky.social spoke with @nytimes.com's Larry Sommer about tracking labor statistics -- read his thoughts at the link! www.nytimes.com/2025/08/08/b...

Following the dismissal of Erika McEntarfer as US Commissioner of Labor Statistics, we've released a statement of concern. At a time of global economic uncertainty, we stress the importance of the independence of statistical institutions and the protection of the integrity of official statistics.

"We are investing so much in driving the capabilities for hundreds of billions of dollars, and we're investing very little in ensuring that leads to widely shared prosperity." Director @erikbryn.bsky.social discussed AI and future of work with @npr.org's Steve Inskeep. www.npr.org/2025/08/05/n...

Lovely interview by @npr.org's @steveinskeep.bsky.social with @stanfordhai.bsky.social's @erikbryn.bsky.social on #AI & the future of work! 🎧 There is also featured nugget in there from @petebuttigieg.bsky.social! 🫶 www.npr.org/2025/08/05/n...

This is a terrific book. I’m reading it now and I recommend it to everyone. It’s filled with practical, science based advice

My @nytopinion.nytimes.com blog on Trump's egregious and senseless order to fire the BLS Commissioner. www.nytimes.com/live/2025/08...

The Friends of #BLS protest the firing of BLS Commissioner Erika McEntarfer. The rationale for firing her is without merit and undermines the credibility of federal economic statistics that are a cornerstone of intelligent economic decision-making by businesses, families, and policymakers.

An app that was supposed to help women be safer leaks their data, including IDs... Don't collect data you can't keep safe (and, you can't keep data safe, so don't collect it) #privacy #AIEthics www.nytimes.com/2025/07/26/u...

Some good tips for reducing your microplastic exposure wapo.st/4o35OId

This is so wrong: “For decades, tech companies have promised that AI will make our computers easier to use. That hasn’t happened yet.” I have intelligent, useful conversations with AI tools via voice every day. If someone had told me five years ago that would be possible, I’d have been blown away.

Can we really understand economic well-being if we only measure what people pay for? Catch up on @erikbryn.bsky.social's (@stanford.edu) keynote interview with @kevinfoxecon.bsky.social on GDP-B at ESCoE's recent conference: tinyurl.com/4729yakf

From @centralbanking.bsky.social: AI demands ‘deep organisational changes’ at central banks (paywall) Lab Research Scholar Sophia Kazinnik and Lab Director @erikbryn.bsky.social's new paper looks at the impact of AI on jobs within the Federal Reserve System. www.centralbanking.com/fintech/7973...

Which Workers Will A.I. Hurt Most: The Young or the Experienced? Check out this @nytimes.com piece on a pivotal time for AI in the workplace, featuring Lab Digital Fellow Sarah Bana and Postdoctoral Fellow Ruyu Chen! www.nytimes.com/2025/07/07/b...

There are hallucinations. And there are hallucinations.

Summary Table of the massive defunding of biomedical research in the United States www.nytimes.com/interactive/...

It's insane how much more popular NYC's congestion pricing got when it was implemented and people started actually seeing it work

Microsoft’s new AI model diagnoses medical cases well. I like the new benchmark where the system starts with limited information and has to ask questions or order tests to learn more. Minor quibble: a better benchmark would include some healthy cases in the mix, not just serious diseases.

OK, then. 😮 @erikbryn.bsky.social www.theguardian.com/technology/2...

Feel like there are a lot of folks on Bluesky who need to read this paper.

A.I. Might Take Your Job. Here Are 22 New Ones It Could Give You. @robcapps.bsky.social wrote an excellent article @nytimes.com with views from yours truly, @emollick.bsky.social, @erikbryn.bsky.social and others. www.nytimes.com/2025/06/17/m...

Today I spoke with @profdevisridhar.bsky.social, Chair of Global Public Health at the University of Edinburgh, about her new book How Not to Die (Too Soon): The Lies We've Been Told and the Policy Changes That Can Save Us erictopol.substack.com/p/eric-topol...

