imatrix

it's not about the punchline, it's about the fact that jokes are an extremely complicated thing. The reason? It's a highly abstract structure of information, not unlike low-level speech (how words are supposed to be put together in different languages), code (different programming languages have different rules for commands and sequences).

Even the best LLm's are still at the level of children in this abstraction, so they can't make "quality" jokes. They also suffer from not having a unique personality and being the average of everything. Until this is addressed, don't expect great jokes to come out of AI. It's almost the most challenging discipline. I wouldn't be afraid to make the comparison that the "joke" is the real Turing test.

Same on iOS and mobile Safari - not working.

Its not working.

Loading model: waiting WebGPU error: WebGPU is NOT supported on this browser.

Im on macOS and Safari.

Why there is no comment?

This is not correct. It is like VRAM, not like normal PC RAM.

Lowest latency of DDR5-6400 on normal PC starting at 60ns+

Lowest latency of VRAM on GeForce RTX 4090 starting at 14 ns

Lowest latency of Apple M1 Memory starting at 5 ns, its more like L3 cache

And on Apple M chip, this ultrafast memory is available for CPU, GPU and NPU.

https://www.anandtech.com/show/17024/apple-m1-max-performanc... https://chipsandcheese.com/p/microbenchmarking-nvidias-rtx-4...

I think that if investors see this, that you only need few millions to be almost number one, they will not give them billions so easily. Deepseek changing the narrative of spending money in AI tech.

Right now, they are standing on the backs of giants, but in the future...it could be opposite.

What they achieved with so few resources, it’s WOW and WTF. Maybe it will cause end of the open source models? Because it’s always about money in the end.