Sam Altman JUST Said The RSI Era Has Begun... As Everyone Turns On Him

Value of the Information & Strength of the Argument

The video provides substantial information, aggregating recent news and primary sources. It effectively argues that RSI is becoming tangible, evidenced by concrete metrics like agent workdays and compute costs. The inclusion of counterpoints, such as the Princeton study showing AI’s poor research skills, strengthens the argumentation by avoiding one-sidedness. The video also contextualizes Altman’s statements within broader debates about AI governance and public trust.

Scientific Rigor, Source Quality, Title Accuracy

The video demonstrates strong rigor by citing official OpenAI blog posts, reputable tech news sites, and scientific publications like Scientific American. It critically examines the Navier-Stokes proof, presenting both OpenAI’s claims and mathematicians’ objections. The title accurately reflects the content, though it uses sensational language. The video also notes when claims are unverified, such as the Navier-Stokes result and the reported funding round target, maintaining scholarly caution.

The title accurately reflects the content, which focuses on Sam Altman's RSI era declaration and the mounting backlash against him, though it is slightly sensationalist.

8/10

The video is well-sourced, citing official OpenAI blog posts, news outlets, and scientific publications. It presents a balanced view, acknowledging critical perspectives and uncertainties.

Markers derived by PSI from the transcript: the creator did not define chapters.

• 00:20Intro: Altman declares the RSI era begins, and discusses his most uncomfortable job aspect.
• 01:45Altman on Raj Shamani's podcast: discussing pacing, loss of control, and concentration of power incidents.
• 03:20OpenAI agent metrics: 3.1 workdays per human, compute costs, and the admission of not knowing how to achieve aligned full RSI safely.
• 05:15Anthropic's similar progress: Claude's speed improvements and safety tests, including a quote from an employee who hasn't coded in 5 months.
• 07:30Mathematicians question OpenAI's Navier-Stokes proof, with Scientific American reporting it solves a modified problem.
• 10:00Public backlash: 41 data center ballot measures, Gallup poll showing 74% worry, and the Molotov cocktail incident at Altman's home.
• 13:00Cyber model exploits: 948 working Linux exploits built in one run, costing $1.2M, plus other news on ChatGPT Images and Xiaomi robots.

• A Thousand Fewer Paths to Root — OpenAI's report on the cyber model that generated 948 Linux exploits.
• Research Acceleration: view inside OpenAI — OpenAI's internal data on agent usage and workday metrics.
• OpenAI research automation intern — News article covering OpenAI's automated research intern progress.
• Did OpenAI Solve the Wrong Navier-Stokes Problem? — Scientific American critique of OpenAI's Navier-Stokes proof.
• No room for debate over human control of AI says Sam Altman — Altman's comments at an energy conference in Abu Dhabi.
• Figuring Out Podcast with Raj Shamani — The interview where Altman discusses the RSI era and related topics.

• Research Acceleration: view inside OpenAI — Supports the claim that AI agents are significantly accelerating research productivity.
• A Thousand Fewer Paths to Root — Corroborates the cyber model's exploit generation capabilities.

• Did OpenAI Solve the Wrong Navier-Stokes Problem? — Raises doubts about the validity of OpenAI's Navier-Stokes proof, suggesting it addresses a modified version.