E90: The State of AI
中文标题
E90：AI 现状
Core conclusion · Tlines extraction
Citi's Research podcast episode 90 features Air Street Capital founder Nathan Benaich discussing the tenth edition of his State of AI report, a 244-page annual survey. The supplied excerpt covers only the opening and closing remarks; it contains no substantive findings, forecasts or asset calls. Benaich notes that the pace of AI news now forces near-daily revisions, making an annual cadence harder to sustain.
- Institution
- Citi
- Published
- 10/22/2026
- Research topic
- Semiconductors · NVIDIA
- Time horizon
- Not explicitly stated
Key arguments
- An annual AI report cadence is increasingly strained because news and research now shift daily.
- The State of AI report has become a foundational reference for the AI community over nearly a decade.
- The 2026 edition spans 244 pages and reflects rapid change over the past year.
Key numbers
- Report length
- 244 pages
- Report edition
- ~10th annual (nearly a decade)
- Podcast episode
- E90
- Recorded
- October 12, 2026
Main risks
- Supplied excerpt lacks substantive findings, so no investment conclusions can be drawn.
Corresponding source evidence
And as you’ll see in the Research slides, most of the benchmarks that were really ambitious 12 months ago when we published the report have now saturated, which means that the model has reached close enough to 100% that any material changes in performance cannot be measured.
Nathan Benaich (05:14) It means that you quasi give up on building frontier models yourself and especially by doing pre-training yourself and you pick up open weight models and provide most likely a fine-tuning service to customers by leveraging all the compute capacity that you built up with the original intention of using it to pre-train.
Heath Terry (10:07) Yeah, yeah, it's going to be fascinating to sort of see how that plays out.
Context: the conclusion is Tlines' automated extraction, not the institution's wording. Quoted evidence is publisher wording. Scope and date must travel with the conclusion.
Verify at the original source ↗Loading the document…
AI analysis
Cross-institution comparison · SOX
How this page was produced
Tlines extracted the conclusion, arguments, numbers and risks above from the report published by Citi on 10/22/2026, then structured them for comparison. The publisher's own wording is reachable through the source link on this page and is displayed in full above. Last updated 10/23/2026. Methodology · Source policy · AI usage · Request a correction