zlacker

>>apsec1+(OP)
Sam wanted to commercialize stuff to shoot for revenue. Ilya wants to keep pushing for gpt 4.5 and beyond, to hell with the revenue. Ilya won the argument, Sam out.

Hell yeah.

It's not safetyism vs accelerationism.

It's commercialization vs innovation.

>>mvkel+1a
OpenAI's mission statement is "Creating safe AGI that benefits all of humanity".

How does an LLM App Store advance OpenAI toward this goal? Like, even in floaty general terms? You can make an argument that ChatGPT does (build in public, prepare the world for what's coming, gather training data, etc). You can... maybe... make an argument that their API does... but I think that's a lot harder. The App Store product, that's clearly just Sam on auto-pilot, building products and becoming totally unaligned with the nonprofit's goal.

OpenAI got really good at building products based around LLMs, for B2B enterprise customers who could afford it. This is so far away from the goal that, I hope, Ilya can drive them back toward it.