I like to use "shoveling dirt" to demonstrate technological progress.
It's unglamorous work but it's something humans have been doing since the very beginning.
Now apply the same scaling metric (shovel to robot fleet) to super intelligence.
Imagine what you could get done.
Theoretically, Canadians should be as wealthy - or wealthier than - their US counterparts.
Indeed, Canadian per-capita GDP was on a tear until about 2011 when oil prices collapsed.
Canada decided to invest in more housing and services for immigrants instead of industrial capacity, believing that was the best way forward.
It wasn't.
My wife's family is from the Toronto area, so I get a direct view into how Canadian govt policies directly affects their lives.
It's very European. They really notice the difference when they come to the US.
It's cold up there, you might say. Things are different. Well, so is Alaska and they're doing far better.
First chart is US vs Canada per capita GDP.
Second chart is Alaska vs Canada per capita GDP.
It looks like Canadians voted to be poor, just like their European counterparts.
Do I believe that there's a well-organized and well-funded group actively planning and sponsoring color revolutions both here and abroad?
The evidence is overwhelming, so -- yes.
And there's ample incentive. Controlling a country's population through force can be very profitable to those doing the controlling.
Just ask any billionaire despot.
What is amazing is that they're confidently moving forward despite having been totally and utterly exposed here on X thanks to @DataRepublican and others.
This means the bad guys all believe
(a) their plan will work despite being known, and
(b) they won't suffer any consequences.
That's a really bad bet on several levels.
First, while the Deep State has enjoyed a great run over the decades, times have changed. Past tactics aren't as effective as they once were. Covert patterns are becoming much easier to detect and thwart.
It's not the 1980s and this isn't Nicaragua.
Presuming the Republicans hold power through 2028 and beyond, I'm sure rooting these people out will become a national security priority.
Side note: these same silly games are sometimes played out in the corporate world. They usually don't end well for the instigators.
SHADY! Elissa Slotkin’s 2024 win always felt off. They needed her apparently by any means necessary. Now the former CIA officer is the Left’s go-to voice for color-revolution tactics.
She flew to Poland this summer to study how to remove or neutralize Trump. Weeks later, Poland faces its own attempt to topple a government the Left doesn’t control.
This isn’t coincidence. The agenda is global: get every country marching in lockstep. The Right won’t go quietly, so they’re accelerating the pressure.
Gaming a betting platform is cheap, just like buying phony polls.
Grok tells me there's about $4.1m open interest on this race, which means $1m would really sway it.
More people should be taking the short odds, it's easy money in your pocket.
BREAKING: James Talarico soars to new high in Texas Senate election odds, just 3 days before the midterms.
70% chance Texas flips blue.
🔗 polymarket.com/r/TXSen26
Elon has figured out how to provision new GPU capacity faster than anyone else -- exactly when the market is paying an insane premium.
Analysts are forecasts ~$200B ARR by the end of 2027. It's a capital intensive but high margin business. Once a data center is built and running, it turns electricity into predictable revenue.
I'm convinced this is a side quest for Elon, simply to generate enough capital to fund the projects that really matter to him, just like "Not A Flamethrower" but on a much grander scale.
He can now borrow massive sums against this revenue stream. He can now more easily fund data centers in orbit, manufacturing on the moon, and eventually colonizing Mars -- all while transforming society here on Earth.
I'm not sure how many people realize how insanely intelligent Elon is.
Replying to @herbertong @CernBasher
Macroharder and Minihard both have 220k GB300s with 800 gig nics configured for training. Gun Range will have the same. Spam (used to house Hormel products) will have ~200k GB300s, but configured for inference, so much simpler networking and less than half the fiber.
I walked the whole Colossus 2 site around midnight and again this morning.
Good progress by the SpaceX team and contractors. Gun Range is tracking to be operational by end of November and Spam in December.
Not a hot take at all. Anyone who has supervised a team of superstars knows that it's really hard work.
LLMs are a reflection of how we think, for better or worse.
