Wait… Tesla Just Added HOW Many Robotaxis?
Overall Thesis
Tesla is making significant progress on robotaxi deployment, Optimus humanoid robot training, and custom AI chip production, with recent fleet additions and public software demonstrations indicating accelerating commercialization timelines.
Narratives
Herbert and guest Jo Bhakdi discuss Tesla's rapid robotaxi fleet expansion, noting the fleet has grown to 476 registered vehicles in Texas including 67 CyberCabs. They argue that crossing the psychological threshold of 100 vehicles in active service could act as a catalyst that reawakens investor interest in the stock.
Key Arguments
- Robotaxi fleet grew from 9 to 28 to 476 total registered vehicles in Texas in a short span
- CyberCab production could reportedly scale to 300 units per week
- Reaching 100 vehicles in service is seen as the first 'mini catalyst' for investor attention
Predictions (1)
The pair debate whether Tesla will eventually sell CyberCabs directly to consumers or keep them as centrally-owned fleets. Jo Bhakdi argues that Tesla profits more from owning and operating fleets itself, especially in dense metro areas, while private fleet ownership may make sense in rural or less profitable regions.
Key Arguments
- Elon has reiterated intent to eventually sell CyberCabs to consumers
- Tesla likely earns more by operating fleets itself in major metros
- Selling individual CyberCabs to random consumers creates network-effect problems
The hosts discuss whether Tesla will license FSD to other automakers, concluding that regulatory hurdles requiring per-vehicle approval make commoditization of self-driving software much harder than commonly assumed. Jo argues Tesla has openly offered licensing for years, but the complexity of hardware and safety certification means competitors won't quickly replicate Tesla's approach.
Key Arguments
- Tesla has openly invited automakers to license FSD for years, but none have accepted
- NHTSA approval processes apply to the full vehicle system, not just software, making licensing legally complex
- Commoditization of self-driving tech is possible but would still require years for competitors to build the necessary hardware and regulatory track record
Jo Bhakdi expects Tesla's Optimus robot program to reach a major commercial launch next year, driven by intense competitive pressure from Chinese robotics makers and Tesla's need to stay ahead. He believes manufacturing capacity could scale into the tens of thousands of units by the end of next year, though he acknowledges near-term execution will remain messy.
Key Arguments
- Chinese competitors are shipping Optimus-like components in the thousands, forcing Tesla to accelerate
- Camera-based worker training for Optimus is already underway at multiple Gigafactories including Berlin
- Manufacturing capacity could reach tens of thousands of units by end of next year
Jo Bhakdi frames Tesla's converging businesses—robotaxi economics, Optimus, and the new AI5 chip—as building toward a major cash flow inflection starting next year, arguing that the main uncertainty left is timing rather than whether these initiatives will succeed. He also highlights that CyberCabs could generate an estimated $30,000-$50,000 in annual margin per vehicle once deployed at scale.
Key Arguments
- Multiple business lines (robotaxi, Optimus, AI5 chip) are converging simultaneously, reducing overall execution risk
- Estimated $30,000–$50,000 in annual margin per CyberCab once deployed at scale
- Favorable regulatory environment under the Trump administration reduces obstacles to autonomous vehicle rollout
Predictions (1)
Hedges & Caveats
- Distinction noted between registered vehicles (476) and actively deployed vehicles in the wild (42 tracked)
- Lag acknowledged between registration and actual deployment
- Tracker limitations may undercount deployed vehicles
- Future catalysts are speculative (100 vehicle threshold mentioned as 'magical number')
- Volume production of AI5 chip targeted for 2027 (future date)