x-Apple / x-Rivian (tech team founder) Disruption happens. Optimism ahead. 🚀🇺🇸 My Articles nitter.net/pbeisel/status/1848807…
Silicon Valley
- Tweets 31,532
- Following 815
- Followers 34,324
- Likes 51,254
Agentic isn't the cleavering point.
And I say "cleavering" purposely because our politicians are really that stupid about technology. They are like pretend heart surgeons applying a cleaver as their primary tool.
Agentic means nothing more than something that can operate asynchronously to their human directors. And, yes, to cure cancer you need AI to operate agentically. Duh.
That said I also agree with @Chamath's read on this. @BarackObama is playing a power game here, like he does always.
Maybe talking ignorantly about something will stop certain politicians from gaining power over it. Though it never stopped them before.
Go away @BarackObama, the world needs to progress without "old men clinging to power".
Chamath on Obama’s AI Warning: Follow the Money
President Obama last week:
“If we are thinking about AI just in terms of how do we cure cancer or get better energy, you can do that without having agentic AI.
That's a misalignment between what our society needs and the commercial imperatives that these companies are facing.”
@Chamath:
“This is a very important moment for a very simple reason, which is we, as in the world, are about to endow three, four, five, six companies with about $10T of wealth.
And what Obama knows very well is that most of those companies are overwhelmingly left-leaning. And what he also knows is that there is a huge portion of that money that will then get put into philanthropic and charitable causes that then he, and the people around him, will be beneficiaries of.
That is the truth. We already know this because we know that some of these frontier corporations actually ask you to stand up DAFs and have a portion of your stock that you're willing to pledge.
So this money is going to go to things other than consumption or savings. And so it's going to go into PACs, it's going to go into political movements, and (Democrats) stand to disproportionately benefit.
So this has nothing to do with prosperity. This is a very simple political calculus.
If you freeze frame the economy the way it is today, a handful of organizations that will disproportionately be able to affect the Democrats will win. They will capture the lion's share of the economic gains, and then they will help the Democrats win power.
That's all this is.”
9
7
71
6,688
The Mandate of Heaven (Tianming) is an ancient Chinese political and philosophical doctrine that a ruler's right to govern is conditional on their virtue and ability to serve the people well.
3
4
116
10,837
Living in a van down by the river has its benefits.
~2.5GW compute by 2027.
When I read Elon’s words, I keep thinking about the Airstream parked somewhere on the Colossus property in Memphis.
It’s pretty clear all signs point to SpaceXAI winning the AI race.
Grok will converge to within spitting distance of the leading models, and likely move into 1st place over time.
It’s not just about the model. It’s equally about running it efficiently, at high performance, cheaply, and at massive scale.
When companies like Anthropic are getting their compute from SpaceXAI, they’re effectively buying a critical input from their competitor.
That makes it very difficult to compete on price.
Only one of these companies has a vertically integrated AI stack — from compute to model to deployment.
2
15
192
8,196
When I read Elon’s words, I keep thinking about the Airstream parked somewhere on the Colossus property in Memphis.
It’s pretty clear all signs point to SpaceXAI winning the AI race.
Grok will converge to within spitting distance of the leading models, and likely move into 1st place over time.
It’s not just about the model. It’s equally about running it efficiently, at high performance, cheaply, and at massive scale.
When companies like Anthropic are getting their compute from SpaceXAI, they’re effectively buying a critical input from their competitor.
That makes it very difficult to compete on price.
Only one of these companies has a vertically integrated AI stack — from compute to model to deployment.
1. We will keep accelerating. Our AI efforts are only 3 years old, vs 6 and 10 years old for Anthropic and OpenAI. If our second derivative remains strong, SpaceX will reach pole position in about 6 months.
2. Once you far exceed the caliber of intelligence needed for a class of tasks, additional intelligence is pointless. You don’t need (and it would be cruel to put) Newton-level intelligence in your toaster!
3. Hardware is hard. Bringing massive compute online rapidly is incredibly difficult. SpaceX has demonstrated exceptional ability in this regard and will only get better.
8
30
263
20,229
"A Bentley with no grill would be like taking the nose off someone's face."
