
AI, work, and what comes next. (Very) begrudgingly back on this platform. Most recently Senior Fellow at @BrookingsInst; building something new.
Washington D.C., DC
Joined June 2025
- Tweets139
- Following301
- Followers2K
- Likes504
Molly Kinder retweeted
Jobs could be the most important youth vote issue in 2028.
Young people around the world are not okay. Incredible graphs from WSJ
1. Unemployment rates for young college grads have been climbing compared with the broader population for years
2. AI might be exacerbating the employment challenges for young ppl in software
3. In Italy and Spain, more than 80% people in their 20s live with their parents (!?!!?)
4. Youth unemployment in China has gone from non-existent to one of the highest in the world during a period when its poverty rate has plunged. Weird.
Molly Kinder retweeted
My latest in the @nytimes: the AI scenario we should pay more attention to is a boom paid for out of the $8–10 trillion US white-collar wage bill nytimes.com/2026/11/17/opini…
Huge news! @jasminewsun is a singular voice on AI right now. Her Substack and writing has been must reads for countless of us over the past year. As a @nytimes subscriber, I am delighted the paper scooped up a generational talent on the defining issue of our time. Huge kudos to Jasmine -- can't wait to follow along in print/ podcast/ in my inbox.
Personal news: I just joined @NYTOpinion! I’m starting a new AI newsletter, writing on tech for the editorial board, and appearing on podcast feeds near you.
I’ve had the best two years of independence & it was not my plan to give that up. But I believe the NYT is the best place to shape the AI conversation as it stretches beyond Silicon Valley and into homes, workplaces, laws, and lives everywhere.
There’s so, so much to write about, and my first story will be out this week.
(As I onboard, doing a bunch of background learning convos from folks around the ecosystem; also always open to tips, feedback, etc - say hi via DM or on Signal at jws.27!)
Molly Kinder retweeted
You don’t have to love powerlooms or hate hand weaving to recognize a powerloom when you see one.
Lots of intellectual tasks that used to be pure craft & which, therefore, could never be industrialized might soon be subject to the logic of industrialization. Not an easy change.
Molly Kinder retweeted
The challenges posed by #ArtificialIntelligence and unregulated automation can put countless jobs at risk and become a real threat to proper integral human development. Technology must always be subject to ethics; an algorithm cannot replace human discernment or a worker's value. Access to work for all must be a high priority for public policies and economic processes, serving as a criterion for evaluating the human quality of any development model. #ApostolicJourney #Argentina
Molly Kinder retweeted
AI is taking jobs in some industries, even if it's not destroying jobs overall.
Molly Kinder retweeted
I've always had substantially longer timelines than my friends at Anthropic — I've been concerned that they're in an echo chamber and believing the hype about their own models too much, and that they're underestimating bottlenecks and unknown unknowns. And there have definitely been specific occasions when they have overstated how much uplift they were getting from their models. Nonetheless, if METR is correct here (and I think they are among the most trustworthy sources on this topic), then my Anthropic friends have been much more right about timelines than me.
This report is also some evidence that a software-only singularity is plausible. I now think that we are more than 50% likely to have full blown superintelligence in the next 18 months, absent a slowdown. I need to call my family.
Molly Kinder retweeted
Most see the election results as driven by affordability, Iran, and corruption. @BetterFuture_AI makes a different case: AI is the culprit!
Just as you can get a sense of the future AI capabilities by extrapolating lines on the graph, folks would be wise to note the rise in salience/voter dissatisfaction outlined here. I disagree that 2026 was "the first AI election" but it's not a bad bet that 2028 will be dominated by AI politics.
2026 was the first AI election, but it won’t be the last.
Our data shows super intelligence was a major factor in the significant losses for Republicans, and winning back voters’ trust requires course-correcting now.
Our @BetterFuture_AI memo breaks down the numbers. Some highlights:
1) AI had the “most significant growth of any issue” this cycle.
Less than a year ago, only 7% of voters called AI a top priority for the government. Now that number is 21%, representing the most significant growth of any issue tested. No other issue moved more than 4%. NBC reported that over half of voters polled (56%) called AI and datacenters a crucial or major factor in their vote. When CNBC asked voters their top fear if Republicans keep Congress, 18% said failure to regulate AI and data centers. This concern tied with the war in Iran.
