George Macgregor<p>Good grief. This is far, far worse than I thought possible. On average, <a href="https://code4lib.social/tags/AI" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>AI</span></a> <a href="https://code4lib.social/tags/search" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>search</span></a> assistants were confidently incorrect on > 60% of queries. Especially pleased Grok 3 was incorrect on > 94% of queries! 😉 All tip of the iceberg stuff.</p><p>AI agents delivered "confident presentations of incorrect information, misleading attributions to syndicated content, and inconsistent information retrieval practices."</p><p>AI Search Has A <a href="https://code4lib.social/tags/Citation" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Citation</span></a> Problem - Columbia <a href="https://code4lib.social/tags/Journalism" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Journalism</span></a> Review<br><a href="https://www.cjr.org/tow_center/we-compared-eight-ai-search-engines-theyre-all-bad-at-citing-news.php" rel="nofollow noopener noreferrer" translate="no" target="_blank"><span class="invisible">https://www.</span><span class="ellipsis">cjr.org/tow_center/we-compared</span><span class="invisible">-eight-ai-search-engines-theyre-all-bad-at-citing-news.php</span></a></p>