micro1

On Wednesday , September 9th at 12PM PT, I'll be sitting down with Frank Nagle, first Chief AI Economist at Microsoft for "Decoding the Economics of AI: A Chief Economists' Dialogue," part of the micro1 forum. Two chief economists, two vantage points on the same question: what is AI actually worth, and to whom. Frank's work on the economics of open source and digital infrastructure has shaped how a lot of us think about value that doesn't show up cleanly on a balance sheet, which makes this a conversation I've wanted to have for a while. We'll get into where AI's economic value is really accruing, what current measurement approaches miss, and where the incentives are pointing next. If you work anywhere near the intersection of AI strategy and economics, this is a dialogue worth carving out some time for. Register here: https://lnkd.in/gp_kcFwR

We are hiring 10,000 robotics trainers over the next 7 days for one of our biggest projects yet We're looking for thousands of participants to help train robots by reviewing and labeling videos of them performing tasks - Role is open to English speaking and Western countries - No prior AI experience required - You'll review and label videos of robots performing tasks to help them improve - Hourly rate: $50-90/hr - Referral bonus: $300 per successful referral If you're interested in applying or referring, use the links in comments

we’re hiring 10,000 robotics trainers in the next 7 days. $50–$90/hour, accepting applicants globally. you’ll review and label videos of robots performing tasks to help them improve. no prior AI experience required. an entirely new category of work is emerging around teaching robots how to interact with the world. application link in the comments below.

fun scroll for our experts: micro1.ai/human :)

When I first took on a teaching appointment at Harvard back in 2011, it felt like exactly the right place to be pushing on the questions I cared about. Years on, the balance had shifted. The opportunities to contribute elsewhere had become too significant to keep dividing my attention, so I made the decision to step down from that role and focus more deeply on the work we’re building at micro1. As a scholar, I’ve always believed there is real value in moving between scholarship and industry, with each sharpening the questions you bring to the other. At this stage, preserving that balance meant giving the industry side of my work the focus it now demands. The questions we’ve been circling as a team aren’t the kind you can properly chase on the side. They require sustained attention. A few of the threads we’re pulling on: 1) Human data as a factor of production. How do the quantity, quality, and price of human-generated data shape AI capability? We understand compute as an input to intelligence; we understand the economics of human data far less well. 2) The economics of evaluation and reliability. As benchmarks saturate, what is trustworthy AI evaluation actually worth? The next competitive edge may be less about scoring higher and more about measuring what matters. 3) Continuous alignment and the infinite last mile. Human knowledge never stops changing, which means neither data nor alignment is ever really “finished.” Keeping models current is a continuous problem. 4) Expert happiness and model performance. How does the experience of the humans producing training data affect its quality, and ultimately model performance? Better AI may depend in part on enabling experts to do better work. These aren’t side questions. I think they sit close to the center of what will determine whether AI systems become better, safer, and more useful. We’ve spent much of the AI era thinking about intelligence through the lens of models: how we build them, scale them, and make them more capable. Increasingly, I think some of the most consequential questions sit one layer underneath, in the human systems that teach, evaluate, correct, and continuously improve them. Since the initiation of my Chief Economist role last summer, I’ve had the chance to start digging into these questions from inside a company operating directly at that layer. Now we’re scaling that work into a dedicated #economic #research #unit, giving these questions a home, a team, and the space to pursue them with the depth they deserve. There’s a lot we still don’t understand about the economics of the human infrastructure behind AI. That’s what I want to spend the next chapter working on.

What if AI research could create more culture, not less? Over two days, we sponsored more than 20 live performances across San Francisco, giving local musicians new stages and bringing more live music into public spaces around the city. Portions of those performances were also recorded to create real-world audio data for AI audio and music research. That data can help advance music education tools, improve how systems understand instruments and musical structure, and create new ways for emerging artists and music to be discovered. We believe nothing replaces human brilliance. The future of AI in music shouldn’t mean less human creativity. It should help people explore more, learn more, and bring more ideas to life.

Perception models for defense are only as strong as the conditions they've been trained and evaluated on. Our latest work in mission AI uses synthetic data generation to help teams fill coverage gaps that are difficult, costly, or impractical to collect in the real world. The goal: help defense programs build more complete datasets and evaluate models across the conditions they're actually expected to encounter. Learn more at the link in comments.