Email or username:

Password:

Forgot your password?
Simon's posts Post Back to profile
Top-level
th0ma5

@simon great documentation ... Any details on accuracy? How much did you have to clean up the output and did you have to check it all by hand?

18 October at 5:44 | Wall-to-wall | Open on mastodon.social
4 comments
Simon Willison

@th0ma5 I checked it all, didn't take long (I watched the 35s video and scanned the JSON) - it was exactly correct

18 October at 6:56 | Open on fedi.simonwillison.net
Felix Westphal

@simon @th0ma5 you being surprised that this actually worked tells a lot about the state we're in and afaik with this technology we can never be sure that the result will actually be correct. So if you have to double-check anyways you can just do it yourself manually (or use another non-LLM-tool).

18 October at 16:15 | Open on fosstodon.org
Simon Willison

@superFelix5000 @th0ma5 right - the single hardest thing about learning to productively work with LLMs is figuring out how to get useful results out of inherently unreliable technology

18 October at 16:30 | Open on fedi.simonwillison.net
th0ma5

@simon @superFelix5000 Sorry my previous reply I didn't realize that you coded. The manual setup and such that you did is the hard part right? Extracting frames and OCRing them can be done with just command line tools. I think your documentation is great, but it doesn't feel like a net gain to me.

19 October at 18:38 | Open on mastodon.social
Go Up