2) Voters are unified: they want Big Tech accountability.
Voters have made their anger about SI/AI clear. An October 27 survey of Ohio and Iowa voters found that 67% say we need more safety measures, even at the risk of losing to China. 91% of Americans consider it important for the U.S. to establish safeguards for AI (Quinnipiac). A separate national survey released by Reuters/Ipsos on September 22 found 73% say AI companies haven't done enough to prevent serious harm. Crucially, 61% of Republicans support stricter government action on AI.
3) Voters don’t trust Republicans on AI. They think the GOP is too close with Big Tech.
Voters are worried about AI, but Republicans didn’t effectively address those concerns. That’s why Democrats now hold a 9-point trust advantage on it (Navigator) and 47% of voters call Republicans the party of tech tax breaks, more than double who say the same of Democrats (Politico). The 2028 election cycle starts now. Republicans must act quickly to gain trust back.
4) Democrats weaponized SI against Republicans, and Republicans didn’t fight back.
Democrats seized on SI/AI as a political vulnerability for Republicans, outspending Republicans 9 to 1 on AI and data center ads, and winning 8 out of the 9 races where ads were run. In Ohio alone, Dems spent $30M portraying Jon Husted as a corrupt Big Tech ally, and he paid the price.
5) It’s not too late to course-correct for 2028, but Republicans must address voter concerns quickly.
SI/AI can transform Americans’ lives for the better, but only if people believe it is trustworthy and secure. We need to act now to make this happen:
✅ Protect children: Pass legislation to ensure that chatbots never addict children, create unhealthy attachments, or coach children towards sexual behavior or suicide.
✅ Defend America’s security: Require national security protections like robust controls and monitoring, incident reporting, auditors, and independent incident reviews to address unintended SI behaviors.
✅ Strengthen the economy: Implement measures that prevent workers from being left behind and require data center companies to pay their own way on power costs.
✅ Benefit everyday Americans: Encourage developers to spend more resources on curing diseases and making products cheaper, and less on experiments that only benefit the developers.
It would be political malpractice for Republicans not to course-correct on the most powerful technology of our lifetime and the issue that will headline the 2028 cycle. While the recent midterm election was about more than just AI, 2028 will not be. We need to get real on balancing innovation with security, and we need to do it fast.
Lots of AI people made it low status to comment on shortfalls and failures of models, or the fact that they're not reallty automating jobs fully and so on. Any scrutiny means you're not 'AGI pilled' (which is not a claim about technology but sometimes a quasi-religious one about its impacts and adoption). It's a shame because it's actually very interesting to explore the spikiness, jaggedness, failures, complementarities and so on; but for a lot of people this is almost a waste of time because 'something something future models!1!!' means they get to just parrot 'this is the worst the models will ever be' whilst dismissing any crumb of empiricism away because they've already worked out the future based on theory and projections alone.
To be fair I sympathise with them to a degree because the 'other sides' aren't any better, essentially over-indexing on short-term failures (Rs in strawberry etc) and extrapolating from these to dismiss the technology completely. But I care less about what these people think about capabilities since time just proves them wrong rapidly enough. Why would you exercise vibe arbitrage over such a cheap target? I also find it painfully boring to constantly litigate 'AI skeptics' - there are more interesting and thornier discussions to be had.
Molly Kinder retweeted
GPT-4 could do 4 minutes of independent work with a 50% chance of success.
Current models can do multiple weeks of work, and January 2026 is a reasonable cut-off for the moment agentic work started.
Large productivity impacts before that would mostly be confined to coding.
Molly Kinder retweeted
My views on AI have radically shifted over the past few weeks, after spending way too much time reading about it and watching interviews of industry leaders and experts.
First, I really love the tools. I spend $200 on Claude every month and the product is worth every penny for me. Probably even more. I think the existing tools can deliver significant increase to future GDP and help us solve major global challenges faced by humanity.
However, I just cannot comprehend why the industry is not much more regulated, as its leaders and employees seem to request themselves. I meet a lot of people who have heard of the hugging face incident but don’t know the actual chain of events. If you don’t, you definitely should spend a least an hour reading about it.
