ExpertsFMNVIDIA CEO Jensen Huang
Acquired · October 16, 2023
5 Selected Lessons:
heuristic · 2 case · 1 reasoning_pattern · 1 priority · 1
What a business plan must actually prove
Source
Acquired — NVIDIA CEO Jensen Huang (October 16, 2023) · at 1:16:55 ▸
When this applies
when deciding what content actually matters in a founding business plan or pitch
The takeaway
Rank the true problem and the defensibility of your solution above financial projections and go-to-market polish when evaluating or writing a founding plan — the latter are learnable skills, the former is the core bet.
Why it works
Financial forecasts in an unproven business are inherently unreliable guesses, whereas identifying a real unmet need and a defensibly hard solution actually predicts survival, because competitors will swarm any easy idea once it's revealed.
Where it stops holding
This stops working when investors specifically require financial credibility for later-stage or capital-intensive businesses where forecasting discipline itself signals competence.
Detail
hierarchy:
- 1. What is the true unmet need you believe will emerge?
- 2. What are you going to do that is sufficiently hard that competitors can't swarm and copy it once it's known to be a good idea?
- 3. Product, positioning, pricing, go-to-market — learnable skills, lower priority
- 4. Financial forecasts — Jensen says these 'turn out not to be that important' since nobody knows if they'll be right
rationale: Jensen: 'what is it that you're going to do that is sufficiently hard that when everybody else finds out it's a good idea, they're not going to swarm it and make you obsolete.'
In their words
What is it that you're going to do that is sufficiently hard that when everybody else finds out it's a good idea, they're not going to swarm it and make you obsolete. And so it has to be sufficiently hard to do.
Moat-as-network thinking
Source
Acquired — NVIDIA CEO Jensen Huang (October 16, 2023) · at 47:42 ▸
When this applies
when defending a lead in any market you were early to create
The takeaway
Reframe competitive defense from 'building walls around your castle' to 'growing a network that others benefit from alongside you' — durable advantage comes from shared ecosystem value, not exclusion.
Why it works
Helping an ecosystem of developers, partners, and customers succeed around your product creates a 'network of networks' that is defensible not because you block competitors, but because switching away means abandoning the whole ecosystem's accumulated value.
Where it stops holding
This stops working when the ecosystem's value is easily replicated by a competitor or when network effects are weak in your specific market.
Detail
approach: View competitive strategy through ecosystem-enablement rather than moat/defense framing; invest in making the whole ecosystem (developers, customers, partners) succeed alongside you.
signals:
- you were early and mindful about helping developers/partners succeed
- a large ecosystem of complementary products/services formed around your platform
rationale: Jensen: 'I don't love thinking about it in the context of a moat... I tend to like thinking about things in the context of building a network.'
In their words
That network is essentially your moat. And so, you know, I don't love thinking about it in the context of a moat. And the reason for that is because you're now focused on building stuff around your castle. I tend to like thinking about things in the context of building a network.
Learn from everything, imitate nothing
Source
Acquired — NVIDIA CEO Jensen Huang (October 16, 2023) · at 31:38 ▸
When this applies
when consuming case studies, best practices, or competitor moves from any source
The takeaway
Treat outside examples as inputs to your own reasoning process, not templates to copy — always ask 'what does this mean to me in my context' before acting.
Why it works
Direct imitation ignores that your context, company age, capabilities, and environment differ, so a borrowed tactic may not fit; translating the underlying lesson into your own context preserves what's actually transferable.
Where it stops holding
This stops working when you lack the judgment or context-awareness to correctly adapt the lesson, and may default back into literal copying.
Detail
rule: Learn from every company example or book you encounter, but always translate it into your own specific context rather than directly adopting it.
threshold: Applies regardless of source — competitor, adjacent industry, or unrelated industry.
rationale: Jensen: 'the whole point of these books is to share their experiences. And you're supposed to ask...what does it mean to me in my world?'
In their words
As in everything you read, you're supposed to, first of all, enjoy it, right? Enjoy it, be inspired by it, but not to adopt it
Target zero-billion-dollar markets
Source
Acquired — NVIDIA CEO Jensen Huang (October 16, 2023) · at 43:14 ▸ · at 47:42 ▸
When this applies
when choosing which markets or customer segments to pursue as a growing company
The takeaway
Prefer serving an unmet future need that has no current market over fighting for share in an existing one — the absence of competitors at entry is itself the strategic advantage.
Why it works
Positioning in a market that doesn't yet exist (non-consumption) means there's no incumbent competition to fight, and if your bet about future unmet need is correct, you'll have accumulated a decade of expertise and relationships by the time the market materializes.
Where it stops holding
This stops working when the anticipated future need never materializes, leaving years of investment with no market to monetize — pure speculative risk.
Detail
rule: Deliberately position the company in markets that don't yet exist ('zero billion dollar markets') where you believe future demand will emerge, roughly a decade ahead of the market.
threshold: ~10 years ahead of market emergence, per Jensen's stated pattern (automotive CTO's 'cars cannot tolerate blue screen of death' objection, proven wrong ~15 years later).
rationale: By the time the market emerges, there aren't many competitors shaped the right way to serve it, because they weren't there early.
In their words
We prefer to position ourselves in a way that serves a need that usually hasn't emerged
Riva 128: betting the company on a simulated-perfect chip
Source
When this applies
when a company has only one shot left to ship a product before running out of resources
The takeaway
Before committing irreversible resources to a single shot, pull every future risk you can into the present and resolve it via simulation or dry-run first.
Why it works
Buying and using an emulator let NVIDIA run the entire future software/hardware validation cycle in simulation before committing physical production, converting an irreversible bet into a near-certain one.
One level up
When facing collapse, redesign the process itself (not just push harder) so the 'all-in' bet is actually a resolved bet in disguise.
Where it stops holding
This stops working when the risks cannot actually be simulated or tested in advance (e.g. market reception, unknowable unknowns) — simulation only de-risks what's technically testable.
Detail
situation: NVIDIA had ~6 months of cash left in 1997 and needed to tape out a graphics chip with no room for a second attempt.
decision: Bought an idle emulator company's inventory, built the full software stack against the emulated chip, and only taped out once everything tested successfully.
outcome: The chip worked essentially as planned and became 'a reset of the company,' setting the pattern NVIDIA still uses for chip design today.
lesson: Betting the company isn't reckless if you've already simulated away the risk — 'when you push your chips in, you're saying I'm going to take everything in the future, all the risky things, and pull it in advance.'
In their words
When you push your chips in, what you're really doing is when you bet the farm, you're saying, I'm going to take everything in the future, all the risky things, and I pull it in advance. And that is probably the lesson.