Ideas From Jensen Huang (Nvidia)

/ With NVIDIA’s recent parabolic increase in value to a market capitalization of $3.4 trillion (in other words, near equivalent to the GDP of India in 2022!), CEO Jensen Huang has been on the press circuit talking the company’s history, GPUs, AI, accelerated computing, and much more.

Recently I found most interesting some of these fireside chats/interviews where he talks about leadership, company culture, and his view on “work” in rather unique ways. Takeways in bold.

Background (in three sentences): Jensen Huang is founder/CEO of technology company Nvidia, founded in ’1993, based in the Bay Area, with general idea to develop graphics cards (graphics processing units, or GPUs) for computer graphics. NVIDIA has ~30,000 employees, and made about $60B in revenue in 2023 and $44bn in profit (after primary costs). NVIDIA has been a more popular name in the press given large appreciation in the company’s value given how their products support development of the ecosystem in Artificial Intelligence (AI).

Note:

Even if you say “NVIDIA is benefiting from AI hype cycle, how much of that value is actually material” the company was already scaled to $150bn market cap pre-COVID doing $10bn in revenue at ~13k employees (~770k revenue / employee). It was already (by most definitions) a successful company at less than 1/10th the valuation it is now! Feel it is safe to assume here that there is something special in the sauce that has enabled the company to acheive said valuation. The question of whether it is ACTUALLY worth that or what it will be worth from here is a completely separate discussion.
Before datapoints and takeaways, I also want to acknowledge that valuation or total assets of a company isn’t a great marker for “success” vs. evaluating what an entity (one person or team) has created/built/done/influenced/supported…or how many hard problems one person or a team has solved…or how much positive material value they’ve created…or how much benefit/quality-of-life increase they’ve brought to others.

Jensen Huang Interview at Sana AI Summit in Stockholm by founder/CEO Joel Hellermark
This is the best ~20 minute clip I found on Jensen talking about corporate culture and organization, filmed 06/2023.
I elect for no summary because it is valuable to watch the whole thing in its entirety (1.25-1.5x doable).

But my favorite points (that really only make sense after you watch it):

He’s creating a company that has goal of solving near-unsolvable problems, so hiring great/right talent is important (a step further, I take it to be you need people with more of an investor/creator/risk-taker mindset, one who is more comfortable than average with iteration and failure)

A company as small as possible, and as large as necessary. You need to empower people with knowledge and access to the company and information.

You want the most informed, most experienced, or best informed people helping you find ground truth.

There is no plan, just what we are doing.

Everything that your company goes after should be first reasoned through with first principles. Establish foundation of assumptions.

The mission of leaders is to create the environment to empower others.

Podcast with Nicolai Tangen, CEO of Norges Bank
Norges = Norway’s government pension fund.
Video links to start, section ends at 45:20 but the remainder of the video is equally interesting.

I think it is also funny that Nicolai can get basically any CEO he wants on his podcast because it is Norges Bank, like on the interview of Satya Nadella (Microsoft CEO) the description casually says “oh yeah we own 1.3% of the company or equivalent $35bn USD”

It just seems funny that, sure you’d expect the 30+ year CEO of a vastly accretive technology company to say they works really hard or works every day, yet Jensen says he’s relaxing all the time! Such a mind shift.

Yes, of course his definition of relaxing may be unique, from being wired differently or having different priorities but it seems so non-common.
I have the assumption that basically most working individuals work hard. Like most people work hard. They spend intentional time and effort in their day-to-day activities. This just seems to be on another level.
- Another connection to the most goated career video I’ve come across here where Jack Altman (Sam’s brother) interviews Sam. This was way before all of his OpenAI hype and before mainstream got to know him. His first sentence is probably the most valuable part of the whole interview and best framework to approach things in, in my view. I’ve talked about it on my old blog here (scroll down to ‘One Of My Top 5 YouTube Videos of All Time - 1/6/23’).
This is also press conference Jensen but he seems much more genuine than average so I keep both of those in mind.

Equally interesting quote: “Your perspective about the future has to be on a fairly long arc…pretty important! And it has to be somewhat directionally right”

This seems obvious but it just keeps pushing us to think longer-term.
A friend asks me for advice because they know I have their best interest at heart: I might suggest something that today doesn’t sound beneficial or is a tough decision but in 3-5 years they’’ll be 2x better off. Frankly I might not even care about their happiness today because enduring happiness is so much more valuable.
This goes back to an idea I wanted to tweet about on “in general, when there is an edge/advantage, the more it becomes readily known, the faster it is competed away” or put in other words “edge is arbitraged away once widely known”. An opportunity is exploited until it no longer exists, leading to a state where no significant advantage remains.

3. Jensen Huang Interview with Acquired, hosted by Ben Gilbert and David Rosenthal.
This is a nice podcast and was one of the first (I believe) longer-form interviews Jensen has done recently. The background of company technology is interesting to learn, but I find more interesting his thoughts on company building (27:13-34:19~), rapid-fire Q&A (1:08:58-1:13:32…and in fact through 1:19:45), and then part of this interview which got more attention, on why not to start a company (1:19:46-end).

Enjoyable listening to these tidbits; I’ll highlight some more commentary on company building from ~27:00. Paraphrased.
Where the “organization should be the architecture of the machinery of building the product”, where having a true flat structure, where having information doesn’t you an advantage / you bias more people knowing information than less, where people work on what they can provide the most specific value, and (important part) you earn what you do by helping others reason through problems and helping them succeed vs. priviliged information.

