Announcements

• Introducing Speculative Technologies, a private DARPA-like research organization. See also coverage in Forbes and Ben Reinhardt’s AMA on the Progress Forum
• I’ll be speaking on how to write about progress at the Thesis Festival this weekend
• Day One policy memo on enabling faster NIH funding timelines (via @LNuzhna)

Opportunities

• Convergent Research is hiring a Director of Development (via @AGamick)
• Dwarkesh is looking for help with his podcast

Links

• A case for more techno-optimistic storytelling
• Patrick Collison interviewed by Reid Hoffman
• David Deutsch interviewed by Naval Ravikant
• Marc Andreessen interviewed by Dwarkesh Patel
• OpenAI will let you “define your AI’s values”
• “Most of the rank and file at the NRC are not anti-nuclear”

Queries

• Is there a historical price series for moving a cubic meter of dirt?

Quotes

• Why was the wind never used on roads? Why no carriages or wagons with sails?
• When Carnegie hired a staff chemist: “great secrets did the doctor open up to us”
• Why corporate R&D in the 1980s was mediocre compared to Bell Labs and GE
• “Even the aspiration” to sustainability is dangerous, says David Deutsch
• Francis Bacon with the understatement of the millennium
• “It can be done”

Tweets & retweets

• The Martian as one of the only tech positive films out there (@brian_armstrong)
• Katalin Karikó has received enough awards to fill a cabinet
• Why is it so expensive to build transit in the US? Summary findings
• LLMs are just a massively scaled up version of the “travesty generator”
• Toolformer lets language models use tools like web search, calculators, and APIs. Original paper on arXiv
• “Within a decade solar will be cheap enough that CO2 will be the best place to get carbon.” And how that will change the economy
• Quantifying healthspan in dogs for longevity research (@celinehalioua)
• Spy balloons have a long history

Charts

• CGO poll: people think social media is important for policy discussions but avoid being political online