How he thinks.

Rather than reasoning by analogy — 'this is how it's always been done' — first principles thinking requires you to break a problem down to its most fundamental physical or logical truths and reason back up from there. Musk credits this as the single most important tool in his arsenal and the namesake of his entire intellectual brand.

The ratio of the cost of a finished component to the cost of the raw materials required to make it. A high idiot index means someone — a supplier, a contractor, or your own team — is being paid a lot of money for something that shouldn't cost much to make. Musk uses it as a rapid filter for whether SpaceX or Tesla should manufacture something in-house instead of buying it.

The most reliable part is the one that doesn't exist. Every component you eliminate removes a potential failure mode, a weight penalty, a manufacturing cost, and a maintenance burden. This principle drives SpaceX and Tesla's extreme integration approach — designing out complexity rather than engineering around it.

Step one of Musk's engineering process is to challenge the requirement itself before trying to meet it. He argues that requirements often carry hidden assumptions, outdated constraints, or someone's preference disguised as a physical law. A requirement is only valid if you can name the specific, smart person who owns it — and have them defend it.

Musk's manufacturing and design process: (1) Make the requirement less dumb. (2) Delete the part or process. (3) Simplify or optimize — but only after steps 1 and 2, never before. (4) Accelerate cycle time. (5) Automate. The critical insight is the ordering: most engineers go straight to step 3 or 5, optimizing or automating something that should have been deleted.

Rather than engineering for perfection before testing, Musk prefers rapid iteration: build, test, fail, learn, rebuild. SpaceX's development model for Starship — flying prototypes at increasing complexity until they explode, then incorporating lessons — is a direct expression of this. Musk views an untested design as a hypothesis, not an asset.

Musk's companies build an unusually high percentage of their own components rather than outsourcing to suppliers. This gives them control over cost, quality, and the pace of iteration. It also means that when Musk applies the idiot index, he has the option to internalize production rather than simply complaining about a supplier's price.

Musk believes that concentrating exceptional engineering talent at a single company, paying them well, and giving them hard problems creates a compounding advantage that is nearly impossible for competitors to replicate quickly. He has described his goal as hiring the 100 smartest people in every discipline he needs.

Musk believes that in industries where hardware iteration is possible, moving fast enough to outpace competitors is more valuable than moving carefully enough to avoid failures. He is unusually tolerant of public, dramatic failures — Starship exploding on the pad — as long as the team learns from them quickly.

Musk expects — and models — working at extreme intensity, often 80-100 hours per week, particularly during critical moments. He has described sleeping on the factory floor during production crises. His 'hardcore' email to Twitter/X staff in November 2022 formalized this as a company requirement: staff had to sign up for 'extremely hardcore' work or be let go.

Musk regularly uses order-of-magnitude estimates to evaluate whether a goal is physically possible before investing engineering resources. Named for physicist Enrico Fermi, who could estimate complex quantities (the number of piano tuners in Chicago) from first principles to within a factor of two. Musk uses this to filter out ideas that are obviously impossible or obviously feasible.

Musk insists that all engineering decisions ultimately run up against physical laws, and that understanding those laws — thermodynamics, materials science, orbital mechanics — is more valuable than domain expertise or industry convention. He is skeptical of 'experts' who cite precedent rather than physics.

If something can be improved, it should be improved — and eventually will be, by you or by someone else. Musk treats continuous improvement as obligatory rather than optional. This is the engineering complement to his speed-over-caution philosophy: failing to improve is itself a form of failure.

Musk often references the principle that the last 10% of any engineering challenge consumes 90% of the time and cost. His implication: don't let perfection be the enemy of deployment. Ship the 90% solution, learn from real-world performance, and close the last 10% through iteration rather than design.

Perhaps the most famous Musk one-liner — simultaneously an expression of his genuine desire to colonize Mars, his dark humor, and his philosophical acceptance of mortality in service of a larger goal. It encapsulates his belief that the multi-planetary imperative is worth personal sacrifice, and that physical risk is not a disqualifier for a mission of civilizational importance.