Jensen Huang Fireside Chat at Stanford Graduate School of Business (GSB)
This is a nice interview, intro through 7:00 or even 15:00. Clips on challenges / team (25:00-30:00), and the clip that went viral from this interview is on “No task is beneath me” (30:00-36:40).

“You can’t show me a task that is beneath me…If I can be of service to you in my review of it, show you how I reason through it, I’ve made a contribution to you. How someone reasons through something empowers you. This is how you reason through something super ambiguous / incalculable / seems to be very scary. So I show people how to reason through things all the time. Strategy, forecasting, breaking a problem down. You’re empowering people all over the place. In that process of doing that, I’ve learned a lot from you…feel rewarded by the process.”

“…What marks a leader’s place in a company is ability to reason through complicated things, help other people acheive, inspire them, empower/support them, in service of all of the other people who work there…they create conditions where other people who volunteer to come work for you and do great work.”

I won’t quote the rest but HIGHLY recommend listening for 30:00-36:40. Really interesting part at 32:26 on connecting to incredibly smart people, and you need to get on the same level as them to understand much less help them think.
- This reminds me of podcast between Joe Lonsdale and Alex Karp here (15:40-20:33) where Karp talks about the same idea. Mental model: You need to be within 1-2 standard deviations of talent to appreciate that talent (i.e., understand what is going on).
  - Second-order thoughts from that:
    - A technical recruiter, who’s job it is to solely find great engineers, must also be a great engineer. I think this is not common nor common practice. “Ok but how do you incentivize a great engineer in a non-engineer role”. Yes fair question. You don’t have a technical recruiter. You just have engineers hire and you keep growing until you can’t do that at scale anymore. If you have a +4 SD engineer and a +2 SD engineer on staff, remember you probably don’t need the +4 SD engineer to run interviews, get a +1 or +2 SD engineer to run interviews, close enough. If an average (or god forbid -1 SD engineer) runs interviews, that’s not ideal.
Around 36:11 he also talks about trusting people to go make smart decisions and ingest complex and “hard-to-hear” information and then thinking on second-order, there needs to be some amount of trust built and established such that you can say “If you are trustworthy and reliable, then we can really make some progress.”

Jensen Huang Interview at Stanford Institute of Economic Policy Research (SIEPR) Economic Summit, interviewed by John Shoven (Professor of Economics, Stanford)
This interview is pretty good but the main clip at ~36:00 that also got some attention is interesting on suffering and executive treatment through ~41:00.

“One of my great advantages is I have very low expectations. People with very high expectations have very low resilience. Unfortunately resilience matters in success. I don’t know how to teach it to you except for I hope suffering happens to you. You want to refine the character of your company. Greatness is not intelligence, it comes from character, informed out of people who suffered. If I could wish upon you Stanford students, I wish upon you ample doses of pain and suffering.”

How he also talks about working with the executive staff (E-staff) is super interesting.

55 direct reports (more than average)
No reviews written for any of them, gives them constant review, and they provide the same to him
Compensation is nearly exactly same for all (you can actually just go and check this, most times found in Proxy Statement but in some companies in the annual report it is formatted nicer, page 80, he’s right! Basically the same for those listed).
Don’t do 1-on-1s unless they need him then he drops everything.
No alone meetings. Never tell any information to E-Staff that isn’t relayed to rest of team/company.

Jensen Huang Interview with Patrick Collison, CEO of Stripe, at Stripe Sessions in April 2024 in San Francisco
I had the opportunity to attend Stripe Sessions (perks of living in SF) and watch this live, and then re-watch it a month later. Was a nice fireside, more comical which was nice. Jensen touched on some points he’s discussed previously on other podcasts, such as on hard work (referencing the Norges podcast), resilience (referencing the SIEPR clip), company culture (GSB fireside chat), and really on reasoning through problems (multiple live examples during the chat).

It’s a comical and interesting fireside chat and is pretty good through 26:38, after which it gets a bit technical, but actually the entire chat is rather enjoyable, but in my view the first half is more.

I liked the part at 18:54, not necessarily for the viral bit where people are like “Thank god I am not competing with someone like this” but rather how work for him is not all solving problems but dreaming and fantasizing about the future and what the company can be.

Thinking from this that life cannot just be about solving problems. It just seems circular (SCRUM anyone?). Drifting away from the normal practices…thinking creatively (what can I add or subtract from my routine processes)…really trying to build fungiblility into your non-fungible day-to-day (i.e., you have been doing something very process-like as a routine for 1 month/3 months/5 years…what would happen if you asked yourself if you could change something about it, what would it be? What could prevent you from making change?).
I think the answer to the debate of one side saying just do things, iterate, figure out vs. step back, think about boundary conditions, plan a bit is that both are equally important.

Some final takeaways / tying this back to my key framework (judgement, decision making, intuition, risk management).

I don’t take everything he says at face value but I try to independently think “Is what this person saying interesting to me or not” and so far some of this is interesting (or unique or contrarian or counter-intuitive or aligned with my prior thinking but more elequently phrased…)

It’s interesting that in grade-school, the goal is to show a kid how to do a math problem, not tell them the answer. While this is sometimes rather difficult (reminiscing about many years as a math tutor/instructor), it gets infinitely difficult when you need to guide/instill good principles and decision making tools toward other people when you are yourself (hopefully) iterating your judgement and thinking.

Until next time

-VS

Published on April 6, 2024.

Tagged: Principles