I changed it to "human authorship for the purposes of this policy".

If Claude suggests a one-line change, and I agree with it, and (after careful examination) would have written exactly that change myself, that's enough for me.  It would be odd to say that I must re-type it myself.

I really don't want to be silent on this issue.  I want us to be explicit that every line of code has a human being standing behind it, vouching for it, taking personal responsibility for it.

If anyone wants to offer a better term than "authorship" to express this aspiration, by all means do suggest it.  (Terminology is always hard.)

I sense that the text has notably moved towards the “pro-LLM” side.

The intent (perhaps badly executed) is to be neither pro nor anti, but rather neutral about whether LLMs are good or bad.  Rather that be pro or anti, since we want the policy to guide the behaviour of a very diverse group of people, it tries to de-emphasis the tools and instead concentrate on (a) the artefact itself (e.g the MR) and especially (b) the effect on our shared discourse and the motivations and agency of both reviewers and contributors.   

That neutrality may not truly satisfy anyone, but I hope that it may perhaps constitute a shared statement of some things we agree about (e.g. respectful discourse, high quality code base), and (because it is neutral) be equally un-satisfying to everyone.    Of course, even what "neutrality" means might be controversial. 

I'm grateful to Wolgfang and others for their willingness to move forward together even if we disagree in some ways.

Simon

On Tue, 21 Jul 2026 at 14:43, Wolfgang Jeltsch via ghc-devs <ghc-devs@haskell.org> wrote:
Am Di 21.07.2026 14:32 schrieb Tom Ellis via ghc-devs:
> On Tue, Jul 21, 2026 at 04:24:53PM +0300, Wolfgang Jeltsch via ghc-devs wrote:
> > I sense that the text has notably moved towards the “pro-LLM” side.
> > Sometimes this led to outright wrong statements, like the following:
> >
> > > If you carefully review and understand every line of code or
> > > documentation produced with LLM assistance, revise it where necessary,
> > > and take full responsibility for every line, that constitutes human
> > > authorship.
> >
> > No, it doesn’t. An author is a person who *created* the work. If the
> > work was assembled by an LLM, based on only guidelines by the person in
> > question, then said person acted only as a reviewer, not as an author.
>
> Would it help if the text were changed to “… that constitutes human
> authorship for the purposes of this policy.”? That is, make it clear

This would still mean using the term “authorship” in contradiction to
its actual meaning. Why not be honest and say that GHC doesn’t require
contributors to be authors? Or just be silent about this issue?

All the best,
Wolfgang
_______________________________________________
ghc-devs mailing list -- ghc-devs@haskell.org
To unsubscribe send an email to ghc-devs-leave@haskell.org