NVDABearunverifiable
Jo Bhakdi“I would be a little cautious on the semiconductors when this whole thing blows up.”
Thesis at the time
BearishJoe expects Nvidia to be unfairly but heavily punished by the market once fears about a halt to frontier AI training and terrestrial compute buildout spread, likely around 2028, even though he believes fundamentally inference demand will keep growing. He sees this as a market overreaction rather than a reflection of Nvidia's actual business prospects.
Key arguments
- He expects the market to overreact against Nvidia once AI-doom/regulation narrative becomes mainstream.
- He believes a halt to training does not mean a halt to inference, which should keep growing, but the market won't fully price that in.
- He expects semiconductors broadly to get hit badly when this narrative 'blows up,' though he thinks that could take about another year.
Counter-arguments acknowledged
- Inference demand will keep expanding even if training buildout halts, which is fundamentally not bad for Nvidia.
Hedges and caveats (from the video)
- Host acknowledges this is macro-level discussion, not tactical trading advice
- Host notes audience may find existential AI risk discussion less exciting than immediate trading opportunities
- Concerns presented are based on whistleblower claims, not definitive evidence
- Host frames this as a conversation to explore implications rather than making definitive claims
The call
- Date said
- Sep 10, 2026
- Timeframe
- within about a year
- Price at prediction
- $218.36
- Confidence
- low
- Specificity
- vague
How it resolved
- Status
- unverifiable
Why this resolved this way(resolution audit)
Full rules: docs/resolution-spec.md.
Confidence Reasoning
Hedged with 'a little cautious' and conditional on a vague future trigger ('when this whole thing blows up'), reducing conviction.
Source
How DOOMERS Will CHANGE SpaceX and AGI
Said on Sep 10, 2026Open on YouTube ↗