
Co-Founder at @VZVCfund. Investing in companies solving humanity’s biggest problems, led by founders hell bent on bending the curve.
Menlo Park, CA
Joined June 2008
- Tweets5.8K
- Following1.4K
- Followers36K
- Likes26.7K
Pinned Tweet
What should we build as AI, biology, and robotics reshape the world?
There is a tremendous opportunity in front of us. AI allows us to rebuild and fix society's greatest problems. Or gives opportunities for others to lock in what we hate.
For the past few years I've been exploring this as a scientist, investor, founder, and writer.
The result is a growing collection of essays on:
• AI and intelligence
• Venture and innovation
• The future of health and biology
• Institutions and society
• Meaning in an age of abundance
• Building the next Renaissance
If you're new here, don't start with the latest post.
Start here: vijaypande.substack.com/p/st…
I'd love to hear what resonates—and where you think I'm wrong.
Palmer Luckey says it well, but I partially disagree. We all know *that* guy:
“The guy who screwed me over is the people that I had hundreds of meals with ... The guy who’s worth tens or hundreds of millions of dollars because I bet on him, I put him in a position, I protected him .... Those are the people that I loathe.”
Where I disagree is that in the end, it's never really about that guy. It's about what you do after (and what they don't, if you're into spite).
Palmer's a perfect example of moving on and being way way better off for it.
Palmer Luckey on teaming up with Mark Zuckerberg again: “He’s not quite a god, but you kind of have to almost act as if he is. He's a force of nature.”
“The reality is Mark Zuckerberg is not the one who fired me. He runs this massive organization. I was, like, an update on his daily brief. ‘Oh, by the way, the people who run this division in your company, that’s a couple thousand people, they decided that Palmer Luckey needs to be terminated.’ ‘Oh, wow. Really?’ ‘Oh, yeah. There’s no way out of it. All of the executives say they’re going to quit if you make them work with him.’ ‘Is this a Trump thing?’ Okay.”
“There’s people who are like, oh, but shouldn’t he have stood up for you? No, man. He’s running a multi-hundred-billion-dollar company. His job is not to countermand the signal he gets from his deputies one, two, three levels down. Imagine if that’s what he did. I tried to do that, and it’s really difficult. So the people who screwed me over are not Mark Zuckerberg. You know how many times I had met Mark Zuckerberg? Like, a dozen. I worked with his lieutenants all the time, and I’d see him around.”
“The guy who screwed me over is the people that I had hundreds of meals with, even in the middle of the night, eating crappy Korean food while we jammed to try and get stuff done for CES, the Consumer Electronics Show. The guy who’s worth tens or hundreds of millions of dollars because I bet on him, I put him in a position, I protected him as the Facebook politics tried to replace his job with someone else. Those are the people that I loathe.”
“Mark, he’s not quite a god, but you kind of have to almost act as if he is. He’s a force of nature. He exists outside of the universe of what Oculus was at the time. And so if I can make a deal with Mark that says that all the tech I sold him can be used to help the US Army, and that all of the IP that I sold him can be licensed into that, and that all the tech they’ve built on top of what I sold to them can also help our troops stay alive, I am going to make that deal.”
Genesis shows they continue to be best in class in potency prediction, while being 40x to 1000x faster.
Huge congrats to @enfeinberg & team, especially with the @Incyte results and partnership.
R U Binding Yet?
Meet Ruby 🔴, our new potency prediction model and the newest member of the GEMS family.
Ruby answers every drug hunter’s first question: R U Binding Yet? 🎯
With our partners at @Incyte , Ruby showed a step-change improvement on real-world drug programs. It runs 40x to over 1,000x faster per compound at its fastest setting, and leads every published model we could find on public benchmarks, free-energy methods included.
📂 We’re also open-sourcing GenesisPotency-2026, part of the audited public training data behind it.
🖥️ Thanks @nvidia for keeping us snappy!
Report: cdn.prod.website-files.com/6…
Dataset: github.com/genesistherapeuti…
Blog: genesis.ml/news/introducing-…
There’s a massive irony here.
NVIDIA’s moat is its software ecosystem: CUDA etc.
But with LLMs coding (on originally NVIDIA GPUs) one can rewrite the back end to run on other GPUs as software is no longer a barrier.
The moat’s skills kills the moat.
SITUATION EXPLAINED: Anthropic's IPO roadshow says Claude got itself running on AMD chips in 3 days with almost no human help.
