When @JensenHuang returns to $NVDA, it is my belief that he will do so with class, contrition, and commitment. I do have one ask: "Lose the leather jacket; your character and accomplishments speak for themselves without the symbolism." @TeslaBoomerMama @elonmusk @DavidSacks @XanderSky @Amy787 @p_ferragu @TeslaBoomerPapa
4
8
3,969
Tom Fraser retweeted
We were taught that Earth orbits the Sun in a closed circle. In reality, Earth has never returned to the same point in space once in 4.5 billion years.
Right now, the Sun is carrying Earth and every planet around the center of the Milky Way at 230 km/s (828,000 km/h), more than 18,000 kilometers every 80 seconds.
Because the Solar System is tilted 60.2° to the galactic plane and bobs up and down through the galactic disc roughly every 30 million years, the planets trace open 3D corkscrew helices across deep space rather than flat loops.
The last time the Sun stood where it stands today, one Galactic Year (230 million years) ago, the very first dinosaurs were only beginning to appear on Earth. Since forming 4.6 billion years ago, our Sun has completed only 20 laps.
Tom Fraser retweeted
All-In Summit: Naveen Rao
-- Why AI is extremely inefficient, and how to fix it
-- AI is about to run out of power
-- Your brain does this on 20 watts, the data center needs a city
-- Every computer since 1945 is built wrong
-- Why the doomers have the energy story backwards
++ much more!
(0:00) Welcome @NaveenGRao
(4:45) Is energy really the problem? The cost of a token, power contracts & the gap to close
(10:24) Cutting out the middleman: abstractions, dynamical systems & a new kind of machine
(19:40) Chamath joins: the path to product, porting existing models & building the team
Be prepared. Wall Street may take us out to the woodshed.
Tesla
Wednesday's earning call
In my opinion, Q2 numbers will show heavy investment mode
- WS sees free cash flow around -3.25B, my estimate is closer to -5B.
- Cash position according to WS ~$41B, think it could drop to @ $39B .
Capex guidance 2026 went last quarter from $20B to $25B … wouldn't shock me if it heads now toward $35 - 40B
This is how the best leaders build the future.
Robotaxi, Optimus, energy — full speed ahead!
5
753
In the 80s and 90s Japan created its competitive advantage brilliantly. They took on the bloated auto monopolies with asymmetric products and used their cultural strengths of pride in quality, teamwork, and an honest days labor. The product differentiation has dissolved. Recommended reading: The Reckoning by Halberstam.
Tom Fraser retweeted
This is Elon telling you the future.
A mass driver is basically a giant electromagnetic catapult. You build a track on the Moon, run current through superconducting coils, and yeet payloads into space at 5,300 mph. No fuel. No rocket engines. Just electricity.
Gerard O'Neill built the first working prototype at MIT in 1976. The tech is 50 years old. The economics are what changed.
Here's the math:
Falcon 9 costs $2,720/kg to low Earth orbit. Starship targets $100/kg once fully reusable. A lunar mass driver, powered by solar and running continuously with zero propellant? The Space Studies Institute estimated $1 per pound back in 1979. Updated for modern superconductors and solar efficiency, we're talking single digits per kilogram.
The Moon has over 1 million metric tons of helium-3 (worth $2,000 to $20 million per kg on Earth). Titanium. Aluminum. Iron. Silicon. Rare earths. Water ice at the poles.
So what's the timeline?
Artemis III lands astronauts at the lunar south pole around 2027-2028. SpaceX wants a permanent base by early 2030s. NASA has plans for a 100-kilowatt nuclear reactor on the surface by 2030. China is racing to build their own lunar station by 2035.
Once you have power and people, a mass driver is just construction. O'Neill's original designs called for a track a few kilometers long. A 160-meter track can reach escape velocity at high g-forces. Modern estimates suggest 12 tons of equipment landed over 20 years could bootstrap a self-expanding lunar industry.
The realistic timeline? First operational mass driver by 2045-2050. Maybe faster if the space race with China heats up.
And here's the part nobody's pricing:
A functioning lunar mass driver can throw 600,000 tons of material per year into cislunar space at near-zero marginal cost. That's O'Neill's 1979 estimate with conservative tech.
At that point, the supply curve for critical resources inverts. You stop asking "how much does it cost to lift this from Earth?" You start asking "can we mine it on the Moon?"
Mining on the Moon, with 1/6th gravity and no atmosphere, gets cheaper every year as robotics and AI improve. Launching from the Moon gets cheaper as solar panels improve. The cost curves only go one direction.
Elon's "money becomes irrelevant" framing sounds crazy until you think about what happens when energy and raw materials both approach zero marginal cost. Every economic system ever built assumes scarcity of stuff. A mass driver breaks that assumption for anything you can make from lunar regolith.
This is why Starship matters as a bootstrap mechanism. You need cheap Earth-to-Moon transport to build the infrastructure that eventually makes rockets obsolete for bulk cargo.
Rockets get you to the Moon. The Moon gets you everywhere else.
And Elon just told you that's the plan.
Replying to @IterIntellectus
When the mass driver on the Moon gets going, I’m not sure money will be relevant
Tom Fraser retweeted
The U.S. Military to Study Bitcoin
In this document...
In section 1543...
The U.S. military is being told to study how to make cyberattacks against America’s military systems so painful and costly that enemies decide it’s not worth trying.
