Gregory's ServerWatching Google I/O's presentations.
They're pushing generative A.I. so much that I'm starting to count the number of fingers on presenters.
 8 comments
It runs the gambit. They had a whole thing about injecting AI improvements before search results (not super helpful) to creating content and code, to some of the photo and voice editing stuff they do on devices. But yeah, every speaker on the keynote so far has mentioned AI at least once. @jrconlin I'm not surprised about them putting an LLM above the Search results, but also not happy. That would divert traffic away from the sites below, which provide the source material for the LLM, presumably without consent, attribution, or compensation. So, not only is this bad for the open web, but in the long run it'll undermine Google's LLM, too.  Honestly, if I was on the fence about the usefulness of AI before the amount of hype from tech companies has totally convinced me it's just the new NFTs. Amount of tech hype = percentage of bullshit. There are actual, useful things that LLM can do. Heck, even for search, LLM could help disambiguate terms. (e.g. What do you mean by "jaguar"? The car, the animal, the football team?) Ducktaping AI into everything, though, is just marketing. The I/O keynote is basically a marketing presentation, but still... They did introduce James Manyika who leads the "Technology and Society" group and spoke about Responsible AI. But, yeah, I too remain skeptical. |
@jrconlin Sigh. Are they talking about doing anything novel or meaningful with it, or just dropping the buzzword at every opportunity and stuffing it into every product?