Noses are for breathing. An EV does not need to breath.
The Bentley Torcal EV's grille sits behind a piece of polycarbonate. Behind it are dozens of illuminated diamonds designed to look like a piece of jewelry.
Bentley on why their EV even has a grille: "A Bentley with no grill would be like taking the nose off someone's face."
7
33
6,549
Apple and its AI Future
Apple is plowing ahead into a particular AI future based on its past strengths. It makes sense.
But I don't think it will be the AI future.
Apple is betting on powerful intelligence at the edge— intelligence resident in the device, tightly integrated with Apple's silicon, OS and app ecosystem.
That's a very Apple-like future.
But I think powerful edge intelligence is ultimately the domain of agency: devices that can do.
Namely, robots.
And Apple further believes in an app-centric future.
That's where I think the thesis breaks.
Intelligence does not need to be resident in a device if the device's primary purpose is intelligence. Nor are “apps” required for intelligence to be useful.
The future device, as I've written about below, is a device that can apply intelligence by doing.
Some of that intelligence will necessarily be local. Perception, control, latency, autonomy.
But much of the intelligence can come from powerful compute fabrics in the cloud (e.g., Starmind).
The device doesn't need to contain the raw intelligence.
It needs access to it.
And it doesn't need an app to invoke it.
It makes all the sense in the world for Apple to pursue the future it is pursuing. I just think it's the wrong future.
9
7
96
9,656
Something insanely cool getting a software update notice for the Tesla on Starlink (United) at 35,000 feet. Like living in the future I guess.
$SPCX $TSLA
Starlink on United is Internet
Compare Starlink vs. ViaSat.
Starlink delivers average download speeds of ~90 Mbps while ViaSat is ~13 Mbps.
Starlink offers latency in the ~30 ms range while ViaSat can be around 770 ms.
That latency difference is the killer.
ViaSat, Anuvu, Panasonic and the other traditional satellite systems rely heavily on geostationary satellites roughly 22,000 miles above Earth. You can optimize the network around that, but you can't optimize away the distance.
It's physics.
At 35,000 feet, a Starlink satellite is hundreds of miles away, not tens of thousands. The speed of light is an unforgiving constant.
That's why Starlink isn't just better Wi-Fi on an airplane.
It's a fundamentally different network.
Measurements from my network testing app NetP. Available at netp-app.com/. Try it on your next flight.
First screen shot, somewhere over Utah last night.
4
7
90
2,969
Starlink on United is Internet
Compare Starlink vs. ViaSat.
Starlink delivers average download speeds of ~90 Mbps while ViaSat is ~13 Mbps.
Starlink offers latency in the ~30 ms range while ViaSat can be around 770 ms.
That latency difference is the killer.
ViaSat, Anuvu, Panasonic and the other traditional satellite systems rely heavily on geostationary satellites roughly 22,000 miles above Earth. You can optimize the network around that, but you can't optimize away the distance.
It's physics.
At 35,000 feet, a Starlink satellite is hundreds of miles away, not tens of thousands. The speed of light is an unforgiving constant.
That's why Starlink isn't just better Wi-Fi on an airplane.
It's a fundamentally different network.
Measurements from my network testing app NetP. Available at netp-app.com/. Try it on your next flight.
First screen shot, somewhere over Utah last night.
3
12
102
6,764
Was asking Grok Bot to look into an issue with my NetP app on a United flight with Starlink.
Oh @bot, the boss might want to have a talk with you about characterizing Starlink on United as "flaky".
It has not been flaky for me, btw. Its been awesome! @Starlink
netp-app.com/
4
4
25
3,549
SpaceXAI’s move: raise Grok quality and cut token prices, Cybercab-style.
They own the compute.
Competitors rent large slices of it for $1B+/month.
That leaves Grok with a structurally lower cost per token while the model keeps improving.
Competitors then have to match both capability and a falling price from a cost base they don’t control in the same way.
That’s a tough spot.
Now maybe you understand why Elon is living in an Airstream.
Grok 4.7 works longer on difficult tasks, checks its work more carefully, and comes with our strongest safeguards to date.
7
17
159
8,428