Post AsJ9mKFUEeF22f0WXo by ianb@mastodon.well.com
 (DIR) More posts by ianb@mastodon.well.com
 (DIR) Post #AsJ9mJ5sWsxgSZVKvg by ianb@mastodon.well.com
       2025-03-22T07:39:34Z
       
       0 likes, 0 repeats
       
       There are people - and I confess I am one - who use AI for search and say to themselves “well yeah, I know it gets things wrong, but I’m smart enough to spot that so I’ll keep doing it”.Reader, I am not smart enough to spot that.
       
 (DIR) Post #AsJ9mKFUEeF22f0WXo by ianb@mastodon.well.com
       2025-03-22T07:39:58Z
       
       0 likes, 0 repeats
       
       But this begs the question: knowing about the factually inaccuracies, why keep doing it? And there are two parts to that. First, Google searching isn’t exactly full of the best answers right now (this is why people add “wiki” to the search query, because they know Wikipedia gives a better answer)
       
 (DIR) Post #AsJ9mLQrpowHiFL7vE by loke@functional.cafe
       2025-03-22T07:42:05Z
       
       0 likes, 0 repeats
       
       @ianb also, I think for a lot of people the time wasted due to errors are offloaded to others. I've often had to spend more time reworking someone's "helpful" assistance than it would have taken to do the work outright.
       
 (DIR) Post #AsJ9mMN0LfRYcYXYXY by ianb@mastodon.well.com
       2025-03-22T07:45:52Z
       
       0 likes, 0 repeats
       
       @loke There is that, but I think the same is true of much internet based "research". It's gone backwards in the past three years broadly, because of the inevitable consequences of monopolisation of finding information.
       
 (DIR) Post #AsJ9mNMgeKmdhrOogS by loke@functional.cafe
       2025-03-22T07:56:03Z
       
       0 likes, 0 repeats
       
       @ianb absolutely. If we move the clock backwards to before the arrival of computer-based searching, the difficulty was in finding the information in the first place, but once you had it you could be reasonably sure that it was true (at least a true statement made by whomever made it, religious texts notwithstanding).Today, this is completely flipped on its head. Information is easy to get hold of (in fact, it's being pushed to you) but assessing its accuracy is becoming harder and harder, even for someone who is actively trying to research a subject.The problem is that people still think research is about finding information, leading them to think they've done 'research' when they followed a series of youtube links.
       
 (DIR) Post #AsJ9mO16E56rjCJ5fs by publius@mastodon.sdf.org
       2025-03-22T09:05:10Z
       
       0 likes, 0 repeats
       
       @loke @ianb I'm to the point of wanting "books and papers published before the year 2000" as references.
       
 (DIR) Post #AsJ9mPzOtMwHqVWlIu by ianb@mastodon.well.com
       2025-03-22T07:40:10Z
       
       0 likes, 0 repeats
       
       But second, the user experience of a plain language search and single answer is a better user experience than ten links of random quality plus ads plus answer boxes plus other misc garbage and more ads. Google enshittified itself into uselessness.
       
 (DIR) Post #AsJ9mVq3931ZzFCNYu by ianb@mastodon.well.com
       2025-03-22T07:40:22Z
       
       0 likes, 0 repeats
       
       And conversational interfaces are great! When you get your answer you can fine tune it, give it more information to get a better answer. For many kinds of information query, it better matches the way our messy, imprecise, uncertain brains work.
       
 (DIR) Post #AsJBPinWzoVNQZHTKC by loke@functional.cafe
       2025-03-22T09:23:32Z
       
       0 likes, 0 repeats
       
       @publius @ianb and when looking up inffrom blog posts and the like, make sure it was written before 2021 or so.Related, if a blog post or article contains "updated 2025" in the title, especially if it's about a topic that doesn't change, you know it's garbage.