What are the tasks workers would happily offload to AIs—and how well is the technology up to it? Research by @echoshao8899.bsky.social et al suggests the match is not that good: buff.ly/pHOPmad via coauthor @erikbryn.bsky.social

A new kind of AI bot is scraping the web. 3/11 www.washingtonpost.com/technology/2...

Is it worth it to be polite to AI? 2/11 www.wsj.com/tech/ai/ai-d...

EXTREMELY BAD NEWS for economic research, per former BLS Commissioner @ericagroshen.bsky.social on LinkedIn. BLS is suspending access to its restricted data "for the forseeable future." Applies to projects through the Federal Statistical Research Data Centers & onsite projects with BLS. #EconSky

Just FYI: Martin Wolf and I will be doing a series of podcasts for the FT. Landing page here, but also on various platforms www.ft.com/wolf-krugman...

There’s no one better than Fiona on the topic of digital platform regulation. Looking forward to reading what she has to say.

From TIME, and featuring Lab Director @erikbryn.bsky.social: time.com/7290751/ai-f...

Steven Pinker, a critic of Harvard in the past, brings some perspective to the recent attacks on the university. And on science and education more broadly. www.nytimes.com/2025/05/23/o...

Good advice here www.washingtonpost.com/wellness/202...

Here’s a thoughtful and balance thread from @TimoBres.bsky.social about what went wrong and what didn’t go wrong with the apparent fraud in the recent paper about AI and scientific discovery

"The machine learning community should use centaur evaluations, in which humans and AI jointly solve tasks. Centaur Evaluations refocus machine learning development toward human augmentation instead of human replacement." Paper by Haupt and @erikbryn.bsky.social www.andyhaupt.com/assets/paper...

Check out this new podcast by two of my amazing former postdocs: Andrey Fradkin and Seth Benzell The format is that they state their priors, read a paper, discuss it on the pod, and then update their priors. It's "Justified Posteriors" 🤣🤣🤣 Subscribe here: empiricrafting.substack.com/podcast

I'm so proud of the Stanford AI Index Report. It's the best one-stop-shop for credible, current data on AI trends. hai.stanford.edu/ai-index/202...

Not a good idea to convert the leaders of the federal statistical agencies into political appointees. It would undermine trust. Send in your comment today to object to this change in the federal civil service. This article has the link. #econsky

What are the best starter packs for Economics, especially productivity, growth, innovation, labor and IO?

Internet users can seek out interactions with like-minded individuals who have similar values and become less likely to trust important decisions to people whose values differ from their own. Published 2005 www.dropbox.com/scl/fi/ji0n0... @erikbryn.bsky.social

Our future workforce could look entirely different with AI’s impact on labor. Join John Soroushian, @erikbryn.bsky.social, Brent Orwell, and Crystal Weise tomorrow for a panel on what that change could look like and what it means for policymakers. Register here: zoom.us/webinar/regi...

“Patriots walked away with the best value out of any team with six draftees ranking in Miller's top 100.” Draft rankings are published, so it would be easy just to copy the consensus draft ranking and take the best player available with each pick. Then everyone would say they “won the draft”. 1/3

The @wsj.com quotes SIEPR's @erikbryn.bsky.social on the “productivity paradox,” in which improvements in AI capabilities aren't leading to a corresponding surge in productivity. Why? Application of the tool. Read the piece: www.wsj.com/articles/com...

I wrote a letter to the editor replying to a JAMA viewpoint that said we shouldn’t experiment with being more generous to nondirected donors of kidneys for transplant. JAMA readers won’t see my letter, but you can see it below. #econsky #lawsky #medicine marketdesigner.blogspot.com/2025/04/jama...

🚨 New working paper 🚨 Demand for LLMs: Descriptive Evidence on Substitution, Market Expansion, and Multi-Homing A key question for the business of AI is the extent to which LLMs are differentiated from each other. I use data from OpenRouter to take a first look. andreyfradkin.com/assets/deman...