The journey is the goal - Jensen Huang

This week at the Perplexity AI Business Fellowship we had a fireside chat with Jensen Huang (CEO of NVIDIA) and Aravind Srinivas (CEO of Perplexity). Lots of great insights and one story that stuck with me.

In 1996, NVIDIA was dying. 

They'd bet everything on the wrong graphics architecture. Microsoft's DirectX had just standardised the industry on triangle rendering, while NVIDIA was stuck with quad-based technology. There were 50 great competitors who were ahead of them. They were out of money.

Jensen flew to meet Sega's CEO, Shoichiro Irimajiri, with three brutal truths:

- Our technology doesn't work

- The product we promised you is doomed

- We need you to pay us anyway, or we're dead

Irimajiri's response was extraordinary. 

He not only paid the contract but invested an additional $5 million. As Jensen said: "His understanding and generosity gave us six months to live."

Jensen's insight: "Confronting our mistake, and with humility asking for help, saved NVIDIA. These traits are the hardest for the brightest and most successful."

Read on for reflections from Jensen on why 

1. Radical candor + vulnerability = extraordinary leverage

2. Ecosystem lens > chasing feature parity

3. Honesty ≠ weakness

4. No quotas + daily/weekly retro = endless drive, zero burnout

5. Internal discipline → external lock-in.

6. 100 % bet > hedge; hedge ⇒ slow bleed.

7. Context framing + tough love = 10 × talent lift

8. LLM reasoning × digital twins → $100 T physical-AI prize

9. Presence > busy-ness.

The man who almost lost everything to a DirectX announcement now runs a $4 trillion company with the same philosophy: 

• be intellectually honest with reality 

• ask for help when needed 

• do your best work with pride and 

• you have plenty of time (Jensen doesnt wear a watch!)

1. Intellectual honesty is the core principle

Radical candor + vulnerability = extraordinary leverage

In 1996, NVIDIA was dying. They'd bet everything on the wrong graphics architecture. Microsoft's DirectX had just standardised the industry on triangle rendering, while NVIDIA was stuck with quad-based technology. 50 competitors were ahead of them. They were out of money.

Jensen flew to meet Sega's CEO, Shoichiro Irimajiri, with three brutal truths

• Our technology doesn't work

• The product we promised you is doomed

• We need you to pay us anyway, or we're dead

Irimajiri's response was extraordinary. He not only paid the contract but invested an additional $5 million. 

As Jensen said: "His understanding and generosity gave us six months to live."

2. The brutal pivot story

Ecosystem lens > chasing feature parity

The NV1 to NV3 pivot context is crucial. 

They went from forward texture mapping/curved surfaces (completely wrong) to having to buy OpenGL textbooks at Fry's to learn triangle-based rendering. 

While 50 world-class competitors were already building it. They were literally last place with the wrong technology, no knowledge of the right approach, and burning cash.

Jensen's insight: "Just because we don't know how to compete doesn't justify doing it the wrong way." They had to abandon their "clever" architecture completely and start over.

3. Strategy lives outside of the product

Honesty ≠ weakness

After NV1/2 flopped, NVIDIA was dead-last among 50 graphics rivals. 

Jensen still found white-space by re-imagining distribution + ecosystem, not just silicon.

Your unique lane often appears only after you’re willing to admit the current plan is “exactly wrong.”

This explains why they survived when others with better technology didn't. The discipline to rebuild correctly, even from last place, mattered more than being first with the wrong approach.

4. The journey is the goal philosophy

No quotas + daily/weekly retro = endless drive, zero burnout

Jensen's management philosophy emerged from nearly dying as a company:

• No annual plans or revenue goals

• No artificial deadlines

• "The journey is the goal"

• "I just wanted to be employed doing my best work"

He measures success by looking backward each day/week asking.

"Did I do important things? Did I help people?" 

Not by hitting numerical targets.

5. Build platform gravity

Internal discipline → external lock-in.

The CUDA bet mirrors the NV3 decision. 

Adding significant cost with no immediate benefit to graphics while under intense competition. 

But they learned from NV1/NV2: stick with the right architecture even when it's hard. 

The discipline of making everything CUDA-compatible came from surviving that early crisis.

CUDA’s moat is that rule. To “everything above and below must be CUDA-compatible.” 

6. All-in beats hedging

100 % bet > hedge; hedge ⇒ slow bleed.

Every inflection (triangles, CUDA, deep-learning GPUs, Omniverse, now CUDA-Q) was a bet-the-company move. 

7. The new leadership literacy = context engineering 

“I don’t fire people; I torture them to greatness.”

Context framing + tough love = 10 × talent lift

Knowing when to over-constrain or under-constrain is a critical leadership skill. Exactly like setting vision for a team. 

“I don’t fire people; I torture them to greatness.” Retaining early leaders compounds expertise across new frontiers (cars, robots, quantum).

8. Physical AI vision will be the next $100 trillion opportunity

LLM reasoning × digital twins → $100 T physical-AI prize

He sees multimodal reasoning models collapsing the entire perception-to-planning stack for robots. Warehouses first, streets later. Massive simulation (Omniverse) will be the compute sink.

Jensen sees the next $100 trillion opportunity in physical AI. 

The convergence moment: reasoning vision models that can drive by understanding language. 

This vision comes from someone who bet the company multiple times and won.

9. Life lessons from unexpected teachers

Presence > busy-ness.

Three people who shaped Jensen:

• His mother - taught him English without knowing English ("How hard could it be?")

• - Roger the Denny's dishwasher - showed up precisely on time, worked with extraordinary pride

• Japanese moss gardener - when asked why he was picking tiny dead moss pieces with a tiny pair of bamboo tweezers from the biggest moss garden in the world - "I have plenty of time". 

Jensen uses that last phrase frequently (~10 times daily). He doesn't wear a watch because he has "plenty of time."

The next time you're staring at failure, remember. 

Somewhere in 1996, the future CEO of the first $4 trillion company in the world was buying textbooks to learn what his competitors already knew and were building.