Changing the value of a few intermediate "neurons" can produce wide swings in behavior.
The more I study LLM behavior, the more I learn about human behavior.
My answer to Fermi's Paradox?
If there are many trillions of galaxies each with many billions of stars, planets, etc. -- why haven't we been contacted by alien life?
I think it's because we haven't advanced nearly enough yet.
Basically, it's an "us" problem. Held at arms length, we're a sloppy, emotional and violent species.
I've found living in Florida generally awesome.
It compares well with the SF Bay Area, the Pacific Northwest, the Boston burbs, Boulder and a few other cool places I've lived.
I'm in a beach resort town where every day feels like, well, living in a beach resort. It doesn't suck.
It's a beach day today. This morning everyone was out in the great warm weather hiking, biking, rowing, fishing or whatever they do outdoors.
Classic cars and motorcycles everywhere. Lots of expensive toys as well. Young families out and about. Everyone does their thing.
When people ask "what do you do?" they mean what do you do to keep yourself busy, and not necessarily what line of work.
There are golfers, boaters, pickleballers, tennis players, cruise enthusiasts, charity leaders, travelers and so on. And, of course, plenty of folks still holding down a job.
I run a great local band. People love good live music here. It's something I've always wanted to do.
By default, most everybody is happy, polite and engaging. It's expected that you'll make small talk with just about everyone you meet, including service people.
There are at least 40x more churches in town than liquor stores or pawn shops. Lots of people carry. It's about a safe a place as you could imagine.
They call my town "Mayberry By The Sea", and there's a certain truth in it. Everyone loves being here and wants to keep it the special place that it is.
The town and county aren't anti-growth, but they do keep a close eye on things so Vero Beach doesn't end up like West Palm Beach.
In all fairness, the summer humidity can break you after a while, though. If you have a place to escape to farther north, that's ideal. The snowbird thing works.
Just my two cents for all you retirement-adjacent folks who might be looking around for your piece of paradise.
Lutz nails it here.
Many conservative voters refuse to participate in the polling charade and lump obviously biased results in the media under "more fake news".
What matters more than someone's opinion is who is engaged enough to vote, and who isn't.
And the only way to determine engagement is via focus groups. You literally pay people to sit in a room and answer questions honestly.
At one point in my career I was running dozens of focus groups a year, then later would write big checks to have professionals do this on our behalf.
The other huge flaw in political polling is that there's FAR more money behind swaying the results than determining reality.
🚨🇺🇸 Pollster Frank Luntz is NOT buying the blue wave...
He says Trump voters have a tendency NOT to participate in polling, and that many of those polled may not actually end up voting.
He also cast doubt on whether pollsters are even using the right measurements or asking the right questions.
Underestimating GOP turnout by even a small turnout could mean the blue wave is more like a blue puddle...
We'll know who's right in 13 days!!!
Writer: Michael
I would suggest that all investors stay a safe distance away from Anthropic.
They're not going to be a winner. It might be a good time to sell the company to IBM or similar.
Training costs wiil continue skyrocket and market prices for inference will continue to drop.
Just like any gold rush, the cost of production continues to skyrocket while market prices fall.
By comparison, Elon is building the full stack: chips, power, data centers in space and rockets to get them there.
Now that's a moat.
Yesterday he shared that he'd put his best team on building gas turbine blades, which is the current bottleneck to AI scaling. And that Terafab is moving ahead quickly.
Open source is moving fast as well. Deepseek is wild, they've released a "service provider" stack: most everything you need to stand up AI as a service.
And there will be hundreds if not thousands of people doing this very inexpensively.
If I were an investor, I'd consider shorting the Anthropic IPO.
🚨 ANTHROPIC's NUMBERS ARE CONCERNING -
Revenue ~$16,300,000,000
Net Loss ~$80,700,000,000
IPO Valuations ~$2,000,000,000,000
@grok How long can it run before declaring bankruptcy if it keeps on buring like this?
