Debunking also doesn't undo the original damage because most people will never see the debunking. For example, a while ago a paper about the oceans supposedly h

Debunking also doesn't undo the original damage because most people will never see the debunking. For example, a while ago a paper about the oceans supposedly heating up much more than previously estimated hit a lot of news sites and HN: https://news.ycombinator.com/item?id=18352506 https://news.ycombinator.com/item?id=18352506 The article currently at that link is not what was there at the time, it's essentially a retraction that almost none of the HN readers who saw the original will have seen. That paper had an estimate of ocean warming that was much higher than that found using actual widespread physical measurements of ocean temperature, based on a weird indirect calculation from global CO2 and O2 levels in the atmosphere. It turns out they'd massively underestimated the error bars on their weird indirect estimate and it was just far too inaccurate to actually conclude that ocean warming was substantially higher than the direct measurements. (Which is not at all surprising really.) Now go and read the HN comments on that science at the time, they're quite something given the subsequent context...