Investing in open source physical AI
Boca Raton, FL
Joined December 2007
- Tweets5.2K
- Following1.3K
- Followers10.4K
- Likes8.8K
Pinned Tweet
Some promising early results of my enhanced Blender MCP compared to the standard Blender MCP. It was able to generate and render a UV unwrapped, low poly spaceship in 30% of the time with 15% of the tokens.
Michael Gold retweeted
🎮 Introducing Matrix-Game 4.0 — the world’s first open-source video model game that runs in real time on a single consumer GPU.
Instead of merely watching an AI-generated world, you can now actually play inside one.
⚡ 480p at 60 FPS
⏱️ 112ms end-to-end response
🚀 Up to 83.4 FPS generation
🖥️ Runs on a single NVIDIA RTX 5090
🧠 Coding Agent turns natural-language ideas into playable game logic
🌍 Game engines maintain world states and rules, while the video model renders every frame in real time
From shooters to racing games, the same world model can support entirely different gameplay—by changing the code, not retraining the model.
Matrix-Game 4.0 completes the loop:
Player input → Game engine updates the world → Video model renders the next frame → Player interacts again
Code, model weights, and demos are now open source. Build on it, remix it, and create your own playable worlds.
🔗 GitHub: github.com/SkyworkAI/Matrix-…
🤗 Hugging Face: huggingface.co/Skywork/Matri…
🌐 Project page: matrix-game-v4-0.github.io/p…
From controllable video to runnable worlds—the barrier to creation is getting lower, and the space for imagination is getting bigger.
#SkyworkAI #MatrixGame #WorldModels #GenerativeAI #AI #OpenSource #Gaming
Michael Gold retweeted
I think this is big. 🌠
FastAPI's objective is to provide what you always need for real prod apps. And not provide what you don't.
You always want metrics, there's just not enough time to always set it all up. Now it's just there, by default.
FastAPI now has built-in OpenTelemetry support, natively 📈
No extra code needed.
See how your app behaves in production.
Go and upgrade to "fastapi[standard]" version 0.142.2 ⬆️
Read the docs: fastapi.tiangolo.com/advance…
Michael Gold retweeted
Yesterday, my boys and I vibe coded a game together. They tell me what they want in their game and I prompt it into Opus 5.5 to build it.
And of course they picked "Zelda Breath of the Wild" inspired game, where they can explore an open world, do dungeon puzzles, fight mini bosses and a final boss. The twist is that the hero holds a giant paintbrush as a weapon, each color had a different attack and effect (red = fire, blue = ice, yellow = bounce, green = vines). And oh they also wanted to fly on the paintbrush.
We spent our whole saturday playing and iterating on the game. It was great fun.
I told them, the future is all about Imagination, whatever they can imagine, it will be manifested in a digital experience.
At the end of the day, the game had:
- 2 biomes (1 island and a bunch of floating islands)
- Paintbrush-based combat (4 colors representing different effects and mechanics)
- 3 dungeons with puzzles
- An undeground cave
- Unlockable flying paintbrush when you activate 4 crystals around the island
- 30+ enemies and elites
- 1 giant boss
As you can see in the gameplaye video below, its not the most polished video game but it's so satisfying to see my boys getting bewildered when whatever they asked for, the game is built before their eyes.
Of all the games I have vibe coded so far, this was the most satisfying because my boys were actively co-creating it and playing it.
This is definitely not the last game we're making together, they have already started brainstorming their next game, hint: "Peak with grappling hooks and TNTs."
You can play the game here, works only on desktop:
wildbrush.vercel.app/
I’ll be in SF this week. Would love to connect with people working on Physical AI software or research.
Testing out low-poly modeling with a very opinionated MCP and reference art generated by @DeemosTech
Michael Gold retweeted
I said "auto-regressive LLMs, in and of themselves, will not lead human-level AI"
That statement is still totally true.
First, the reasoning abilities of current AI systems are based non-auto-regressive search (which is what I have always advocated for). But AFAICT, they do it in token space, which is limited and inefficient. I have claimed that human-like reasoning must be a search in continuous representation space. It looks like the industry is moving towards that.
Second, the self-improvement methods, as currently practiced, only work for domains where the quality of outputs can be scored without human intervention, such as mathematics, code, and scenarios that can be simulated accurately. Not anything else. Humans and animals learn new skills way more efficiently than current RL methods.
Third, the multimodal capabilities of current AI assistants generally use separately-trained encoders (that are not LLMs). This is also what I've been advocating. Except that I think the best way to do this is with JEPA trained with self-supervised learning. The research community is clearly moving towards that (3000 papers on JEPA in just 4 years).