I'm still processing everything Elon said during the first 20 minutes of the SpaceX earnings call.
It was surreal.
It should be named themost mind-blowing earnings call in history (after his rocket car demo) and ended with him throwing out a $500B - $1 trillion ARR.
Apple is running at $466B, Amazon at $717B.
Just watching him talk and think is fascinating.
teapawt.org/i/broadcasts/1PKqrNbLZ…
And with DeepSeek's release, they've created a new category of "independent AI service provider" -- smaller players who can provide a higher level of customer service at a more reasonable cost than the big guys.
Wait six months, they'll be everywhere.
After watching Dems get over-polled for decades, I asked myself "why is this"?
It can't be completely explained away by "pay to say" pollsters. Respected names get it wrong regularly as well.
Frank Lutz nailed it: there are conservatives who won't share their opinions with pollsters. I am one of them, and I know there are many others.
Some will try to apply a correction bias of ~3.4%, supposedly based on past polling failures, which is a weak attempt to recognize the phenomenon.
You can't measure what refuses to be measured. there is no way to know how many of us there are, or if we will turn out to vote.
The only poll that matters is on election day.
💥NEW: NBC's Steve Kornacki *DUMPS COLD WATER* on The Today Show viewers' hopes of Dems winning Senate💥
Kornacki noted Democrat leads in key Senate races are "tenuous" — and that polling frequently underestimates Trump voters.
He asserted Republicans now confidently view polls where Democrats have slight leads as likely GOP victories.
Kornacki gave specific examples, such as Ohio and Iowa.
Remember the infamous Ann Selzer Iowa 2024 poll?
Kornacki particularly highlighted the Kansas Senate race in 2020. He said the Democrat was polling well — but Republican Roger Marshall ultimately smoked her in a landslide.
Marshall is now supposedly polling extremely tightly against his Democrat opponent once again.
Kornacki also pointed out that even blue states like Michigan and Minnesota appear remarkably close in polling.
He reminded viewers that Abdul El-Sayed barely won his primary after polling projected a huge victory.
Due to the tightness of these Senate races, underestimating Republican turnout by even a small margin could make all the difference.
Frank Luntz also recently warned Trump voters "don't respond" to pollster phone calls and asserted he was "skeptical" that polls were accurate.
Nobody is buying the polls anymore.
Democrats could be in for a rude awakening in 2 weeks.
One of my favorite Muskisms is the ability to predict the future being the highest form of intelligence.
I've given that a lot of thought, and it's painfully true.
I've seen many otherwise bright people get blindsided by easily predictable events. "What happens next?" can be really hard for some.
AI (now evolving into true super intelligence) won't trigger mass unemployment, but will force many to be supervisors rather than individual contributors.
Every bot team needs a human involved. There's a limit to how much bot work a human can inspect. More bots, more human supervisors/inspectors.
No supervisors paid to pay attention, no useful bot work. Just like with human workers.
The total amount of useful work performed by the workforce (a rough proxy for GDP) will increase by one order of magnitude, or more.
Plenty of jobs for those who can do it, or are willing to learn. Bot herders will be paid much more than non-bot herders for obvious reasons.
The predictions around having to work less and make more are likely the most accurate ones.
🤖 Made with AI
Their "networks" will likely be rebranded Starlink services.
The only company putting enough satellites in space is SpaceX for the foreseeable future.
Mobile companies have two core competencies: running a network, and billing for it.
Now they're just going to be retail outlets for Starlink services.
BREAKING: Four European telecom giants are joining forces to challenge Elon Musk’s Starlink.
Deutsche Telekom, Orange, Telefónica and Vodafone have signed a letter of intent to explore a joint bid for spectrum to deliver satellite connectivity directly to phones.
Four telecom giants. Still no Falcon 9. Still no Starship. SpaceX has no competition. Period.
This is great.
Very few mid-level dweebs are going to risk federal prosecution by knowingly committing a crime.
The obfuscation is fun to watch.
