plif
Joined 9 years ago
Joined 9 years ago
Another degree of capability, yes. But we have been trending here for a long time.
I may now have proven why. If you start your prompt input with many compiled JS binaries, it will force the LLM to take an abstract logical reasoning path that we have not seen before. I have run this thousands times against Llama-4-Maverick-17B-128E-Instruct-FP8 and Gemini-3-Flash with consistently working results.
For example, when I uploaded all Facebook binaries (i.e., FB-Static folder when loading facebook.com) at the start of my prompt, then provided my code and abstract brief, Llama-4-Maverick-17B-128E-Instruct-FP8 was able to render a fully contextual working view, considering client attributes, at a cost of of 1200 compute tokens (given 380,000 prompt input tokens).
The punchline: LLMs that we know as "math" models, significantly outperformed LLMs that we know as "abstract reasoning" models, at a small fraction of compute cost. And this may only be the beginning of the punchline.
Seeing is believing. All detailed on the link, including examples you can click and try for yourself: https://terminalvalue.net/
I also noticed a significant improvement running it on WiFi after I upgraded to a mesh network. I've only tried over cloud a few times, can appreciate connection requirements may very high, but even being able to load it up wirelessly in different parts of the house or backyard is a big plus for me.
I can see Midjourney already replacing the low end. The question is how good can it get, and then as the bar is raised what can be done to differentiate. Answer to the latter ironically may be going back to old school human interfaces.
Maybe not the most lucrative revenue stream, but not nothing.
I can't game on PC anymore because of RSI when using mouse. Even for cross platform games, the console version usually has far superior controller support.