Tesla Earnings Call Clearing Up Key Points *The key message of this earnings call was

26/4/23 #Tesla News Summary

• Tesla Earnings Call Clearing Up Key Points *
  The key message of this earnings call was “investment and cash flow over short-term gains and losses.”
  Despite increased SBC, increased AI investment, and SpaceX equity investment ($2B), it has secured operating cash flow of $3.9B and FCF of $1.4B.
  Significantly expanded beyond CapEx $25B in 2026 → It clearly stated that it is a phase of sacrificing short-term profits and all-in on long-term growth.

1. The center of Tesla’s strategy = AI + robots + semiconductors

• Investing in all directions from battery, powertrain, AI, software, and chip design/manufacturing
• In the midst of a structural transition to “automobile company → physical AI infrastructure company” (Emphasis on Elon directly)
  → Chip Shortage Is Not a Cost Issue, It’s a Survival Issue

1. FSD is already in the monetization phase

• 1Q New paid user record increase (active total 1.3M, net + approximately 200,000)
• 14% of the total adoption rate (in the 30% range in North America)
• Reduce turn + increase usage time (expand long-distance utilization)
  → Tech Verification Is Over, Now How Big Is Expanding

1. Robotaxi is already in operation, and the expansion is very conservative

• Austin unsupervised 운영 → Dallas, Houston 확대
• Aim to gradually expand to 12 regions this year
• QA Vehicle + Simulate Safety Verification
  → Strategy for incremental expansion after safety verification rather than “fast expansion” (Elon: no accidents/injuries first)

1. The real problem with robo-taxi is not safety, but “natural.”

• Excessive conservative driving → Stop/stop behavior occurs
• Exceptions such as signal failure, construction section, and infinite loop are bottlenecks
• Elon in a nutshell: “convenience drops to maximum safe mode… Lack of confidence is key”
  → Technology is almost complete, leaving only the stage of driving naturally like a human

1. HW3 has clear limits → HW4 upgrade required

• Unsupervisorated FSD not possible with HW3 (Call acknowledges clear for first time)
• Computational power/memory bottleneck
• HW upgrade program operation for customers (trade-in discount + replacement of computer/camera)
  → Key variable for FSD large scale = speed of hardware transition

1. FSD Roadmap: V14 → V15 → Oversight Expanded

• V14: Structural improvement + safety improvement (safer than human already)
• V15: Year-end to early 2027 Large Architecture Update
  → Unsupervised FSD is on the verge of full commercialization after gradual expansion

1. Optimus Is Tesla’s Most Important Future Product

• Under internal testing, external customer utilization aims for 2027 summer
• Gen3 Unveil July-August, Production Starts in the Second Half of the Year (2026)
• Completely new structure + 10,000+ new parts → Very slow to produce at the beginning
  → Typical S-curve (Initial Slow → followed by explosive escalation)

1. Optimus is a “local AI-powered device”

• Works without network (on-board compute required)
• Self Intelligence + Grok Orchestration
• Data center inference demand expected to surge in mass production
  → Not a robot, but a moving AI computer

1. AI chip internalization is an essential survival strategy

• External supply alone cannot handle AI demand
• AI5 tape-out completed, AI6 and Dojo3 are being developed
  → Without its own chips, growth itself is limited (Bottleneck)

1. Building a research chip Fab (TERAFAB)

• Giga Texas Three-Year Project Starts This Year
• Thousands of wafers per month (for research and development)
• Intel 14A Process Cooperation + mask/logic/memory/packaging integration
  → It’s not for mass production, it’s for development speed innovation

1. SpaceX Cooperation + Energy Business

• Fab initial infrastructure cooperation (requires board approval)
• Megapack Production Expansion (Megapack 3 ramp), Energy Distribution 8.8 GWh Recorded
• Reacting to short-term adjustments in the solar market with release models + their own panels
  → Maintaining a Second Stable Growth Axis Outside of Automobiles

1. Regulatory risk → transition to a collaborative phase

• NHTSA aggressively responds, approves Netherlands, and plans to expand China Q3
• Apply conservative policies such as FSD restrictions in case of camera field of view problems
  → Not regulatory conflicts, but manageable areas