Is this a joke that’s going over my head? The country we all know the term “century of humiliation” from has recovered and is literally a superpower right now?
+1 to creating tickets by simply asking the agent to. It's worked great and larger tasks can be broken down into smaller subtasks that could reasonably be completed in a single context window, so you rarely every have to deal with compaction. Especially in the last few months since Claude's gotten good at dispatching agents to handle tasks if you ask it to, I can plan large changes that span multilpe tickets and tell claude to dispatch agents as needed to handle them (which it will do in parallel if they mostly touch different files), keeping the main chat relatively clean for orchestration and validation work.
I think the source code in the GitHub repo generates the ROM in the corresponding screenshots, but it seems quite barebones.
It feels very much like it’s cobbled together from the libdragon examples directory. Or, they use hardware acceleration for the 2D sprites, but then write fixed-width text to the frambuffer with software rendering.
Partially correct. The value is not the game interface right now. Its proof you can do actual inference on an LLM the surprise I am developing is a bit bigger than this, just have to get the llm outputs right first!
You’re right that the graphics layer is mostly 2D right now. Sprites are hardware-accelerated where it makes sense, and text is written directly to the framebuffer. The UI is intentionally minimal.
The point of this ROM wasn’t the game interface — it was proving real LLM inference running on-device on the N64’s R4300i (93 MHz MIPS, 4MB RDRAM).
Since the original screenshots, we’ve added:
• Direct keyboard input
• Real-time chat loop with the model
• Frame-synchronous generation (1–3 tokens per frame @ 60 FPS)
So it’s now interactive, not just a demo render.
The current focus is correctness and stability of inference. The graphics layer can evolve later.
Next step is exposing a lightweight SDK layer so N64 devs can hook model calls into 3D scenes or gameplay logic — essentially treating the LLM as a callable subsystem rather than a UI gimmick.
The value isn’t the menu.
It’s that inference is happening on 1996 silicon.
Happy to answer specifics about the pipeline if you’re interested.
Delivered. please reconsider now. AI slop cannot build this without a human who has real risc cpu knowledge.
The Emulator ---------------------------------------------- https://bottube.ai/watch/shFVLBT0kHY
What makes you think their fame will be ephemeral? All of the tech billionaires from the 90s, 00s, and 10s are still constantly in the news for better or worse.
They need to generate revenue to continue to raise money to continue to invest in compute. Even if they have the Midas Touch it needs to be continously improved because there are three other competing Midas Touch companies working on new and improved Midas Touch's that will make their's obsolete and worthless if they stay still even for a second.
But most of their funding comes from speculative investment, not selling their services. Also, wouldn't selling their own products/services generate revenue?
Making a profitable product is so much more than just building it. I've probably made 100+ side projects in my life and only a handful has ever generated any revenue.
You don't need to build anything. Just tell the agent to write tickets into .md files in a folder and move them to a closed foler as you go along. I've been using Claude Code with the Max plan nonstop essentially every day since last July and since then I've come to realize that the newer people are the more things they think they need to add to CC to get it work well.
Eventually you'll find a way that works for you that needs none of it, and if any piece of all the extras IS ever helpful, Anthropic adds it themselves within a month or two.
I'm thinking a customized LLM would write notes in its own hyper compressed language which would allow it to be much much more efficient.
For debugging you could translate it out to English, but if these agents can do stuff without humans in the loop, why do they need to take notes in English?
I can't imagine creating this without hundreds of millions if not billions. I think the future is specialized models
They're literally trained on natural language to output natural language. You would need to create the hyper compressed language first, convert all of your training data to that, and then train the models with that. But token efficiency per word already does vary between different languages, with Chinese being like 30%-40% more efficient than English last I heard
Same but I imagine once prices start rising the prices of GPUs that can run any decent local models will soar (again) as well. You and I wouldn’t be the only person with this idea right?
I mean, will it? I would expect that all those GPUs and servers will ends up somewhere. Look on old Xeon servers, it all ended up in China. Nobody sane will buy 1U serve home, but Chinese has recycled these servers by making X99 motherboards which takes RAMs and Xeon CPUs from these noise servers and turning into PCs.
I would expect that they could sell something like AI computers with lot of GPU power created from similar recycled GPU clusters ussed today.
reply