Cyberattacks are now a real battlefield. The U.S. military needs a clear plan to scare enemies away by making cyberattacks against us extremely costly.
SEC 1543 says the U.S. military should find ways to make cyberattacks so costly that enemies stop wanting to do them. Bitcoin helps do exactly that.
Doing things that cause economic, political, information, or military pain to an attacker so big that they change their behavior.
Bitcoin doesn’t attack anyone - but it makes attacks pointless and expensive.
How Bitcoin “imposes costs” on cyber attackers
Most cyberattacks on critical infrastructure try to cause financial damage:
- Freeze money
- Steal funds
- Demand ransom
- Crash payment systems
- Create panic and chaos
Bitcoin protects against this because it doesn’t have a central owner or switch to turn off. There’s no bank to hack, no headquarters to shut down, and no government that controls it. If the U.S. uses Bitcoin for backup payments, reserves, or emergency operations, attackers lose the ability to cause financial paralysis, which makes their attacks less effective.
This directly raises the economic cost of attacking U.S. infrastructure.
Why attacking Bitcoin is too expensive
To seriously damage Bitcoin, an attacker would need:
- Massive amounts of electricity
- Specialized computer chips
- Control over computers spread all over the world
That’s not a cheap cyberattack anymore - that’s a huge, real-world operation. Instead of writing software from a laptop, attackers would need factories, energy grids, and global coordination. That’s exactly what SEC 1543 means by “increasing the costs” of cyber aggression.
Bitcoin forces attackers out of low-cost digital attacks and into high-cost physical reality.
How this protects critical infrastructure
If U.S. military systems know they can pay people, move resources, store value and keep operations running even when banks, networks, or governments are under attack, then cyberattacks don’t achieve their goals. When attacks don’t work, attackers stop trying.
So, Bitcoin helps the U.S. military impose costs by turning cyberattacks into losing strategies. It removes easy financial targets, forces enemies to spend real-world resources, and keeps critical systems working even under attack. When attacks stop paying off, incentives disappear - exactly what SEC 1543 is asking for.
Tom Fraser retweeted
Link to referenced document: armedservices.house.gov/uplo…
364 days ago on 12/17/24 $TSLA posted its all-time high closing price of $479.86.
1
43
Tom Fraser retweeted
The IIHS has released a new list of the top 5 safest large pickup trucks you can currently buy.
1) Tesla Cybertruck (electric)
2) Toyota Tundra (gas)
3) Rivian R1T (electric)
4) Ford F-150 (gas)
5) Ram 1500 (gas)
The @Tesla Cybertruck just received a Top Safety Pick+ rating from the IIHS, the best possible rating, and was the only electric truck to get that rating. Earlier this year, the Cybertruck also received a 5-Star safety from the NHTSA. The Cybertruck had the lowest overall probability of injury & lowest chance of rollover of any pickup truck tested by the NHTSA.
Tom Fraser retweeted
Replying to @bonchieredstate
I have a hard time understanding cartoon Trump.
He's done so many things that I agree with and then he comes up with the most perplexing ideas that I cannot wrap my brain around.
I'll just have to stick with "better than Harris" and grit my teeth, but my gums are getting sore.
Tom Fraser retweeted
Best two minutes you’ll watch today. xitter.cf/gotrice2024/status/196…
Tom Fraser retweeted
Fox News Reporter Jacqui Heinrich: The way that it felt in the room was not good. It did not seem like things went well. And it seemed like Putin came in and steamrolled, got right into what he wanted to say and got his photo next to the president and then left
Tom Fraser retweeted
Replying to @elonmusk @CernBasher
We still would like to be able to design and save our own routes please Elon!!!
Tom Fraser retweeted
“Ph.D in math” yet he said below one of the stupidest things I’ve ever heard a person say.
Replying to @CynicalPublius
There's nothing intrinsically superior about wetware electrochemical brains over the silicone based ones. That humans can do it means it's a solvable problem. I bet it'll be solved even more quickly than some can now imagine.
Tom Fraser retweeted
An interesting footnote in the AI wars will be how The Bitter Lesson applies to the human capital working on AI.
On one side is this hiring spree (Google, Microsoft, Meta) and the kinds of hires (PhDs, human experts) being made right now vs the other side the xAI approach (computer and synthetic approaches).
Pay one person $1B OR buy $1B more compute.
Tom Fraser retweeted
This is despite @SpaceX competitors increasing their launch rates. SpaceX is an anomaly in the matrix.
Nothing can compare to a system designed to make life/consciousness multiplanetary.
Tom Fraser retweeted
TESLA ENERGY
Obvious to everyone that has researched the scope of the planned AI compute centers vs the available power from grids—there will be a shortage of power.
HAVES —AND HAVE NOTS
Companies that move fast to secure energy, and stationary storage like MegaPacks (to both ensure steady power and buffer it) will be the haves. Everyone else will be the have nots.
UTILITIES ARE BEHIND
Rather than planning new power plants to meet anticipated increase in demand, some US states—and some countries—are shutting down power plants because they think they have “enough” energy. Meanwhile. China is building new power plants at a rapid pace. The US, and most of the West, risks falling behind. This doesn’t have to happen. The problem is foreseeable, and therefore avoidable.

























