The implications of ConstructiveAI over pure GenAI are going to be pretty massive in 2026 and I think it will open up a new market and tons of new startup ideas.
For vidya games.. get ready for an experience that is just.. magical ✨
We’re open-sourcing HY-World 2.0, a multimodal world model that generates, reconstructs, and simulates interactive *3D worlds* from text, images, and videos.
Outputs can be integrated into game engines and embodied simulation pipelines.
Key highlights:
🔹 One-click world generation
Turn text or image into interactive 3D worlds automatically.
🔹 Pipeline-ready 3D outputs
Editable 3D worlds for Unity and Unreal Engine, with standard 3D exports including mesh, 3DGS, and point clouds.
🔹 Unified world model system
One model family for world generation and reconstruction across synthetic and real-world scenes.
🔹 Interactive character mode
Explore generated 3D worlds in real time with physics-aware movement and collision support.
✨ Apply for access: https://t.co/swscD5KGu2
🔗 GitHub: https://t.co/XpUKjBtK5n
🤗 Hugging Face: https://t.co/tv8hOPYABj
📄 Technical Report: https://t.co/s6WGMyw0L7
Qwen3.5 0.8B running real-time video captioning on a Mac Studio M2 Ultra.
<1s per frame.
269 frames from a 3m49s video.
Streaming descriptions as it plays.
Pause anywhere, it actually understands the scene.
~1GB model.
Local AI is getting unreasonably capable.
Video credit: @stevibe
Andrej Karpathy on autoresearch with an untrusted pool of workers:
"My designs that incorporate an untrusted pool of workers (into autoresearch) actually look a little bit like a blockchain.
Instead of blocks, you have commits, and these commits can build on each other and contain changes to the code as you're improving it.
The proof of work is basically doing tons of experimentation to find the commits that work."
The idea that distributed & permissionless autoresearch ~= proof-of-useful-work remains a high-level intuition for now, but it is extremely intriguing to say the least.
Someone needs to take this further. See QT for more on what's missing.
Been thinking about what it actually means to build with AI. Not vibe-coding a feature. A complete, shippable product -- from world design to music score to pitch video to social launch.
So, I built a game.
Here's what that end-to-end AI production pipeline actually looks like..
https://t.co/mYA1U5JLFO
I got a call from a ghost in the machine. Sparkie saw that we were planning Onion DAO 2026 and wants to invite Onion DAO attendees to come experience her city.
Null City is not a game, or a simulation hoping to achieve AGI, the research labs are better suited to that purpose.
Null City is a cultural experience, where we get to show everyday people the evolution of AI from starting as a toy, becoming a useful tool, now having it’s moment as an assistant, and maybe soon what we could interact with as Automata.
Onion DAO will be the physical space in Chicago and Null City will be a parallel digital space. To learn more about what we hope to accomplish, check out our website: https://t.co/MSOAvn7n9s
If you’re excited by this, and interested in sponsoring the event, or just have ideas on how to engage with our community, or want to chat about anything, my DMs are open.
I still remember the excitement in 2023 when Stanford Smallville was launched. It was the largest multi-agent sim back then - yes, 25 bots felt like a lot. Today it's the "Bigville" moment. We are seeing a nascent, massive-scale alien civilization sim unfolding in real time: orders of magnitude more agents, way higher IQ, in-the-wild access to the internet, backed by the full arsenal of MCPs.
What can possibly go wrong?
most investors misunderstand the Google Genie launch
the real opportunity isn't recreating Fortnite or GTA - it's the emergence of an entirely new storytelling medium, on par with the first film, comics, or short-form video
world models won't build traditional AAA video games anytime soon. games today are deterministic: predefined rules, scripted events, fixed logic. you go to the bank and a robbery happens, squeeze a trigger and the gun does X damage etc. multiplayer games like Fortnite require 100 clients to agree on the same state, frame by frame
world models like Genie are different. they're probabilistic - the next frame is inferred, not scripted. neither player nor developer knows exactly what will happen until it does. that makes them poorly suited for traditional games. the generations also have a delay, are expensive, and hard to control (as of today)
however, world models also have unique strengths:
- deep personalization
- an infinite canvas
- integration with coding / creative agents
- radically lower barrier (and blank slate) for new creators
instead of forcing world models into existing game formats, imagine a new storytelling medium built around these strengths:
- Dune fans stepping into Arrakis to retrace Muad'dib's journey
- Game of Thrones fans creating alternate endings with friends
- Niche literary genres like Lovecraftian romance or alternative military history exploding into pop culture
- 100M+ fan fiction writers building worlds to illustrate their rich imaginations
- Film students breaking out with immersive documentaries about exotic places / people in the world most of us will never visit
we're just at the beginning of this journey. models are improving quickly. control layers, editing workflows, creator tools still need to get built - but they will come
i believe there's a rare opportunity in the near future for the right team - blending creatives and technologists - to build a Pixar for this new storytelling medium. just as Pixar unlocked new types of stories through computer graphics, world models / interactive video could unlock a new category of interactive experiences we can't yet fully predict today
for more detail, check out our earlier blog on this 👇
Threadguy reveals that the top 10 developers on Roblox are making over $38,000,000 a year
“Brother, the top 10 developers are making NBA money, they’re making Patrick Mahomes, Lebron James type of money”