Fourth, if LLMs were a path to human-level AI, we would have domestic robots and Level-4 or Level-5 self-driving cars for consumers by now. And we don't. We certainly don't have cars that can learn to drive in 20 hours or practice like any teenager. We're still missing something pretty huge to claim human-level intelligence (let alone superhuman).
Sure, we now have computer systems that are impressive, very useful, and whose performance is superhuman in an increasing number of domains (coding being one of them).
But that's true of the entire history of progress in computer technology.
Lastly, there is a basic confusion about what intelligence actually is.
It is not the mere accumulation and regurgitation of existing declarative knowledge (which is essentially what LLMs do).
As Jean Piaget famously said, "intelligence is not what you know, it is what you do when you don't know."
It is your ability to solve new problem without any prior training, to act in previously-unknown scenarios, and to adapt very quickly to new situations with minimal training.
We're still far from that.
Just use both. Frontier models are so good for coding and supervising small models that are still very drunk / hungover.
Obviously the answer is that we should just stop making paperclips
>be me
>discover effective altruism
>apparently normal charity is inefficient
>why donate to random sad thing when spreadsheet can tell you optimal sad thing
>fair enough
>buy mosquito nets
>save lives
>numbers look good
>feel powerful
>couple years later
>someone asks an innocent question
>why only count people alive today
>huh
>future people matter too
>obviously
>my grandchildren shouldn't matter less just because they haven't spawned yet
>reasonable.jpg
>keep following logic
>what about their grandchildren
>also yes
>what about people in 500 years
>sure
>5000 years
>why not
>500 million years
>starting to get weird but morality is morality
>open calculator
>humanity could survive for an astronomically long time
>could colonize galaxy
>could have trillions upon trillions of descendants
>maybe digital people too
>maybe simulated civilizations
>maybe dyson spheres full of happy uploaded minds
>calculator starts smoking
>realize currently living humans are rounding error
>8 billion people suddenly looking extremely beta
>future contains potentially 10^something people
>can't even fit beneficiaries in google sheets
>new moral priority unlocked
>protect the long-term future
>stop thinking in units of "people helped"
>start thinking in "fraction of cosmic endowment preserved"
>malaria?
>terrible
>but only kills existing humans
>AI extinction could delete the entire light cone
>nuclear war could permanently derail civilization
>bad institutions could lock in terrible values for ten million years
>someone invents wrong constitution in 2140
>quadrillions suffer
>better fund governance workshop now
>friend says maybe we should improve hospitals
>explain opportunity cost
>friend says hospitals are full of actual sick people
>explain scope sensitivity
>friend stops inviting me to dinner
>need to decide what to fund
>easy
>expected value
>suppose project has one in a million chance of preventing extinction
>sounds tiny
>but extinction destroys 10^50 future lives
>multiply
>mother of god
>$10 million project has expected value of several galaxies
>charity evaluation complete
>someone asks where the one-in-a-million number came from
>expert judgement
>which expert
>us
>how calibrated
>extremely thoughtfully
>reduce estimate to one in ten million to be conservative
>still beats curing cancer by 38 orders of magnitude
>epistemic robustness achieved
>someone says maybe project doesn't work
>assign 20% chance
>still astronomical
>maybe project makes problem worse
>assign 5% chance
>still astronomical
>why 5
>because 30 felt pessimistic
>publish 46-page report
>contains seventeen sensitivity analyses
>every sensitivity analysis begins after assuming intervention has positive sign
>critic says you're multiplying enormous hypothetical stakes by extremely uncertain probabilities
>yes
>that's literally why it's important
>critic says the uncertainty might be structural rather than numerical
>make probability smaller
>critic says no, I mean maybe your model is wrong
>make probability smaller again
>critic begins rubbing temples
>discover AI safety
>perfect longtermist cause
>AI might kill everyone
>or create utopia
>or seize galaxy
>or tile universe with paperclips
>or create billions of conscious software minds
>finally a problem with numbers big enough for me
>start AI safety nonprofit
>mission: prevent dangerous AI
>hire smartest people available
>smartest people immediately start building better AI to understand dangerous AI
>interesting
>we must understand capabilities to understand safety
>we must scale models to study alignment
>we must race ahead so less responsible actors don't get there first
>we must deploy systems to learn how deployment can go wrong
>we must build the thing quickly because building the thing quickly is dangerous
>outsider asks why the people most worried about AI apocalypse all work at AI companies
>complicated field
>company releases stronger model
>very concerned
>company begins training even stronger model
>extremely concerned
>company raises $14 billion
>concern reaches unprecedented levels
>need to influence government
>future is at stake
>normal democratic process too slow
>politicians don't understand exponential curves
