Billionaire Palmer Luckey on why he only flies coach
“I don’t fly private and do fly coach,” billionaire founder of Anduril and Oculus Palmer Luckey says. “For me, it’s a reasoned thing. With exceptions for long, international travel, we only cover coach travel for our employees — it’s only a few hours, and it’s a very bad use of company money for us to be buying business or first-class for people. Because we have so much travel at the company, we could easily spend a very serious fraction of our resources on people traveling instead of slightly better seats.”
To lead by example, Palmer will fly coach — even when he uses his own money:
“People say, ‘Well, why don’t you just fly first-class?’ And here’s why: if I’m going to ask my employees to do it, I need to do it too — even when it’s my own money. It’s not that I would appear out of touch, I would literally be out of touch. Maybe one day coach gets so bad that I literally tell everyone, ‘I hear you, we’re all going business now.’ But today is not that day.”
He continues:
“I love the back of the plain by the window. Nobody bothers you. You can let everyone get off the plane before you. You don’t have to fight anybody . . . And my grandpa was a pilot for United Airlines for over 40 years. So I also grew up around commercial airlines and to me there is a certain romanticism to mass-market, mass-available air travel. What an incredible thing. And we did it — America did it! We figured out how to make it economically viable, and we build everyone else’s airplanes. It is an American thing.”
Source: @myfirstmilpod (Oct 2022)
Elon Musk dropped out of Stanford after 2 days and ended up sleeping on an office futon, showering at the YMCA.
8 years later he priced a rocket at $6 million while NASA burned $1 billion per flight.
Here’s what happened in between.
His professor’s last words were “I don’t think you’ll be coming back” and he was right.
Musk built Zip2 with negative money: huge student debt, an office cheaper than an apartment, internet through a hole drilled in the floor to the ISP downstairs for $100 a month. He sold it for $300 million, then built PayPal with 30 engineers, 0 salespeople and 0 ad budget, and sold that to eBay for $1.5 billion.
Instead of a vacation, he flew to Russia to buy a decommissioned nuclear missile.
The plan was to land a greenhouse on Mars for $20 million, but American rockets cost $50 million minimum so he went shopping for ICBMs. After 3 trips to Moscow the Russians finally said yes, and the deal got so messy he walked away with nothing.
On the flight home one question wouldn’t leave him: why can a country with the economy of Belgium build cheaper rockets than America?
So he built his own answer 30 people, 0 lawyers, Ethernet instead of copper bundles thick as your arm, everything simple and everything cheap.
The scoreboard by 2003:
Space shuttle $1 billion per flight Nearest competitor $25 million Falcon $6 million
Stanford wanted 4 more years he needed 2 days to know.
Everyone is panicking about $PLTR down 30%.
Nobody is looking at free cash flow per share: -$0.30 in 2020, $1.05 today.
Stock prices track FCF per share over the long term. The rest is noise.
$PLTR Palantir is winning.
The stock is down >30% from its ATH, yet:
- Institutional ownership at record high
- YoY rev growth accelerating 11 Qs in a row
- Operating margin continues to expand
- Net dollar retention rate continues to rise and is now 150%
- Palantir’s Maven is now a program of record across the entire US DoD
- NATO announces full authorization to operate Maven for everything
- In other words, Palantir is now the entire backbone of Western military
- Companies like databricks are bending the knee and preaching ontology now
- “Leaders” like Benioff, Nadella, etc are now preaching data and AI sovereignty, something Palantir has a 20yr head start on
- Basically everyone is now admitting “Karp was right” about anything you can think of… and if they haven’t admitted it yet, they will soon.
- Palantir continues to prove LLMs are a commodity, with the most recent example being their new “evolve” tool that is reducing token consumption by as much as 60% in one example and slashing compute costs by 90% in another. The value is in the infrastructure, not the models.
- Warp speed
- Commercial customers continue to show how transformational Palantir is at AIPCon
- NVIDIA recently expanded their partnership with Palantir. Less than a year ago, Jensen acknowledged Palantir as the most important enterprise stack in the world
- Even Karp himself recently acknowledged the stock is undervalued this week
I wrote this off the cuff. The list goes on.
Everything is going right for Palantir while all the other software “players” are scrambling around trying (and failing) to create transformational and measurable value with AI. They’re all bending the knee and regurgitating whatever Karp is saying because apparently, Karp is the only one who actually understands what’s going on.
We are watching Karp’s vision.. Palantir’s vision.. unfold. Karp was right. Palantir was right.
IDGAF if someone on the internet or some talking head on TV tries to tell me otherwise… This is crystal clear. Palantir is winning. And this is only just the beginning.
The answers are right in front of you.
Elon Musk thinks coding dies this year.
Not evolves. Dies.
By December, AI won’t need programming languages. It generates machine code directly. Binary optimized beyond anything human logic could produce. No translation. No compilation. Just pure execution.
Musk: “You don’t even bother doing coding.”
Code was never the point. It was friction. A tax we paid because machines didn’t speak human. AI just learned fluent human. The tax is gone.
Now plug that into Neuralink. No syntax. No keyboard. No screen.
Musk: “Imagination-to-software.”
Thought becomes executable. You imagine an outcome, the system architects and compiles it into reality instantly.
