Podcast
NVIDIA Predicts $1TRN in Revenue: Everything You Need to Know From GTC & Anduril Lands $20B Contract
- NVIDIA capital expenditure is projected to sustain unprecedented levels for four to five years, with cumulative bookings potentially reaching $10 trillion over a five-to-seven-year horizon, though there is a 30% probability these spending trends may not materialize.
- Revenue growth is forecast at approximately 60% for the current year before attenuating to a 20–30% range within a couple of years, while inference usage is expected to increase by three orders of magnitude over the next five years.
- The cost per token is projected to decrease by a factor of 6x concurrent with a 3,000x increase in token volume, potentially driving cumulative revenue to $10 trillion within five years starting from a $1 trillion base.
- Data center demand may expand into space despite initial skepticism, and "everyone" is expected to require gigawatt centers regardless of the underlying model type.
- Current product offerings from major AI companies like OpenAI and Anthropic are expected to become obsolete within 12 months, prompting a "deck cleaning" where companies re-engineer teams over the next 36 months.
- The labor market faces significant disruption as brute force human labor becomes unnecessary, with job reductions estimated at 10%, 40%, or 15% depending on the efficiency gains, and layoffs expected to cascade through the Atlassian and Meta sectors.
- The "prompt engineer" and "go-to-market engineer" roles are predicted to become obsolete within 12 months as tools simplify, while a massive unmet demand is forecast for "forward deployed engineers" and "agentic deployment experts."
- AI fluency is identified as the primary hiring criterion, with single-digit percentages of interviewees currently meeting the standard and approximately 30% of management teams failing to meet it.
- Investors unwilling to learn new AI tools are expected to retire within three to six months, while venture capital funds in mid-tier markets risk terrible returns if they over-swing for the fences or apply power-law theses to markets incapable of supporting $100 billion outcomes.
- Growth stage funds are viewed as the winning strategy due to their wait for product-market fit proof points, with a "wall of money" anticipated to return and drive valuations for late-stage companies similar to 2021.
- Adobe faces significant disruption risk over the next five years, with a prediction that CEO Shantanu Narayen will step down in 2026, potentially causing a stock decline, while Intuit is expected to demonstrate greater durability.
- Autonomous driving for mining and industrial equipment is considered a current viable category, whereas highway autonomy is viewed as on the edge of viability, and a new company by Travis Kalanick, Atoms, is predicted to fail within eight to ten years.
- NVIDIA's stock valuation currently implies a price-to-earnings multiple consistent with a mid-300s revenue forecast for next year, suggesting the trillion-dollar demand figure was already priced in.
- Technical barriers to AI adoption are expected to vanish by Q2 2026, with LLMs performing code reviews and AI agents replacing humans in coding generation, allowing non-engineers to build software via "vibe coding."
- Major risks for NVIDIA include potential competition from Google and Amazon TPU customers and the possibility that inference consumption rises faster than cost reductions might suggest.
- The speaker asserts that any management team failing to introduce a commercial AI tool to their organization within the last 30 days is at a severe disadvantage, and companies not undergoing a talent reshuffle for AI fluency risk being deemed unimportant.
- The market for seed funds sized $50 to $100 million is forecast to underperform other vintages, while the small and medium business software sector remains difficult for AI competitors to dominate.