AI Summit 2: It's Gotten Worse

Gemini 2.5 Pro Experimental, released on March 25th, 2025, crossed an important and concerning threshold on one of The Silver Age's internal LLM benchmarks. Shortly thereafter, OpenAI's o3 and o4-mini models made significant further advancement on our benchmark, placing well above the threshold. We will meet to discuss the benchmark, the threshold crossed, its significance, what may be coming next, and what (if any) action we should take to mitigate associated risks. DM us on Instagram or text 484-800-2275 for the address.

Recommended reading:

Tim Urban's introduction to the subject from 2015
Early discussion of the benchmark
Announcement of the first meeting
AI 2027