ZephyrXero

Allowing LLMs web and terminal access was one of the dumbest moves humanity ever made

To use their own words against them…

We need to see pain in the economy. We need to remind investors that the workers have the power, not the other way around.

Sounds like more people who have bought the hype, despite their pedigree. As while the business world is quickly learning AI can not replace all their staff like they hoped. LLMs are never going to become AGI, and the ML models that might actually have a shot are still in their infancy.

Eric Schmidt being amongst the signers makes me extra skeptical

Do not want

Because the average Instagram user has stopped using Instagram like Instagram, they just watch Reels now

It is, this is from the same developer

Another cool thing about Loops. They are working with the lead developer of DiVine to allow them to federate

If the model dumps over to system ram it gets super slow, you ideally want it to fit completely in your VRAM

I got Qwen 3.5:9b running on my 8GB GPU the other day, and it still has some room left over

There are efforts there. The new Deepseek 4 compresses a lot of its knowledge using something they call engrams. But it’s unfortunately still too big for a consumer GPU.

Gemma 4 is small enough to run on your cellphone.

If your GPU has at least 8GB there are a lot of options for self hosting your own local models

Hmm, fertility dropped “from June 2007 through February 2011”. And they attribute it to the iPhone. But what else happened during that period? Oh right a big financial crisis.

And there’s already data out there that shows an economic decline correlates to a drop in births

I’m seeing reports of a 20-30% improvement in bitrate usage. But the encoders are also still early at this point