I also do not know why we need to follow the path that frontier labs are following, to develop ever more powerful models. It seems to me that they are enormous gains that are readily available if we deploy the current technology in economic and scientific organizations without pursuing a super intelligence that they say might kill us and more likely wreak havoc in society. I know a lot of people think these warnings are marketing tricks, and that’s a reasonable assumption. But I don’t see the point of taking such risks.
My median scenario is that of many economists: AI will diffuse slowly and lift gdp growth by a percentage point or so, will cause some economic disruption on the way, but ultimately people will get new jobs and maybe work less.
However, I am deeply uncomfortable with the absolutely wild level of uncertainty around this prediction. And to be honest, even setting the doom scenarios aside, I find the utopia scenarios of some industry leaders actually dystopian. And I think a large majority of the public would agree if they paid attention to what they’re saying. To their credit, they themselves agree publicly that they should not be the ones making these choices. A large majority of the public is now very skeptical of AI as a source of human progress, in particular among younger adults who use it the most. They should have a say.
I don’t want to convince you of anything but to get informed about this topic, express yourself and vote accordingly.
Molly Kinder retweeted
So I agree with a lot of what Jason has written here…but I do want to take on the question of hiring juniors.
I think the skillset of which juniors you’ll hire will shift but also that (in the specific case of CEA) it should have shifted anyway. (This is about CEA specifically, not the economy overall).
The most effective CEA juniors were like the most effective CEA seniors in that they understood how to work the building and get themselves into policy processes. Yes, even the juniors.
✍️ New from Jason Furman: AI Could Do My First Job. Not My Last.
Why I still think AI could do most of what the Council of Economic Advisers did, and what that means for the people who learn by doing that work.
econvitals.org/posts/2026-11…
I can't stop thinking about this spot on Wile E. Coyote metaphor from architect Stephen Gallagher.
For many professionals testing these tools in their own craft, AI models can feel mediocre, mediocre, mediocre… and then poof, all at once, scary good.
First coders and creatives, now mathematicians and scientists (social scientists included). I felt glimpses of this in my former research role at Brookings.
Stephen’s excellent essay tracks this trajectory in architecture through the lens of technical drawing, and shows a profession now suspended in midair over the canyon, legs still running, not yet looking down.
Stephen's full essay: stephenjgallagher.substack.c….
2
2
588
Nearly two years ago, I interviewed a woman who runs a successful modeling agency.
She started as a receptionist and worked her way up. Most of their work was shooting models for catalogs. Each shoot required an entire cast: the model, yes, but also the stylist, makeup artist, photographer, lighting.
At the time, I was at Brookings, poring over exposure data from OpenAI. In that data, fashion models didn’t look “exposed” to AI at all. Models couldn’t use ChatGPT to save time on doing their job tasks, like a software engineer or social media manager could.
But the woman who ran the modeling agency had no hesitation. She saw what was coming (per this video) a million miles away. She posited: Why would the Nordstroms of the world pay hundreds of thousands of dollars for a shoot if they can just use AI models? She told me it was bleak. If she wasn’t so close to retirement, she would have been distraught, especially if she was in the thick of raising her kids.
I have never forgotten that conversation. Whenever I would hear someone repeat, tritely, “you won’t lose your job to an AI. You will lose your job to someone else using AI,” I would think of her. All the AI skills in the world weren’t going to save her livelihood and the business she had poured blood, sweat and tears into.
The data we are using to track “exposure” misses an example like this when AI is so disruptive the whole industry shifts. The makeup artist, the photographer, the stylist, the lighting person: AI might not be able to be used to replace those jobs directly. But if AI simply automated modeling, it could take demand for those roles with it.
1,000% this.
@kevinroose and @CaseyNewton are spot on. We do not have enough good answers right now on what to do when AI takes jobs. Not from the labs, not from policymakers. Everyone is anxious, but no one (yet) has a plan.
(Quietly whispers: this is why I quit my job at Brookings to build something new. More to share in a few weeks...)
In the meantime, I just wrote about AI's impact o mathematicians, why I cried about it, and what we are getting wrong in our response, in my latest Substack: substack.com/home/post/p-221…
What happens when AI comes for the job you planned to spend 30 years doing? Mathematicians and scientists are already finding out.
@KevinRoose and @CaseyNewton on why governments and AI labs still have no answer.
New Machine Gods with @NPR is out now.
15
1,268