Previously, the CA Secretary of State "enforced" the rules. Hahaha. Ooops, now it's the DOJ.
It's simple: if any of the 20k suspect ballots get counted the DOJ will prosecute them with a federal crime.
Checkmate.
Looks like the organized protest industry is falling on hard times.
They can't seem to get it up, no matter how hard they try.
The dark money is gone along with the paid protestors. It just ain't the same without millions in funding.
Remember when everyone was protesting data centers? That money is gone too, so far fewer protestors.
Same with Palestine or whatever cause-of-the-day used to be protested.
When the money leaves, so do the protestors.
You'll see the Roadster flying car tech on the moon and Mars before long.
The existing pressurized gas system has 6x the range on the moon, maybe 30-40 sec of light as there's 1/6th the gravity. Fans, of course, are useless.
That could get you out of a few tight spots.
The main drawback of the tank is its weight fully loaded: 250 kg or 530 lbs here on Earth. But on the moon, a tank with six times the capacity wouldn't be hard.
Now we're talking 3-4 minutes of flight. Yes, you'd need to provide your own gas to compress, presumably by heating lunar rocks.
Very useful for exploring on the Moon.
Mars has 38% of the gravity of Earth, so a bigger/heavier tank could be supported as well.
I could easily see a lunar or martian vehicle built on Roadster tech that would drive normally but be able to jump obstacles and traverse canyons.
As Elon said, "where we're going, there are no roads".
I got an up-close look at the engineering INSIDE the new Tesla Roadster at yesterday’s unveil, and it’s pretty crazy. After watching this thing drive upside down, launch into the air, and do flips, I got to see the actual hardware that makes it all possible.
1/ The Electric drive unit is the heart of the Roadster’s electric powertrain, delivering power directly to the wheels.
2/ The battery pack has rows of cylindrical battery cells, with cooling systems and electrical connections running throughout. This is what powers the car before the rocket tech comes into play.
3/ The rocket thrusters has multiple cold-gas nozzles. Tesla says the four main rear thrusters blast compressed air faster than the speed of sound, producing more than 80,000 newtons of thrust.
4/ The SpaceX-inspired pressure tanks has two massive carbon-fiber-wrapped pressure vessels with a combined capacity of 600 liters, holding approximately 250 kg of compressed air. And the wild thing is the Roadster can recharge its own air supply, no special rocket fuel is required.
5/ An active rear wing has a variable-pitch wing that automatically adjusts downforce depending on what the car needs. It allows it to get more grip when cornering, less drag when going fast.
6/ The underbody engineering is the hardware underneath that brings everything together, including the systems that help the Roadster stick to the ground and launch into the air. And then there’s the fan suction system, which Tesla says can generate up to 2 g’s of suction force, enough to hold the car upside down.
Tesla and SpaceX have built an electric sports car with a battery, electric motors, rocket thrusters, high-pressure air tanks, active aerodynamics, and a suction system powerful enough to let it drive upside down. And yesterday, I watched it actually happen with my own eyes…
I’ve seen some incredible engineering from Tesla over the years, but getting this close to the hardware and then seeing what it can do was on a whole another level.
This is what happens when you bring together some of the best automotive and rocket engineers in the world and let them push the limits of what’s possible.
Great job Tesla! 👏
One of the reason many people irrationally dislike Elon is that he's so damn successful at so many things.
He's redefined existing industries and created entirely new ones.
And having the time of his life doing it.
Flying cars, rockets to Mars, robots, space AI ... he's the guy who is making the future real. He makes the impossible look easy.
Untold legions of young builders have now been inspired by his ethos.
History will be kind to him.
Elon Musk walked onto the Tesla Roadster stage so casually that people thought the unveiling would be boring.
Little did they know he was about to deliver one of the most INSANE unveils ever!
Then he said:
“You’re gonna see some crazy stuff tonight. We need an audience to tell if it’s not SI/AI.” 😂
ELON IS A F*CKING GENIUS!

