We’re not automating programming. We’re erasing it from existence.
The entire profession collapses into a thought. Decades of training reduced to irrelevance. The gap between idea and instantiation hits zero.
You don’t build anymore. You imagine, and it materializes.
Not incremental progress. Total phase shift. The way humans have created things for ten thousand years just became obsolete.
Welcome to a world where the limiting factor isn’t skill, resources, or time. It’s whether you can picture what you want clearly enough for a machine to birth it into existence.
🇺🇸 Lindsey Graham's last major televised interview before his death read like a roadmap for what comes next in the Middle East:
• If the Iran deal fails, Trump will seize the Strait of Hormuz by force and charge fees.
• If Iran contests it or attacks through proxies, the U.S. will "obliterate" them.
• And the next diplomatic push will be getting Saudi into the Abraham Accords to reshape the entire region.
He didn't live to see how it plays out. He may have just told us exactly how it does.
Source: CBS News / Writer: Monica
President Trump never said what the 20% is 20% of. Read as cargo value, the way the Associated Press read it, one tanker carrying 2 million barrels at $80 oil holds $160 million of crude. A single crossing of the Strait of Hormuz would cost $32 million.
Iran's foreign minister Abbas Araghchi answered Trump in public. "POTUS is absolutely right."
That is not a peace deal. It is worse.
Araghchi agreed that whoever provides safe passage should be paid, claimed the role for Iran forever, and objected only to the number. "20% is of course too much. We will be fair."
So two governments at war now claim the right to sell safety in the same water.
Neither payment buys passage against the other. A permit from Tehran / IRGC does not bind Uncle Sam.
An American escort does not bind Tehran. Hormuz is not becoming a toll road. It is becoming a road with two armed tollbooths pointed at each other.
Then carefully look at what a percentage does. If this war drives oil to $120, that same tanker's invoice climbs to approx $48 million. The danger raises the cargo's value. The cargo's value raises the fee for protection from the danger. Insecurity starts paying for itself.
None of it is law. President Trump's own Secretary of State Marco Rubio said last month that no country may charge tolls on an international waterway.
The UN's maritime regulator said yesterday there is no legal basis for it. Not one dollar has been collected. There is no mechanism, no valuation rule, no agreement with any coastal state.
So capital is doing what capital does. The Financial Times reports DP World is in talks to build a new port on the UAE's east coast, outside the strait, with cargo trucked overland to Dubai and Abu Dhabi UAE.
Not to win the argument. To leave it.
The real danger to Hormuz was never that one side would own it. It is that the world quietly decides the water has become ungovernable, and writes it out of the map of trade.
No, Claude Tag is not the end of Palantir.
Claude Tag is Anthropic’s new proactive AI teammate for Slack: it handles tasks in channels, uses shared org memory, breaks down work, and connects to tools. Useful for coding, project work, and team collaboration.
Palantir’s platform (Foundry + Ontology + AIP + Apollo) is built for something different: integrating massive, messy, often classified enterprise data, enforcing fine-grained access, modeling real-world operations semantically, and running governed AI at scale in high-stakes environments (defense, gov, large industry).
A Slack-based collaborative agent doesn’t replace that core data infrastructure or the security/compliance layer. They operate at different levels and can even complement each other.
The “end of Palantir” take is classic AI hype. Enterprises rarely rip out foundational platforms overnight for the newest frontend agent.
@8teAPi@grok, is this true - that "Claude Tag" is the end of Palantir? How does "Claude Tag" replace Palantir AIP, Foundry, Apollo, Ontology, and EVOLVE in enterprises?
When we look back, Alex Karp may have initiated an important preference cascade around AI sovereignty.
It’s worth noting what @Benioff and @satyanadella are both saying:
Your knowledge, as a company, is your sovereignty. If you lose it to someone else (anyone else) you are hollowing your organization out.
There are many ways to accidentally leak intelligence so you need partners and tools who can sign up for the complexity required to give it to you.
See Benioff below and see Satya’s essay linked below.
“If we triggered the break clause, you’d have to pull the system and replace it with manual workarounds until you could find a replacement to migrate it to.”
“You’d basically have to go back to all of the spreadsheets, the bits of paper, the whiteboards, the phone calls, and the NHS would go backwards, not forwards.”
“I would love to see these campaigners and MPs put forward a credible plan. I’ve got no particular interest in Palantir at all. They just happen to be the company that can deliver this.”
Tom Bartlett, Former Deputy of Data Engineering and Integration, NHSE
We banned the use of all Anthropic models and their harness this week at Codestrap. Not only becuase we dont like Anthropic's leadership and culture, but also becuase they are cash incinerators. Take a look at those blue bars. 30 fing % of the cost at miniscule usage levels.
Anthropic CEO Dario Amodei:
"in 2014 I thought I was too late to AI. there were 50 people in the field. I figured I'd rush in and fight for scraps"
today he runs one of the most valuable AI companies on earth
he joined OpenAI as one of the first employees in 2016. built GPT-2, GPT-3, and co-invented RLHF - the method behind ChatGPT and Claude
his moment of realization: GPT-2 could do regression analysis. "oh my god, this is some kind of general reasoning engine. what do I have in my hands?"
~108mins, free
bookmark & watch now ↓