I have yet to use auto-voting. I signed up for hive.vote and set up a few auto-vote criteria, but I have kept the threshold high enough that it never (to my knowledge) has kicked in.

With that said, under the HF25 scheme, if I were to use an auto-voting bot, I would want to configure it to follow and support manual curators, by auto-voting at the 23:45 mark (just inside the 24-hour threshold) and only voting on posts that had already received both a high number of raw votes (a lot of people voting for it) and, to a lesser degree, a decent $ amount of votes.

Under that scenario, my auto-voting would merely be an amplifier for early (presumably mostly manual) curation.

From a pure 'curation reward maximization' strategy standpoint, I think the above approach would be maximal. This is because I would be voting on posts that already have received large numbers of upvotes (so it is relatively good content), and my vote will amplify that, and amplify it just at the moment when later curation rewards are modestly shifted toward my benefit.

So, in short, I believe HF25 will essentially accomplish what I was arguing for here when I talked about incentivizing auto-voting bots to vote at a threshold window that was a day or more removed from the initial post.

With all that said, I would support changing the threshold from 24 hours to 36 hours.

That way the auto-vote amplification (assuming others follow my logic from above and implement it) will occur at a 180-degree phase shift on the global time scale. In other words, folks that engage on Hive at the same time each day will have their posts amplified at a completely different (opposite) time of day; that is more likely to cross-pollinate the visibility of posts across time zones and across 'habit zones' (is that a thing?).