The public deserves the backstory
October 24, 2026 · OSASI Commentary
Insiders keep sounding alarms about AI safety, and the people building AI expect far more from it than the public does. The public deserves to know how AI is being used, and the backstory of how it is made. Appraisers have a backstory that deserves greater discovery.
How a brief is made
A member reads an article or listens to a podcast and sends the snippets, key thoughts or concerns tied to a passage or a timestamp. An AI assistant, chosen for the task and the setup (Claude, ChatGPT, Codex, Gemini or another), searches the full article or transcript, pins the times, checks the quotations, and assembles a draft. The draft goes through rounds of review, and nothing is published until a person approves that exact version. The cues on each post start the player at the passage discussed. Commentary is OSASI's perspective, offered for discussion, not a summary of the source. Each brief is co-authored by a person and an AI assistant. AI assistant for this brief: Claude (Anthropic).
Source podcast · The Artificial Intelligence Show
This episode prompted the commentary below. It is one perspective, offered for discussion.
The Artificial Intelligence Show · October 20, 2026
#247: OpenAI’s Mounting Safety Concerns, Anthropic’s AI Expectations Gap, MAICON 2026 Recap, AI Score & Gemini 4 Argon
Listening highlights · Paul Roetzer and Mike Kaput. Times from the SmarterX transcript, checked against Spotify screenshots (within a few seconds).
Tap a time above to start the player there. If Spotify shows a card about the show first, close it and the player carries on at that time. Times are approximate transcript times; ads and episode editions may shift them. This plays the full episode, not an extracted clip.
Open episode in Spotify · Show notes and transcript · Anthropic: The AI Expectations Gap · Anthropic: How employment is changing at Anthropic · The Ezra Klein Show: 'This Is Nuts.' An OpenAI Insider Explains Why He Quit
Alarms from the inside
David Robinson led the writing of OpenAI's model safety reports until he left. His view, from his interview with Ezra Klein as read on The Artificial Intelligence Show: "I don't think that we, our peers, really anyone in the industry are being safe enough." The same week, OpenAI fired three safety researchers, who warned that their dismissals discourage colleagues from speaking openly.
The gap between the labs and the public
Anthropic compared its own staff with nearly 11,000 US adults. They agree on what AI can do today but split on what comes next: 91% of Anthropic staff expect AI capable of a Nobel Prize-level discovery by 2030, against 39% of the public. We found no comparable survey from the other frontier labs. Paul Roetzer's view of the work in front of most people: "the average worker doesn't need recursive self-improvement."
One Human Backstory
At MAICON 2026, journalist Karen Hao, author of Empire of AI, spoke about "the extractive behavior of leading AI labs." Roetzer warned the audience that some of it "might make you uncomfortable." His 14-year-old daughter was about to walk in, so he texted her that introduction first: the backstory before the hard part.
What this means for appraisers
Voltaire popularized the saying "Le mieux est l'ennemi du bien" ("The better is the enemy of the good"); in plainer language, better is the enemy of good enough. That can be framed in many ways, and here it applies two ways. On one hand, today's AI models are more than good enough for 99% of productivity and information technology work; on the other, the capability still to come matters most on the cybersecurity side. In the classical sense, the saying warns that perfection impedes progress. Applied to the rapid push toward recursive self-improvement and possible AGI, it warns that chasing better can cost us the good we already have.
In our own work, that cuts the other way: good enough has been the enemy of good. That is our backstory: some practice had holes, and those holes skewed public perception leaving cracks to be leveraged against our foundations. Admitting that makes room for better analysis of harder problems, and for appraisers with a little more intelligence in their own hip pocket.