> On 5 Jun 2026, at 13:15, John R Levine <[email protected]> wrote: > >>> So anyway, while I think it is a good idea to keep abreast of all this, it >>> is >>> not a good idea to design processes based on our guesses about what the >>> legal >>> treatment of LLM output will be because whatever our guess is, it will be >>> wrong. >> >> What concerns me is the inadvertent creation of a legacy mess from now until >> that is resolved. i.e. wishing that we had asked some basic questions early >> on to inform later decisions. I can see that this might be considered >> premature optimisation. > > We can certainly ask, but it's up to the contributor how much to answer. > > Honestly, if it turns out that we're publishing big chunks of AI slop > unedited,
This is based on the assumption that AI produces slop. In my experience the more effort you put into it, the better the quality of output. It will not be too long before many, many people can instruct AI well enough to produce high quality output. Jay > or so lightly as not to pass the Feist* threshold, unauthorized translations > will be the least of our problems. > > R's, > John > > * Feist vs Rural Telephone was the phone book case > -- Jay Daley IETF Executive Director [email protected] -- rswg mailing list -- [email protected] To unsubscribe send an email to [email protected]
