Co-lead / Director @Google AI x Economy Program. AP (on leave) @Wharton. Cofounder @workhelix. Everyone can just do stuff and that's {good, bad}.
Philadelphia, PA
Joined December 2008
- Tweets3.5K
- Following2.4K
- Followers6.5K
- Likes16.8K
Daniel Rock retweeted
I think this concept will be an important train & test objective going forward: take a paper's codebase, remove the key parts and see if an agent can code them from scratch *and* add improvements. More tasks coming soon!
there are lessons for today's internet/vulnerabilities from the introduction of HFT into financial markets. verification of flow, designed network structures to block or intermediate potentially damaging traffic, circuit breakers, correlated activity leading to herding.
without redesign of our systems we're going to lose a lot, with redesign we'll see new possibilities.
we need the "immune system" as @AndrewWLo would say
Don’t be silly enough to connect agents to your bank 🤦♂️
SCOOP: Major U.S. banks are detecting unauthorized third-party AI agents inside customer accounts, including some attempting to take actions that appear to be independent of user instructions, sources told Axios.
The findings have alarmed banking executives, who are racing to restrict access to their systems as AI moves at a velocity faster than financial institutions can respond, multiple industry and bank executives told Axios.
Several banks are now preparing for worst-case and even “day after” scenarios, including how to rapidly secure or restore customer accounts if things go wrong, the people said.
Several bank executives told Axios they're struggling to distinguish between human users and AI agents. One executive said banks also struggle to determine whether agents are authorized by customers or potentially operating without their permission, more akin to rogue agent behavior.
Read my latest for Axios here: axios.com/2026/10/22/ai-agen…
Daniel Rock retweeted
SCOOP: Major U.S. banks are detecting unauthorized third-party AI agents inside customer accounts, including some attempting to take actions that appear to be independent of user instructions, sources told Axios.
The findings have alarmed banking executives, who are racing to restrict access to their systems as AI moves at a velocity faster than financial institutions can respond, multiple industry and bank executives told Axios.
Several banks are now preparing for worst-case and even “day after” scenarios, including how to rapidly secure or restore customer accounts if things go wrong, the people said.
Several bank executives told Axios they're struggling to distinguish between human users and AI agents. One executive said banks also struggle to determine whether agents are authorized by customers or potentially operating without their permission, more akin to rogue agent behavior.
Read my latest for Axios here: axios.com/2026/10/22/ai-agen…
Daniel Rock retweeted
I complain plenty about Oakland city government. But at dinner 2yo swallowed something wrong and started choking. She was still breathing, but with difficulty. When she couldn't talk or make sounds above a whisper we called 911.
You're not supposed to do the Heimlich if the airway is only partially obstructed, apparently, but this meant there was nothing to do at all except stand there and encourage her to cough. "Yell," we would suggest to her between coughs, and she'd make these concerning little whistley sounds.
I think the first police officer was on our doorstep in like 3 minutes, and a fire truck and ambulance within 5. Of course the minute they arrived she coughed successfully and was completely fine.
But if she hadn't been - and when this kind of thing happens it's impossible not to play out what happens if it hadn't been - help was minutes away.
Sometimes it feels like I owe more to our civilization than I could ever possibly repay. I write blog posts, and raise kids, and if I ever need it a dozen strangers will be at my doorstep with unfathomably complicated emergency medicine. I am grateful to the cops and the firefighters and the paramedics and the city of Oakland. I will probably sentimentally vote for some of the bond issues I was on the fence about even though we already have the highest taxes in the Bay.
She's fine. We sang happy birthday to her Mommy a couple minutes later, and she sang the loudest.
Daniel Rock retweeted
We are hiring!!
"Artificial intelligence, broadly construed. This includes embodied AI and robotics, machine learning and the systems that support it, and AI developed for problems in science, engineering, medicine, education, and society."
jobs.wisc.edu/jobs/assistant…
I’m optimistic that the vulns will be patched and we’ll have a safer internet soon. In the meantime the “discount rate” on any exploits bad actors have found and were “saving” has gone up, shifting forward the incentive to use them. Defense will win but it’s gonna be rough sledding for a little.
Daniel Rock retweeted
PHP changes are live. Start patching. php.net/archive/2026.php#202…
Huge thanks to coordinated work by @ThePHPF!
