3 ms·
The opposite, is claiming that OpenAI could have now built better performing, cheaper to run model (when compared to what they published) training it at 1% cost
by scotty79 2y ago
The opposite, is claiming that OpenAI could have now built better performing, cheaper to run model (when compared to what they published) training it at 1% cost on output of their previous models. ... But they chose not to do it.