Will ChatGPT conversation history be critical to the success of future OpenAI models? (Ṁ200 subsidy)
➕
Plus
17
Ṁ469
Jan 1
69%
chance

Resolves YES if OpenAI publicly acknowledges this data being particularly useful for training further iterations of GPT4 or different successors, or trustworthy leaks (my credence), stating that this was an integral component of success.

Resolves NO if other factors (architecture, training techniques, other token sources) are judged as having been largely more important.

Resolves N/A if market closes and question is unsanswerable due to insufficient information being public.

Get
Ṁ1,000
and
S1.00
Sort by:

I understand this is going to necessarily be fuzzy, but of it’s responsible for a significant minority (say 20%) of the performance improvement for GPT-5 does that count?

@MatthewRitter I think if there was significant evidence that it was 20% of performance improvement I would still count that!