Plain inference is profitable actually, that's why there are a hundred inference providers on OpenRouter who compete by undercutting each other. The labs however aren't profitable because training the models is a huge drain.
pooterbroo
joined 1 day ago
I know you might be joking and it's funny, but I don't think we should make AI nudes of anyone even if they're attractive.
Well they didn't even use the latest models in Feb 2025. They should've used DeepSeek R1 and OpenAI o3-mini which use additional test time compute to arrive at better answers. They used GPT 3.5 which was about 2½ years old at the time.
They presented five generative AI chatbots—Gemini (2.0, Google; version available December 2024), DeepSeek (V3, High-Flyer; version available December 2024), Meta AI (Llama 3.3, Meta; version available December 2024), ChatGPT (3.5, OpenAI; version available November 2022) and Grok (2, xAI; version available August 2024)—with a series of closed- and open-ended prompts across five misinformation-prone categories.
Couldn't the researchers at least bother to use the latest models?
I think the preferred term for that is "gooning" nowadays.