Re: Google AI Overviews – Genuinely, Why?
This might come across as a blatant attempt to bash Google's AI Overviews, but is actually a genuine question that I've been asking myself constantly over the past few days. Wanted to post here to see if the HN community had any takes/insight that I might be missing.
My confusion comes from multiple places:
1. Destroying the trustworthiness of the Google brand. It's hard for me to believe that Google would intentionally destroy decades of earned trustworthiness and goodwill toward its core product and money maker by pinning AI answers that will always inherently (due to the technology) have a chance of being completely false. Google search has become a meme, not just in tech circles but in the wider world. Every news publication seems to have reported on this. People that I've never heard talking about technology are aware of this. I simply can't wrap my head around the calculus of this - Sissie Hsiao (VP of Gemini/Assistant) was the one to announce AI Overviews at I/O, so is this some sort of internal power struggle where the Search product is caught in the crossfire?
2. Who benefits from AI Overviews? Usually when you think about controversial features, there is a clear beneficiary. In this case, the main parties involved are Google, users of Google search, advertisers, and site owners, and I think all of these parties are suffering (though maybe not short-term Alphabet shareholders). As mentioned above, Google's brand is being obliterated, but not just on the user side - they've also lost goodwill with advertisers and site owners, the other two critical components of the search ecosystem. What do AI Overviews actually do for Google for it to be worth this destruction of this trust? Is the main objective to show the world that Google is good at AI? Is it to release a product that's clearly inferior to traditional Google search to anchor the world into thinking that similar offerings from OpenAI/Perplexity would be equally inferior to Google search? Google continues to claim that AI Overviews are a better experience because they make users search more (I'm skeptical that these metrics mean what they think) [1] and that unsafe/incorrect answers are an edge case [2], while dodging questions about the effects on publishers [3].
3. Google's AI panic. It's weird to see Google obsess over showing the world their AI chops - no one clued into the tech industry ever doubted that they have always been among the industry leaders in AI. It truly feels that someone high up (Sundar?) panicked. Microsoft has been upselling Copilot in Bing/Edge/Windows for ages, and yet even they haven't tried to stick on-by-default AI Overviews to the top of their searches. Let that sink in - Bing is arguably a better search experience than Google now, and it's not because Bing got better. Did Microsoft intentionally leak a rumor that OpenAI was working on a ChatGPT-powered search engine [4] in an attempt to goad Google into spending resources on this?
4. AI Overviews - on by default, no opt out. No one forced Google to do this, and this seems to be a very strong signal that Google thinks these overviews are an objective improvement on the existing search experience. They certainly knew the consequences here, yet they did it anyway. Was this a calculated strategy, or just a "fingers crossed that it works out" fumble?
Personally, the rollout of AI Overviews has been the only time I've ever considered using a search provider other than Google - I've been trying Bing/DDG for some searches and am looking into Kagi as well. I'm sure I'm not the only one, so I'm curious what I'm missing here.
[1] https://arc.net/l/quote/voybpunr [2] https://arc.net/l/quote/logaroxb [3] https://searchengineland.com/google-users-click-links-ai-answers-440731 [4] https://mashable.com/article/chatgpt-search-engine-google-rival-launch-before-googleio