Excited to share I've just joined @Google on the new Econ x AI research team. Excited to follow this space for your research and please feel free to connect with any interesting work!

Today, we’re continuing to push the boundaries of AI with our release of Gemini 3.1 Pro. This updated model scores 77.1% on ARC-AGI-2, more than double the reasoning performance of its predecessor, Gemini 3 Pro. Check out the visible improvement in this side-by-side comparison, showing Gemini 3.1 Pro’s crisp animation built with pure code. Read more about today’s 3.1 Pro update: blog.google/innovation-and-a…

The first cohort of Stripe econ of AI fellows: ( thread, with some thoughts on the state of the field)

The past 18 months have seen the most rapid change in human written communication ever By. September 2024, 18% of financial consumer complaints, 24% of press releases, 15% of job postings & 14% of UN press releases showed signs of LLM writing. And the method undercounts true use

🚨 New research: We analyzed 1.5M documents to track LLM-assisted writing adoption across society from 2022-2024. The results? 📊By late 2024, LLMs assist in writing: - 18% of financial consumer complaints - 24% of corporate press releases - Up to 15% of job postings (esp. in small/young firms) - 14% of UN press releases All showing similar adoption trajectories despite diverse contexts. PDF: arxiv.org/abs/2502.09747 🧵

Entrepreneurship PhD students: apply for the 2024 @nberpubs Entrepreneurship Research Boot Camp (week of July 15th). My student experience 10 (!) years ago was instrumental to my career. Thanks to @NBER and @robinson_fuqua for keeping it going. conference.nber.org/confsubm…

"Rationalizing Entrepreneurs’ Forecasts" by Mihai Codreanu @m_codreanu Nick Bloom @I_Am_NickBloom & Robert Fletcher @StanfordEcon @Instacart

Sorry to learn that Max Corden has passed away. His book "Trade Policy and Economic Welfare" was an inspiration, and the idea of effective protection is a lasting part of the trade canon

Conference on Field Experiments in Strategy (CFXS) is off to an amazing start @HarvardHBS with @hyunjinvkim walking through her design choices for our PhD workshop.

Some lessons of the insane past 4 days of generative AI, as someone who had access to Bing during and after the "Sydney" era. (Trying this as a long tweet rather than a thread...) 1) Bing AI was two things: a chatbot and an evolution of ChatGPT into a web-connected, supercharged form. There is no reason these two things had to be connected, but they were. 2) The new Bing AI version of search and retrieval (without the chatbot) is much more powerful than ChatGPT. It has some of the same issues (like hallucination and terrible math) but less so, and is capable of some really extraordinary tasks. When I put it to the test, it can do things like read multiple research papers and identify gaps; improve its own writing by asking it to look at online examples of good writing; and do complex analyses integrating diverse information. The work was really, really impressive. 3) The Bing Chatbot was often unsettling. I say that as someone who knows that there is no actual personality or entity behind a LLM model. But, even knowing that it was basically auto-completing a dialog based on my prompts, it felt like you were dealing with a real person. I never attempted to "jailbreak" the chatbot or make it act in any particular way, but I still got answers that felt extremely personal, and interactions that made the bot feel intentional. 4) The lesson of the Chatbot was that we can very easily be fooled by an AI into thinking it is sentient. It isn't just Turing Test passing, it is eerily convincing even if you know it is a bot, and even at this very early stage of evolution. Even if Bing isn't doing this anymore, there is no doubt other AI bots will come along, and may already be deployed (I assume governments have LLMs at the level of Bing, but with less guardrails). We should be considering about what that means. 5) The lesson of the Bing AI version of ChatGPT is that many of the things we thought AI would be bad at for awhile (complex integration of data sources, "learning" and improving by being told to look online for examples, seemingly creative suggestions based on research, etc.) are already possible. There is no doubt it will have a large effect on anyone doing information-based work. Early AI assistants, like Copilot, already cut the time for complex tasks like coding in half. This will do the same, or more, across many industries. I think every organization that has a substantial analysis or writing component to their work will need to figure out how to incorporate these new tools fast, because the competitive advantage gain is potentially enormous. And there is no instruction manual. You can only learn through trial-and-error. We got a glimpse of the future in the past few days, and the gap between ChatGPT (which is already causing waves in many industries) and Bing AI remains enormous. I was not expecting things in AI to keep moving this fast, but now there is every indication they will continue to do so. I don't think anyone knows what this all means, but I think we should be ready for a very weird world.

I mean, seriously, what other conferences do you know that include as a social event "trip to Dracula's castle"? (and resemblance with contemporary characters is purely fortuitous)

My co-author @TomWatersEcon is excited to speak about our paper on work search requirements for single parents in the Grand Salon. [If you want to catch me I'll be mostly Churchill 413 for the next few days. Or text to grab a drink!] #ASSA2023

Yesterday I wrote my first lines of code with ChatGPT. It saves so much time for tasks you know how to do & it's truly transformative for what it means to be a good coder I almost understand why you'd get so excited to spend Christmas programming with it 😅