Zvi Mowshowitz

“I'd bet big against a Chinese lab actually having the best model at any point in 2026, but benchmarks are not leaderboards. I'd be very surprised if this happened on Artificial Analysis... I'd be less surprised if this happened on LM Arena, since it is rather hackable... I still think this is an underdog. [On the broader 'a Chinese lab tops a global leaderboard' wording: 'I'll give them maybe 55%... Manifold says 34%, and on reflection yeah I was being a coward and moderating my instincts too much, that's more like it.']”

Zvi MowshowitzThought Leader

Author of the 'Don't Worry About the Vase' newsletter; AI writer and forecaster

Written in response to the State of AI Report 2025's prediction that 'A Chinese lab tops a global leaderboard'; resolvable at the end of 2026 against Artificial Analysis and other leaderboards.

“On the prediction that a real-time generative video game becomes 2026's most-watched title on Twitch: 'I'll go ahead and take the no here. Too soon. Generative games are not as interesting as people think, and they're doubling down on the 2024 mistake... GPT-5-Pro has this at 14%, Sonnet says 3%. I think Sonnet is a bit overconfident, let's say 5%... Not gonna happen.'”

Zvi MowshowitzThought Leader

Author of the 'Don't Worry About the Vase' newsletter; AI writer and forecaster

A rebuttal of the State of AI Report 2025's 2026 prediction; resolvable from Twitch's most-watched-title data at the end of 2026.

“On 'Datacenter NIMBYism takes the US by storm and sways certain midterm/gubernatorial elections in 2026': 'If we're talking at least two races for governor, house or senate, I think this is not that likely to happen, nor is it likely to be very high on the list of issues in general. I'm on no... I'd probably say more like 15%... I'll adjust to 25% on that.'”

Zvi MowshowitzThought Leader

Author of the 'Don't Worry About the Vase' newsletter; AI writer and forecaster

Resolves after the November 2026 US midterms; data-center siting already surfaced as an issue in the November 2025 Virginia, New Jersey and Georgia races, but Zvi's own bar is at least two governor/House/Senate races being swayed.

“On 'A major retailer reports >5% of online sales from agentic checkout as AI agent advertising spend hits $5B' in 2026: 'I think this is ambitious, but variance here is really high and the correlation between the two numbers is large... Maybe 25%? Manifold says 23% so that seems good.'”

Zvi MowshowitzThought Leader

Author of the 'Don't Worry About the Vase' newsletter; AI writer and forecaster

Resolves at the end of 2026 on whether a major retailer reports over 5% of online sales via agentic checkout while AI agent ad spend reaches $5B; Zvi puts it at roughly 25%.

“I do not mean to imply that my median timeline to High Weirdness or everyone potentially dying remains unchanged at 2029... while I did find 2025 advances very impressive and impactful, I do think in terms of timelines events last year should on net move us modestly farther back on full High Weirdness expectations to something like 2030, still with high error bars, but that number is loosely held.”

Zvi MowshowitzThought Leader

Author of the 'Don't Worry About the Vase' newsletter; AI writer and forecaster

Mowshowitz lengthened his median for 'full High Weirdness' from 2029 to roughly 2030 after reviewing 2025's advances, explicitly holding the number loosely; the target year has not arrived.

“My current median time-to-crazy in this sense is roughly 2031, but with very wide uncertainty and error bars... and I don't feel I would 'have any right to complain' if the outcome was very far off from this in either direction.”

Zvi MowshowitzThought Leader

Author of the 'Don't Worry About the Vase' newsletter; AI writer and forecaster

Consistent with his September 9, 2025 post, where he wrote that 'if you put a gun to my head for a typical AGI definition I'd pick 2031,' loosely held; target year has not arrived.

“Dario Amodei says AI will be writing 90% of the code in 6 months and almost all the code in 12 months. I am with Arthur B here, I expect a lot of progress and change very soon but I would still take the other side of that bet.”

Zvi MowshowitzThought Leader

Author of the 'Don't Worry About the Vase' newsletter; AI writer and forecaster

Neither leg of Amodei's forecast came true: by September 2025 AI-written code was, in Zvi's own words, 'a lot less than 90%', and by March 2026 AI was nowhere near writing essentially all code. Zvi reviewed his call in September 2025: 'On reflection yes I was correctly betting no, but that was an easy call.'