I’ve joined @AnthropicAI to build AI + education! 🌎 4 years ago, I cofounded @schoolhouse_edu with @salkhanacademy. Schoolhouse is now the largest free human-to-human tutoring platform on the Internet, serving 190 countries. I’m excited to join the board and take on a new challenge. Thrilled to work with @alexalbert__, @YoMaggieVo, @AlexTamkin, @wenquai, @sammcallister, and many others on ensuring AI benefits everyone — including by transforming the way we learn and teach!

If you’re using Claude for Teachers and have feedback, I’d love to hear from you. Curious what’s working and what can be improved. Can jump on a call too. Please reach out!

(1) A foundation in curriculum and learning science. The Learning Commons connector brings standards from all 50 states, the prerequisite skills beneath each one, and curriculum from Illustrative Math and OpenSciEd. Ask for a lesson on constructing linear functions: Claude pulls the standard, notes the misconceptions students typically hold, and builds the lesson around them.

I've been using Claude Design and while it's still early, I'm seeing some of the same acceleration as I first did with Claude Code a year ago, but now for design. Presentations, mockups, brainstorming -- its versatility is what makes it so powerful. My first thought was, can't this be done in Claude Code? But the design-first UI means you can make inline comments and direct edits just like you're in a design critique. It also relies on your design system. I'm particularly excited to see how teachers leverage this for producing polished classroom materials.

Thoughts on developing AGI-pilledness as a skill

It's time to go beyond the chat interface in AI. Claude now creates interactive charts and diagrams directly in the conversation. Artifacts was v1. This is v2. We're finding it especially powerful for learning use cases.

Wrote down some reflections after 1 year at Anthropic (feels like a lifetime): 1) For each breakout success, it started as 1-2 people's side project. True for Claude Code, Cowork, MCP, Artifacts 2) Being AGI-pilled is a skill & you can get better at it 3) We humans adapt surprisingly well. SWEs today look very different than a year ago 4) Roles are somehow becoming both more manager-like (directing agents) and IC-like (everyone’s a builder) 5) Most people I know have had their roles change at least a few times over the past year, whether in name or practice 6) Fond memories of the colleague who used to set up a 1:1 with every new hire, as well as the one who would read every slack message in every channel. Neither are possible today… for humans at least 7) I’m surprised how I went from knowing almost no one here to now having a friend/colleague join every couple of weeks 8) Strategic thinking matters a lot at the AI labs 9) It’s worse to underestimate a technology’s potential than overestimate it 10) Initiatives in a company can go from super underresourced to overresourced in a short time, which you have to watch out for 11) “Antfooding” of products internally seemed silly at first, too insular. But I now see its merits for AI labs. 12) Writing culture is big at Anthropic, although I’m not sure how long that will last 13) Internal dissent is alive and healthy, and often make up the most lauded docs/slack posts 14) Work-life balance seems to have gotten worse across the company as we progress along the exponential 15) Being an IC with nothing on your calendar is still one of the most sought after roles 16) Take the stairs whenever possible 17) The weight of the technology we’re building is becoming more difficult to grapple with

We at @AnthropicAI are partnering with @Pratham_India, one of India's largest education NGOs, as their first strategic AI lab partner. Their Anytime Testing Machine, powered by Claude, is a formative assessment tool that will give personalized feedback to young women who have dropped out of school and are working toward India's 10th-grade board exam through the Second Chance Program. Here's what I think makes this partnership work: - shared values - rigorous assessment of where AI helps (and where it doesn't) - deeply embedded technical collaboration

Opus but 2.5x faster feels qualitatively different when coding Cost is still a big issue but I’m optimistic