- cross-posted to:
- fuck_ai@lemmy.world
- cross-posted to:
- fuck_ai@lemmy.world
You know how Google’s new feature called AI Overviews is prone to spitting out wildly incorrect answers to search queries? In one instance, AI Overviews told a user to use glue on pizza to make sure the cheese won’t slide off (pssst…please don’t do this.)
Well, according to an interview at The Vergewith Google CEO Sundar Pichai published earlier this week, just before criticism of the outputs really took off, these “hallucinations” are an “inherent feature” of AI large language models (LLM), which is what drives AI Overviews, and this feature “is still an unsolved problem.”
Google providing links to dubious websites is not the same as google directly providing dubious answers to questions.
Google is generally considered to be a trusted company. If you do a search for some topic, and google spits out a bunch of links, you can generally trust that those links are going to be somehow related to your search - but the information you find there may or may not be reliable. The information is coming from the external website, which often is some unknown untrusted source - so even though google is trusted, we know that the external information we found might not be. The new situation now is that google is directly providing bad information itself. It isn’t linking us to some unknown untrusted source but rather the supposedly trustworthy google themselves are telling us answers to our questions.
None of this would be a problem if people just didn’t consider google to be trustworthy in the first place.
I do think Perplexity does a better job. Since it cites sources in its generated response, you can easily check its answer. As to the general public trusting Google, the company’s fall from grace began in 2017, when the EU fined them like 2 billion for fixing search results. There’ve been a steady stream of controversies since then, including the revelation that Chrome continues to track you in private mode. YouTube’s predatory practices are relatively well-known. I guess I’m saying that if this is what finally makes people give up on them, no skin off my back. But I’m disappointed by how much their mismanagement seems to be adding to the pile of negativity surrounding AI.