The speaker argues that a Tesla‑SpaceX merger will not happen; instead the two companies will create a joint venture—dubbed “Terraab” (or “Terapab”)—to build a semiconductor fab that supplies AI chips to Tesla and SpaceX at far lower cost than buying from Nvidia, AMD, etc. The venture would be funded by roughly $15 billion each from Tesla and SpaceX plus about $20 billion from outside investors (potentially private equity, venture capital, or strategic partners like Intel), giving Tesla and SpaceX a combined 60 % stake. By cutting out design‑house and fab margins, the chip cost could drop from ~$50 k to a few thousand dollars, saving both companies billions and enabling cheaper Optimus robots, Cybercabs, AI satellites, and terrestrial data centers. The speaker also notes that Tesla is hiring heavily for Robo Taxi fleet support and Optimus roles, signaling imminent rollout, and briefly mentions recent AI‑stock movements (ASML, Broadcom, TSMC up; Nvidia down; Marll Networks up 32 %). Overall, the joint‑venture model is presented as a way to raise needed capital, align incentives, and win the AI‑compute race without a merger.
1. A merger between Tesla and SpaceX is not going to happen.
2. People are saying a merger is inevitable between Tesla and SpaceX.
3. Larry Goldberg stated that a true merger of equals between Tesla and SpaceX would occur.
4. Bloomberg reported that the merger is inevitable.
5. Tesla and SpaceX are planning to build a joint venture called Terraab.
6. The first phase of Terraab is expected to cost over $50 billion.
7. SpaceX plans to raise $75 billion in an IPO, with $50 billion earmarked for a new data center.
8. Tesla holds $44 billion in cash.
9. Tesla is allocating funds to Optimus and Robo Taxi.
10. Tesla and SpaceX are heavily capex‑intensive for this year and next year.
11. Neither Tesla nor SpaceX has enough free cash to contribute $25–$30 billion to Terraab.
12. A joint‑venture structure can raise the needed capital from outside investors.
13. It has been stated multiple times that the arrangement is a joint venture, not a merger.
14. Bloomberg reporting has mentioned staff working for the joint venture between Tesla and SpaceX on several occasions.
15. A joint venture involves forming a separate company, not merely a partnership.
16. In one funding example, Tesla contributes $15 billion, SpaceX contributes $15 billion, and outside investors contribute $20 billion, totaling $50 billion.
17. In that example, SpaceX owns 30 %, Tesla owns 30 %, and outside investors own 40 % of the joint venture.
18. Another theory adds Elon’s $5 billion personal contribution, bringing the total to $55 billion, matching a Grimes County filing for phase‑one cost.
19. For phase two, another $55 billion is needed, with each company contributing $15 billion and outside investors $20 billion, or self‑funded if cash permits.
20. The primary purpose of Terraab is to secure chip supply for Tesla and SpaceX, giving them priority access.
21. Tesla and SpaceX receive first access to chips produced; outside buyers wait in line.
22. If no chips remain after Tesla and SpaceX take their share, none are available for others.
23. Currently, there is limited external demand for these chips.
24. Tesla and SpaceX together will own 60 % of the joint venture; thus, 60 % of its profits go to them.
25. When Tesla or SpaceX purchase chips from the joint venture at a discounted price (e.g., $700 vs $1,000 market), the net profit goes to outside investors.
26. If the joint venture sells chips to third parties at market price, profit is shared equally among all parties.
27. The structure prevents the joint venture from having an incentive to prioritize third‑party sales over Tesla and SpaceX.
28. One major advantage of Terraab is the reduction in chip cost for Tesla and SpaceX.
29. Market prices for chips such as Nvidia H100/H2/GB300 range from $30,000 to $50,000; Vera Rubin is around $50,000 per chip.
30. Chip design firms (e.g., Nvidia, AMD, Broadcom) typically earn a 70‑80 % margin on chips they design but do not manufacture.
31. Fabricators (TSMC, Samsung) earn roughly a 50‑60 % margin on chip manufacturing.
32. By eliminating design‑firm and fab margins, chip cost could drop from ~ $50,000 to roughly $1,000‑$2,000.
33. Lower chip cost enables cheaper products such as the Cyber Cab and Optimus robot.
34. An Optimus chip costing $2,500 instead of $50,000 makes a $25,000 sales price feasible.
35. A SpaceX AI satellite with 100 chips would cost $5 million at $50,000 per chip, but $500,000 at $5,000 per chip.
36. The same cost savings apply to terrestrial data centers that use hundreds of thousands of chips.
37. Tesla and SpaceX can achieve the lowest‑cost compute by avoiding design‑firm and fab margins.
38. Job postings for Robo Taxi fleet support exist in Palo Alto (already operating in the Bay Area), Tampa, San Antonio, Orlando, and Miami (not yet operating).
39. There are over 30 Robo Taxi‑related job listings.
40. Specific Robo Taxi roles include real estate manager, architecture and design manager, supervisor, and fleet support operations manager.
41. Tesla is scaling up hiring for Robo Taxi.
42. A high‑intensity action‑oriented leader role is advertised for a Robo Taxi fleet team in Orlando, Florida.
43. There are 197 Optimus job postings.
44. Meaningful financial numbers from Robo Taxi are not expected until the Q4 2026 earnings call (January 2027).
45. The AI stocks primary index rose almost 1%; ASML up ~5 %, Broadcom up ~5 %; TSMC and AMD had a good day; Nvidia was an exception; Palunteer down 5 %; Tesla up slightly.
46. Marll Networks stock rose 32 % in one day after Jensen Huang stated at Computex that the company, valued at ~ $200 billion, aims to become a trillion‑dollar company.
47. Jensen Huang’s comments can move stock prices.
48. Microsoft stock was down slightly.
49. A recommended podcast episode features Ethan He (or Ethan Haye) from XAI discussing XAI’s work on Grock, video generation, word models, and video agents.