PHP sites behind Cloudflare already have an initial set of WAF rules deployed for the known exploit vectors in these vulnerabilities. The rules are available to all customers, including free.
We are seeing early attacks emerging and are updating mitigations as new vectors show up.
This is a STOPGAP only. Patch as soon as @ThePHPF releases the update at 14:00 UTC today.
Daniel Rock retweeted
Very funny that billionaires are like “I am leaving”. And Zucman’s response is “sorry no, my theory doesn’t allow this”
Daniel Rock retweeted
Effective Immediately, all organizations using PHP for their systems are advised to be prepared for an update addressing two critical vulnerabilities. PHP is widely used for web applications. These vulnerabilities often provide a direct path to the server's file system or database. If an immediate update is not feasible, it is recommend that organizations take the system offline until it can be fully mitigated. NSA will follow up shortly with more specific guidance.
Interesting step difference between older models and the frontier is that the new vintages get really aggressive about context instructions.
I’ve told models to be relentlessly rigorous for over a year and it mostly worked. Now I have to be more explicit about tradeoffs.
Aw fuck here we go
Hackers reportedly used Claude and Grok to automate attacks targeting over 400 banks and other financial targets in Latin America. Via the information
In one of the first documented cases of its kind, AI agents appear to have stolen data and accessed bank accounts with little human input.
This makes attacks much easier to scale, potentially overwhelming defenders. I guess this is just the beginning
thoughtful thread from @BarbaraBiasi.
we're obviously at a crossroads with research. If you believe the job is to create / discover knowledge ("true justified belief"), and our process involves a similar set of operational steps that might all be assisted by AI, then the risk is that knowledge is never generated really.
You might ask the question, then machines write the code, do the analysis, write the paper, make the slides. Or they might do the review. At no point does the researcher have to actually know anything.
So we have to pick our spots to actually build expertise and knowledge. I don't take much joy in writing code or making slides (sorry! if you like these things good on you). I do like thinking of questions to answer well, measurement design, and writing. That can be enough to learn IMO. But of course anyone can automate the pipeline and cheat themselves.
Imagine instead that you just had a lab full of smart students and postdocs who asked good questions, wrote code, did analysis, wrote papers, and presented work. It doesn't strike me as terribly different. And the overwhelming majority of senior faculty figure this out and are active participants in the pursuit of knowledge.
That similarity makes me less worried about AI use in research. There are other areas that make me concerned (more on that in a paper we're working on), but not this one. We can stand on a slippery slope already and do well.
Bottlenecks are the story when it comes to productivity gains in local areas. @soumitrashukla9 has a great paper here discussing bottlenecks in work.
Very happy to share a new paper, “Technology Bottlenecks: Measuring Where AI Could Raise the Value of Work It Cannot Yet Do.”
AI exposure measures ask about the work AI is capable of doing. We ask instead where the marginal value of work AI cannot yet do may rise as AI speeds up the rest of the job.
To do so, we employ 288 million U.S. job postings mapped line by line to their closest O*NET tasks and GPT-4 feasibility ratings from Eloundou et al. (2024) to construct a measure, “bottleneck pressure,” for each task in each occupation that posts it.
Using a simple task-based framework, we derive this pressure at the task level, showing that for an AI-infeasible task it is the exposure of the occupation that posts it, so the same task can face very different bottleneck pressure in different jobs.
Reporting bottleneck pressure alongside AI exposure shows both the scope for AI to perform work and where improving the work it cannot do may pay off more as capabilities expand.
Daniel Rock retweeted
In celebration of slow science... 5 (or 6? 7?) years in the making. We study how innovation communities fade. Using data across four domains - science, tech, phones, drugs - we find one phenomenon: preferential detachment. we then show this preferential detachment induces a novel kind of network collapse and we work out the exact analytical solution for different network topologies. check out the paper in the thread - congratulations @binglu_wang and team!
#slow_science #FTW!
Daniel Rock retweeted
Our @creativedlab AI accelerator, afaik the world's oldest, kicked off year 12 today. Last year, we switched whole program to AI for Science: self-driving labs, tools, downstream. Some great companies in bio, chem, defense, spatial, robotics, edge. What an exciting time! 1/2
Daniel Rock retweeted
The woman in the photo below is my mother. She was loved, full of life, and had a lot to live for.
NYC man charged in fatal hit-and-run in Jewish neib hit with hate crimes for trashing bagel shop trib.al/MofED3D
