Public transcript

JUST RECORDED: Elon Musk Interviewed by JP Morgan's Jamie Dimon

By: The Money Investing
Length: 21 min
Language: English
Text from: the video's captions
Transcribed: 14 October 2026

Read the full transcript Open on YouTube

Notes in English.

Summary

Jamie Dimon interviews Elon Musk live at JP Morgan headquarters in front of thousands of individual investors. Musk explains why SpaceX is going public now, describing a major capital growth phase involving over 100,000 Starlink V3 satellites, AI data centers in space, and solar power generation beyond Earth. He also discusses moon and Mars colonization, Starship reusability, chip fabs in New York, and his views on patriotism, talent and leadership.

Key points

SpaceX has been positive cash flow since around 2014-2015, and its private rounds were liquidity rounds rather than fundraising rounds.
Musk says SpaceX is going public now to fund a growth phase of over 100,000 V3 satellites and AI data centers in space.
Starlink V3 is 10 to 20 times more capable than V2, with 100 times the bandwidth and half the latency.
Starship will be the first fully reusable orbital rocket, making the cost of reaching orbit essentially the cost of propellant.
Musk believes solar or 'star power' could scale to a million times Earth's economy in space while using a tiny fraction of the sun's energy.
A moon base could support a thousand terawatts per year of AI space compute using electromagnetic mass drivers.
Musk says there is not a single high-volume computer memory fab in America today, motivating new fabs in New York.
SpaceX AI satellites will allow customers to run Nvidia GPUs, Google TPUs, Amazon Trainium chips, or SpaceX's own chips and software.
Musk describes himself as more laid-back than 20 years ago and says hiring people with a good heart matters as much as IQ.

Questions it answers

00:00Welcome and introductions

How was Elon Musk welcomed at JP Morgan?

Jamie Dimon welcomed Musk on the 51st floor of JP Morgan headquarters before an audience of 3,500 top individual investors, with Musk's mother May Musk standing up to greet her son. Dimon called Musk the Edison of our time.

3,500 individual investors watched across 100 branches, part of democratizing finance.
May Musk joked about rolling her eyes when Elon said he wanted to build rockets.

02:56Why SpaceX public now

Why is SpaceX going public at this moment?

SpaceX has been positive cash flow since around 2014-2015 and self-funding, but is now embarking on a major capital growth phase with over 100,000 V3 satellites and AI data centers in space. Musk also says revenue is now much more predictable.

Private rounds were liquidity rounds, not fundraising rounds.
V3 offers 100 times current bandwidth at half the latency.

08:07Moon and Mars roadmap

What is the bridge from Earth to the moon to Mars?

Musk says a self-growing city on the moon can be built faster than on Mars, and the moon's lack of atmosphere and low gravity allow electromagnetic mass drivers to launch AI data centers into deep space, potentially supporting a thousand terawatts per year. Mars is a longer-term 'fixer upper' that could one day be warmed to have liquid oceans.

Moon vacations could one day be possible.
Musk compares SpaceX to the Union Pacific railroad.

10:28Starship breakthroughs

What is the key breakthrough of Starship?

Starship will be the first fully reusable orbital rocket, making the cost of access to orbit essentially the cost of propellant. Liquid oxygen and methane are cheaper than jet aviation fuel, so cargo to space could cost less than trans-oceanic air freight.

Falcon 9 achieved partial reusability; Starship achieves full reusability.
Musk recommends the SpaceX webcasts for learning more.

12:19Starlink V3 and V4

What is next for Starlink?

V3 satellites are 10 to 20 times more capable than V2, the size of a small bus, and only launchable on Starship, which should carry about 50 at once. Starship V3 targets 100 tons to orbit and V4 over 200 tons per mission with hourly launches.

V3 has bigger phased array antennas, more lasers, and W-band and E-band.
V3 could replace some undersea cables, a security concern after Baltic cuts.

14:59AI data centers in space

How hard is building data centers in space?

Musk says it is easier than communication satellites: it is basically solar power, radiators, basic satellite equipment, and laser links to the Starlink constellation, which reaches the ground with cloud-penetrating frequencies so weather never breaks the link.

Space is seen as the primary means of expanding AI compute.
Ground power plants face community resistance; space solar could scale a million times.

15:18Chip fabs in New York

Why build chip fabs in New York now?

Musk says chips are the limiting factor: there is not a single high-volume computer memory fab in America today, and even best-case assumptions from existing makers cannot meet anticipated AI demand. Micron's Idaho fab won't reach volume production until around 2028.

SpaceX plans fabs for logic, memory and packaging.
Micron's stock has risen sharply on memory demand.

17:26AI strategy, patriotism and leadership

How does Grok fit in, and how has Musk changed as a leader?

SpaceX AI satellites will let customers run any GPU or TPU, including Nvidia, Google TPUs, Amazon Trainium chips, or SpaceX's own chips and software. Musk calls himself super pro-America, cites the classified Star Shield division, and says he is more laid-back than 20 years ago, valuing a good heart as much as IQ in hiring.

Star Shield provides military communications for the Department of Defense.
CFO Brent Johnson has served 15 years; mission belief drives long tenure.

Notes

Why SpaceX is going public now

Musk says people have suggested it for almost 10 years; SpaceX has been positive cash flow since around 2014-2015.
Private rounds were liquidity rounds for investors and employees, with SpaceX buying back stock.
Now the company is entering a major capital growth phase needing funding: over 100,000 V3 satellites and AI data centers in space.
Revenue is now much more predictable than before.

Starlink V3 and the future of bandwidth

V3 is 10 to 20 times more capable than V2, offering 100 times the bandwidth and half the latency at about half the altitude.
V3 satellites are the size of a small bus and can only launch on Starship, which should carry about 50 at once.
Starship V3 aims for 100 tons to orbit with full reusability; V4 targets over 200 tons per mission and hourly launches.
AI and robots will demand enormous bandwidth compared to humans' few hundred bits per second.

Energy and space

Musk argues the future of energy is solar, or 'star power': the sun is 99.8% of the solar system's mass.
Space could scale to a million times Earth's economy in harnessed power.
A moon base with electromagnetic mass drivers could support a thousand terawatts per year of AI compute, versus about one terawatt per year from Earth.

Starship breakthrough

Starship will be the first fully reusable orbital rocket, reducing cost to orbit to roughly the cost of propellant (liquid oxygen and methane).
Cargo to space could cost less than trans-oceanic air freight.

Chips and AI strategy

There is no high-volume memory fab in America today; Micron's Idaho fab won't reach volume production until around 2028.
SpaceX plans fabs in New York for logic, memory and packaging.
SpaceX AI satellites will run Nvidia GPUs, Google TPUs, Amazon Trainium chips, or SpaceX's own chips and software.

Leadership and culture

Musk is more laid-back than 20 years ago and values hiring people with a good heart, not just high IQ.
Long executive tenure, like CFO Brent Johnson's 15 years, comes from belief in the mission.

Text from the video's own captions. Summary and notes written by AI from the transcript, so check anything important against the recording.