Sabitlenmiş Tweet
The world is going to keep getting better because people are going to keep making it better.
Techno-Industrial Deep Dives:
English
Packy McCormick
46.1K posts
Pumped to be expanding to IL as state #2!
IL is birthplace of my wife and parents, plus my home for 12 years, and most importantly a forward-thinking electricity market.
@basepowerco is just getting started
Travis Kavulla@TKavulla
Delighted to announce @basepowerco has received its first state license to operate as an electricity retailer outside Texas. We're heading to Illinois! Thank you @ILCommerceComm for approving our application! Let me explain why we're heading to the Land of Lincoln🧵
English
@EricWollberg “We are living in a computer-programmed reality, and the only clue we have to it is when some variable is changed2, and some alteration in our reality occurs.”
English
Is it like something to be a world model?
Packy McCormick@packyM
There is a tremendous amount of progress happening in World Models. Multiple labs have raised more than $1B. WMs were the star of GTC. They are a real path to embodied AI. So @PimDeWitte & I wrote a comprehensive 19k word overview of World Models. notboring.co/p/world-models
English
Best thing I’ve read in a while
Packy McCormick@packyM
There is a tremendous amount of progress happening in World Models. Multiple labs have raised more than $1B. WMs were the star of GTC. They are a real path to embodied AI. So @PimDeWitte & I wrote a comprehensive 19k word overview of World Models. notboring.co/p/world-models
English
Language models predict a sentence whereas world models predict the future
Packy McCormick@packyM
There is a tremendous amount of progress happening in World Models. Multiple labs have raised more than $1B. WMs were the star of GTC. They are a real path to embodied AI. So @PimDeWitte & I wrote a comprehensive 19k word overview of World Models. notboring.co/p/world-models
English
There is a tremendous amount of progress happening in World Models.
Multiple labs have raised more than $1B. WMs were the star of GTC. They are a real path to embodied AI.
So @PimDeWitte & I wrote a comprehensive 19k word overview of World Models.
notboring.co/p/world-models
English
@jmwass lol that idiot lost more money today than i'll make in my lifetime
English
@packyM He is so freaking stupid, he lost almost $7 billion.
English
I hadn't checked out the Real-Time Billionaires List in a while.
Elon moneymaxxxxxing and networthmogging.
English
This is very cool.
In the World Models piece today, Pim and I wrote that Pi's VLAs are a pragmatic approach to embodied AI and that the company seems to be making a very strategic bet. They keep unhobbling VLAs.
Physical Intelligence@physical_int
We developed an RL method for fine-tuning our models for precise tasks in just a few hours or even minutes. Instead of training the whole model, we add an “RL token” output to π-0.6, our latest model, which is used by a tiny actor and critic to learn quickly with RL.
English
@KaranKunjur @nunzi46 @K2SpaceCo Mike if K2 is now too vertically integrated and you want to sell me your shares, you know where to find me
English
English
Can’t understate how insane this is
@K2SpaceCo team is bananas
And their ambition is beyond inspiring
Karan Kunjur@KaranKunjur
We have shipped our 20kW satellite - Gravitas - to the launch site. Given the supply chain to operate at this power regime doesn’t exist, we had to build 85% of the satellite in-house. This includes building our own large solar arrays, high power propulsion system, large batteries, large reaction wheels and much more. This launch will represent the first time all of these systems are test on orbit together. Internally at @K2SpaceCo, we’ve thought about a few levels of success for this mission - we expect mission success to fall somewhere along this spectrum: - Tier 1 (Baseline mission success): Deploy solar arrays, establish comms, operate the satellite —> we’ve now got an operational 20kW satellite on orbit - Tier 2: Power on the payloads, activate the 20kW propulsion system —> we’re completing payload missions and have fired the highest power hall thruster ever flown on orbit - Tier 3: Orbit raise the satellite, test performance in high radiation environments (like 2,000km) —> we’ve collected massive amounts of data on the performance of the platform in very very difficult environments More than anything, Gravitas represents the start of an iterative journey, where we will take the data we receive from this first satellite and incorporate it into the next wave of satellites launching next year. We’re excited to start this journey, we’ll report back as we get more data. Thanks to Tim for covering our story on TechCrunch techcrunch.com/2026/03/19/k2-…
English
Thorough read! But a good one so far.
World models can sorta boil down to something simple: can you predict how the world will change in response to an action? (or other condition)
Packy McCormick@packyM
There is a tremendous amount of progress happening in World Models. Multiple labs have raised more than $1B. WMs were the star of GTC. They are a real path to embodied AI. So @PimDeWitte & I wrote a comprehensive 19k word overview of World Models. notboring.co/p/world-models
English
after the scale ai and stripe pieces @packyM got the holy trinity with this one on general intuition - so much signal to noise again
Packy McCormick@packyM
There is a tremendous amount of progress happening in World Models. Multiple labs have raised more than $1B. WMs were the star of GTC. They are a real path to embodied AI. So @PimDeWitte & I wrote a comprehensive 19k word overview of World Models. notboring.co/p/world-models
English
I think I have the coolest job in the world. I get to borrow geniuses’ brains to learn about things I’m fascinated by and then share what I’ve learned.
Pim de Witte@PimDeWitte
Packy and I spent the past month unpacking world models from first principles. This piece is the result of that exploration. We go into why, what, how, and look out into the future on the implications of our work.
English
Packy McCormick retweetledi
Everyone says they’re building a world model. Very few actually are.
Most AI learns to see the world.
A world model learns to predict what happens in it — specifically, what happens when people do things. That’s a different problem.
Seeing is passive. Consequences require understanding cause and effect Any AI can learn to read a scene. A world model learns INTUITION about what changes it. @gen_intuition
Pim de Witte@PimDeWitte
Packy and I spent the past month unpacking world models from first principles. This piece is the result of that exploration. We go into why, what, how, and look out into the future on the implications of our work.
English
@nunzi46 @K2SpaceCo necessity is the mother of vertical integration
English
@packyM @K2SpaceCo I was able to talk them out of doing silicon ingots
...for now.
English
@bilawalsidhu Thank you Bilawal. “Ill tailored tapestry” is great.
English
If you've been curious about world models, read this. Got an early preview of the blog and it does a thorough job of unpacking the ill tailored tapestry of world model initiatives.
Packy McCormick@packyM
There is a tremendous amount of progress happening in World Models. Multiple labs have raised more than $1B. WMs were the star of GTC. They are a real path to embodied AI. So @PimDeWitte & I wrote a comprehensive 19k word overview of World Models. notboring.co/p/world-models
English
@packyM @PimDeWitte This is a really impressive piece of writing. Major kudos.
English