> None of the big American companies want to release a leading edge model as open source because this would drive the price of inference to $0
Meta/Llama: "What am I, chopped liver?"
I thought the thing keeping inference above $0 was the hardware, and even if that were free there's still the tyranny of the Landauer Limit.
When was the last Llama release? Meta have abandoned it and reportedly they've had a shift in their AI strategy.
Meta Llama is free for many uses but it doesn't even remotely meet the definition of "open source".