• Anthropic runs Claude on 4 chip families: Amazon Trainium, Google TPUs, NVIDIA GPUs and AMD GPUs
• Google and NVIDIA are among 7 partners buying $26B of Anthropic stock at the IPO price
• Claude wrote or ported almost all of the low-level code for the AMD switch, a job that used to take dozens of engineers months
• AMD agreed in July to supply up to 2GW of its MI450 GPUs and invest up to $5B in Anthropic
• Anthropic set its IPO range at $70-$78 a share, valuing it at up to $1.4T fully diluted, with trading expected around Nov 19
• Its annualized revenue run rate hit about $81B at the end of October, up from about $71B a month earlier
@theojaffee: "That seems to imply a pretty sizable weakness for NVIDIA if you could just easily run your models on any kind of chip."
Jensen is impossible to replace, but Ian Buck would be amazing.
I worked with him at Stanford and was on his thesis committee. His work with Pat Hanrahan was the engine behind the GPU core of Folding@home, PRE-CUDA.
A+ execution and vision.
Ian Buck looks like a fantastic candidate to take over as CEO of Nvidia.
He's been working there since before he finished his Phd in CS from Stanford in 2005.
He helped create CUDA (Nvidia’s biggest competitive advantage) and has been there for the entire transformation from a gaming GPU company to AI. He's given a bunch of keynotes and is a capable speaker and has technical credibility, which is great for talent and the developer ecosystem and customers.
linkedin.com/in/ian-buck-192…
Somewhere, someone is saying "Mission Accomplished"
@WSJ says "Nvidia is Preparing to Oust CEO Jensen Huang"
wsj.com/business/nvidia-is-p…
Everyone who won on Tuesday is about to call it a mandate. That’s step 1 of 10 of How to "Win" Alone.
I've seen these same mistakes in meetings and at home. The 8th is the one that costs you the room.
10 steps, 10 drawings, whoever you voted for: vijaypande.substack.com/p/ho…
In an election, “How do we win together?” makes no sense. There are winners and losers. I expected books on war to agree. They didn’t.
The answer starts with who the “we” is. MLK turned that into a strategy that changed our nation. So can we.
Here’s how:
vijaypande.substack.com/p/ma…
Good arguments don’t move people first. Feelings do and the reasons come after.
Florence employed beauty to build a Renaissance. Nazis used it to drive their evil.
How we can build a Renaissance scale movement and avoid humanity's most notorious traps:
vijaypande.substack.com/p/th…
For decades I've felt like a robot dedicated to getting my calendar done. Sound familiar?
There are two ways to lose yourself to AI, and only one of them is losing the job. The other is keeping it.
A new way out:
vijaypande.substack.com/p/th…
Terrence Tao nails it I think, not just for math but other fields driven by creativity -- from heroes to industrialization.
Some good parts will be lost (especially for who would have been the future heroes) but much will be gained, especially for humanity as a whole.
Mathematician Terence Tao:
"we're seeing the end of the romantic era of mathematics, the age of heroes. we're moving into an industrial era where math becomes large-scale, done in big teams like the other sciences"
We're entering, rather unwillingly, a new era where exciting new things are possible, but some things are being lost
Why did the Star Wars prequels fall short? By 1999, nobody could tell George Lucas how to do Star Wars.
AI is about to create that at scale: power without anyone who can tell you you’re wrong.
What it takes to create people we can trust with power: vijaypande.substack.com/p/wh…
How to Engineer a Soul.
We didn't stop having religion. We unbundled it. Therapy kept confession. Wellness kept ritual. Politics kept the tribe. Then each part was sold back to us separately.
The whole machine, and the seven steps to rebuild it:
vijaypande.substack.com/p/ho…
AI Will Make Your Kids Powerful. But Who Will They Become?
Every renaissance gave its teens someone better to become, defining its "ideal person" from what the machines can't do.
Tonight, machines nobody answers for are drawing it.
Here's a way out:
vijaypande.substack.com/p/ai…
I forgot about this tweet, but the Doomer Industrial Complex was clear to many of us even in 2023
2
5
1,719
France and Germany did not achieve peace by trusting each other. They began by sharing coal.
Trust was not the prerequisite. It was the product.
My "Counterfeit Hero" finale asks the question that changes the game and player:
How do we win together?
vijaypande.substack.com/p/we…
History did not wait for saints.
Florence was full of ambition, vanity, and rivalry. But so was Rome.
What differed was the Status Market (what an ambitious person had to do to become important).
We don't need better people.
We need a better game:
vijaypande.substack.com/p/we…







