Ex-Tesla president Jon McNeill reveals Elon's five-step 'algorithm' for scaling hardware companies

Summary

Jon McNeill, former president of Tesla and now a venture investor, lays out the five-step operating framework he developed alongside Elon Musk — an approach he argues is as relevant to software and AI companies today as it was to scaling hardware.

The algorithm

The framework emerged from postmortems on self-inflicted mistakes at Tesla, with the goal of pushing decision-making to the edge of the organization. The five steps:

1. Question every requirement. If a requirement isn't mandated by physics, safety, or law, consider deleting it. McNeill's example: it took 64 clicks to buy a Tesla online, 44 of them inside auto loan and lease documents. When lawyers confirmed none of those paragraphs were legally required, Tesla reduced the loan disclosure to four sentences and saw a material lift in online sales of $100,000 vehicles.
2. Simplify the product. Remove everything the customer doesn't pay for.
3. Speed it up. Speed reveals flaws. If something can move fast, it has to be simple and good — McNeill's explicit argument that the "good, fast, cheap — pick two" trade-off is false.
4. Automate last. This is the counterintuitive one. DoorDash launched with a PDF and a phone number to map the workflow before writing code. Amazon's founders ran to a bookstore to fulfill early orders by hand. Tesla automated the Model 3 line prematurely and it failed; the company was saved by tearing down the automation and building cars by hand in a tent outside the factory.
5. Then automate. Once the workflow is understood and simplified, automation compounds the gains.

McNeill traces Claude Code's current lead in AI coding tools to exactly this logic — long internal debates about simplifying the architecture before shipping, which allowed it to do things Codex and Cursor can't.

Software-first hardware

Tesla's edge over traditional automakers came from hiring orthogonally — bringing in people with no auto industry experience, including McNeill himself, a six-time software entrepreneur. The concrete payoff: Tesla concluded that autonomous driving required a single chip to eliminate latency and synchronization overhead. Competitors running 18 to 36 chips, McNeill argues, are structurally unable to deliver autonomy because they're thinking hardware-first.

Keeping culture sharp post-liquidity

On whether SpaceX faces a motivation collapse as it approaches a $1.5–2 trillion IPO valuation, McNeill draws on the Tesla post-IPO experience. Most mission-oriented people stayed engaged regardless of bank account size. The structural mechanism Musk used to maintain urgency: Tesla operated on a single quarter of cash even after going public — with 70 days of payables outstanding, that left less than three weeks of actual liquidity at times. McNeill wanted more breathing room; Musk refused, arguing that proximity to death sharpens decision-making.

McNeill's own entrepreneurial itch was channeled inward. He built mobile service and Tesla Insurance inside the company — the insurance business now represents roughly a third of Tesla's cash flow.

China's industrial playbook

McNeill, who sits on GM's board, frames Chinese EV competition as structural rather than cyclical. The playbook: every five years, the government announces five industries to enter and dominate, subsidizes roughly 100 entrants, then lets competition run until three to five winners emerge. Those winners receive the consolidated manufacturing capacity of all 100 companies for free and are directed to export. The result is companies like BYD entering global markets with heavily subsidized cost structures, proven domestic competitiveness, and no-cost scale.

Chinese brands are already on American roads — Volvo is owned by Geely, and Zeekr vehicles are the new platform for Waymo following Jaguar's discontinuation of the iPace. McNeill's position is that this is not theoretical.

The manufacturing gap

The sharpest competitive risk McNeill identifies isn't tariffs or subsidies — it's engineering talent. Xiaomi already runs a lights-out factory where raw materials enter one end and finished phones exit the other with no humans on the floor, managed by an AI plant operator. Tim Cook's oft-cited observation frames the scale of the problem: all the manufacturing engineers in the US would fit in Stanford's auditorium; China's would fill three Stanford football stadiums. McNeill argues the US needs demand-pull from companies like GM and Ford to drive university and trade school investment in manufacturing engineering before the gap becomes unrecoverable.