weekend ai reads for 2023-04-21

🏗️ FOUNDATIONS

• via Rahim: Large, creative AI models will transform lives and labour markets The Economist

• and How to worry wisely about artificial intelligence The Economist

• Well, that was quick. OpenAI’s CEO Says the Age of Giant AI Models Is Already Over Wired

• Inside the secret list of websites that make AI like ChatGPT sound smart Washington Post

• I Heard Talking was Dangerous, a webinar on May 3, 9:15 am - 10:00 am PT

🎓 EDUCATION and AI

• A high school senior's point of view: Banning ChatGPT will do more harm than good MIT Technology Review

• 8 in 10 teachers approve of student use of ChatGPT, nearly all use it themselves, survey results. Intelligent.com

• (press release) New AI Enhancement to McGraw Hill's ALEKS Math and Chemistry Program Leads to Notable Increase in Student Learning

• Democratizing the future of education Ethan Mollick, Substack

📊 DATA & TECHNOLOGY

• An interview with Katy Ilonka Gero, computational poet and human-AI interaction postdoc on LLMs as creatives. "On tradeoffs in language models, weirdness of decoding methods, and writing as a communicative act" Embeddings, Substack

• Reflecting on Reflexion Nanothoughts, Substack

  • The references look pretty good too (we haven't read them all).

• 'Reform' AI Alignment with Scott Aaronson; long conversation with Scott Aaronson, UT professor who studies quantum computing. The AI X-risk Research Podcast

• We thought we linked to this before: AI Incident Database

• 🎉 FUN and/or PRACTICAL THINGS

• Patients were told their voices could disappear. They turned to AI to save them. Washington Post

• Offal, an AI-generated podcast. (contains swearing)

  • Surreal

• Dr. Gupta.

  • DO NOT give this any PII.

  • +1pt for the name. -5,000pts for being run by Martin Shkreli.

  • But a good indicator how terrible GPT-enabled "personalization" is going to be before it's not.

• Multi-on, a GPT-powered web browser

  • Just a demo video right now and a waiting list