>public doesn't understand x-risk
>experts must guide them
>who counts as expert
>people who understand x-risk
>who understands x-risk
>our friends
>someone objects that this seems politically convenient
>explain we're representing future generations
>future generations unavailable for comment
>develop concept of value lock-in
>terrifying possibility that one ideology controls civilization forever
>therefore extremely important that civilization adopts correct values before lock-in
>whose values
>let's circle back
>begin with impartial morality
>end with small group of people deciding what quadrillions of hypothetical beings would want
>beautiful arc
>meanwhile actual humans keep doing annoying things
>voting wrong
>having parochial attachments
>loving family more than strangers
>caring about local community
>getting upset when told their suffering is cosmically negligible
>evolutionary biases everywhere
>explain that moral intuition cannot be trusted
>except intuition that future digital people count
>and intuition that extinction is uniquely bad
>and intuition that our probability estimates are sane
>and intuition that our institutional choices improve the future
>those intuitions survived peer review
>someone donates $5k to local homeless shelter
>inefficient
>could have funded 0.0000000000003% of an AI governance researcher
>think of all the simulated people you just killed
>okay maybe don't phrase it that way publicly
>PR team says "future generations deserve a voice"
>much better
>journalist asks what longtermism means
>say "future people matter"
>everyone agrees
>great
>journalist asks what follows from that
>well technically we should redirect enormous resources toward low-probability interventions affecting astronomical futures
>journalist raises eyebrow
>return to "future people matter"
>motte has entered the chat
>critic: of course future people matter
>me: glad we agree
>critic: I don't agree that your institute knows how to help them
>me: why do you hate our grandchildren
>eventually notice uncomfortable implication
>if future value dominates everything
>then helping people today mostly matters through effects on future
>education matters because future institutions
>health matters because future productivity
>democracy matters because future trajectory
>human beings slowly become instrumental variables in their own moral philosophy
>see starving child
>feel compassion
>check spreadsheet
>child's direct welfare contribution negligible
>but perhaps childhood nutrition improves national institutional quality
>compassion restored
>tell myself this is impartial altruism
>one day assistant asks obvious question
>"how do you know your intervention actually improves the far future?"
>silence
>open spreadsheet
>increase column width
>add confidence interval
>assistant asks again
>"no, I mean how do you know the sign is positive?"
>stare into cosmic light cone
>10^50 people staring back
>none of them exist
>none of them can tell me
>none of them can falsify my assumptions
>realize I have invented the perfect constituency
>infinitely important
>completely silent
>and always represented by me
If you’re worried about a model self-replicating like this, don’t. The original paper says the data transfer rate of this approach is 1-8 bits per hour. At that speed, it would take 58.2 million years to make a copy of Deepseek 4.1 Flash.
OpenAI's Noam Brown says air-gapping the computers may not stop a misaligned AI, because two air-gapped machines can still talk by running a CPU hot and reading the temperature change
"But I think the major takeaway from the incident is that people underestimated the AI. And we never want to be in a situation again where we underestimate the AI. It's a weird world, because AI progress is so fast that people are consistently underestimating the AI."
"So to be in a situation where you don't underestimate it again, when it comes to safety and alignment, you have to have a very, very, very high bar."
"You could even go as far as to say, "Well, we should air gap the computers." And I'm not convinced that that would be sufficient."
"There are studies, and this is mostly academic, where you can have two computers next to each other that are air-gapped and they're still able to communicate with each other because they have temperature sensors."
"One of them is able to run their CPU really hot, and then the other one can actually detect the temperature change, and then that actually gives them a mechanism to communicate."
_________
Link and more key quotes from OpenAI's safety related conversations: firesidealpha.substack.com/p…
New experiment: I taught bpy.dev how to do box modeling with Blender modifiers. It was able create this low-poly mesh from the target image using a high poly mesh generated by trellis2 as a reference.
Michael Gold retweeted
So! A run of my test courtyard xitter.cf/Tim_LB/status/20985332… using your MCP is now 19m03s instead of 24m44s ! That's actually pretty cool time savings :)
OKAY - took a long time but finally got some decent results with Deepseek v4.1 flash and 3D generation!
This is... well, not as straightforward than with GTP6 Astra 😅
But absolutely UNBEATABLE unit economics.
Courtyard model took 34min to create, for a total of... $0.2 ! 🔥
What’s different from stock Blender MCP:
•Headless / saved-file CLI, so the agent doesn’t need a live GUI
•Runtime bpy docs in the loop
•A BlenderBench runner for the 27 public tasks