Generative AI Weekly - June 06, 2024

This week:

• Two new (to me) incredibly fun products from Google. Illuminate turns academic papers into AI-generated audio discussions (!!) and Notebook LLM takes many documents you gather into a research assistant. I will be playing with both this week for both work and pleasure!
• An automation tool I am currently demoing - Gumloop - integrates LLMs into drag-and-drop workflows!
• Delve - a non-linear LLM UI (I love where this is heading)!
• I just came across Jevons Paradox which "occurs when technological progress increases the efficiency with which a resource is used (reducing the amount necessary for any one use), but the falling cost of use induces increases in demand enough that resource use is increased, rather than reduced."
• "In short, a spot-market for hiring AIs is developing."
• Co-founder of Anthropic does a 5-year retrospective on GPT-2.
• LLMs are W.E.I.R.D.
• The value of in-person meetings continues to rise.
• Ex-OpenAI researcher sets out a vision for the next decade with AI. A very long, but very scary/engaging podcast discussion here.
• Ethan Mollick's new post on experiencing AI through play. I find demos in play territories are always the most interesting to new audiences - even professional ones!
• A survey from February has some helpful outlines of LLM capabilities (somewhat outdated already).
• And for papers this weekend (possibly via Illuminate) - Daron Acemoglu writes a NBER paper on "The Simple Macroeconomics of AI" - Goldman Sachs is more optimistic.