Gavin comments on Steelmanning Inefficiency - Less Wrong

18 Post author: Stuart_Armstrong 03 July 2014 12:17PM

You are viewing a comment permalink. View the original post to see all comments and the full post content.

Comments (25)

You are viewing a single comment's thread. Show more comments above.

Comment author: Gavin 03 July 2014 06:19:02PM 8 points [-]

Isn't that a necessary part of steelmanning an argument you disagree with? My understanding is that you strengthen all the parts that you can think of to strengthen, but ultimately have to leave in the bit that you think is in error and can't be salvaged.

Once you've steelmanned, there should still be something that you disagree with. Otherwise you're not steelmanning, you're just making an argument you believe in.

Comment author: Wes_W 04 July 2014 06:08:47AM 6 points [-]

Part of the point of steelmanning, as I understand it, is to see whether there is a bit that can't be salvaged. If you correct the unnecessary flaws and find that the strengthened argument is actually correct (and, ostensibly, change your mind), it seems appropriate to still call that process steelmanning. Or rather, even if it's not appropriate, people seem to keep using it like that anyway.

Comment author: formido 04 July 2014 06:25:57PM *  5 points [-]

If you take a position on virtually any issue that's controversial or interesting, there will be weaknesses to your position. Actual weaknesses. I thought the purpose of steelmanning was to find and acknowledge those weaknesses, not merely give the appearance of acknowledging weaknesses. If that's not right, then I think we need a new word for the latter concept because that one seems more useful and truth seeking. If you're stretching things beyond the domains of validity and using tricks, it sounds awfully like you're setting up straw men, at the very least in your own mind. Seems more debate club than rationality.

Comment author: Stuart_Armstrong 04 July 2014 09:51:46AM 1 point [-]

A much better phrasing of what I was thinking.

But I think Kaj approach has some merit as well - we should find a name for "extracting the best we can from opposing arguments".