So, am I wrong in thinking that things like the above (Meta adding their Llama models into all of their apps: Insta, WhatsApp FB) is a play to take Google Search Volume? Like an expanded version of folks searching for "best restaurants in new orleans" type IG searches.
Well, I think there's the "sprinkle some AI on it" startups that make a few AI calls and tell the world they've pivoted and then I think there's the Meta + Apple's + Microsofts + Amazons of the world that look at this as a the first real wedge into Google's slice of the pie.
And I'm trying to figure out what that looks like - for example Amazon (Bezos) is funding Anthropic + Claude and that's now the "Rufus" AI shopping chat app that's all over the Amazon site + app.
Yeah, maybe. Good topic. But this isn't the first go around with AI for those larger companies. Did they really just sit around and wait for an AI startup to build something just to invest in it when they had the budget to build the entire time? Amazon and the like benefit from searches that do not happen in the app because of Google. Wouldn't taking search share from Google be like stepping on your own throat?
yes, they are and in theory the margins would be higher for sure. I would have to believe that most traffic goes through the mobile app, however. Now I am going to have to do some digging. :melting_face:
Those cameras for the Amazon example do happen in their own search already. It's just that for some searches people aren't going to have the association of Amazon being the place to go. Like for the best restaurants query. And the WhatsApp AI thing is annoying. I use WhatsApp for one thing - communication with contacts. Having some stupid AI thing in the way helps me none.
Well, take Instagram, there are already searches like: "Best restaurant in New Orleans" happening on it - my assumption was that Meta is going to expand that out with the AI responses to capture more of it
I'm not sure understand "follow on questions" for IG. Maybe you meant follow up? So are you saying basically secondary search filtering against the initial result?
So wrt Rufus and Amazon - I'm currently buying an upgraded streaming and video camera. So I'm trying to figure out if I should get a "dummy battery" and external charger. Until Rufus - I would have needed to ask that question of Google and then come back to Amazon. Now it's all within Amazon.
That implementation and use case makes sense. I'd call it a shopping assistant in this example.
It's still a bit clunky though. I think they may be concerned about disrupting the traditional purchase path until there is more data / if it supports better integration.
But you're not going to Amazon to look for restaurants.
Yeah, but also this goes to my "AI is just going to be everywhere" thing - like how databases are today. There's not that many people who _make_ databases, but every app on your phone and every saas has a bunch they use. AI is just going to be in everything.
I don't disagree on that. It's just about how it's implemented and whether it's helpful/improves CX in a particular implementation. On Amazon it totally can. Their review summaries are helpful. Other than they don't use it to detect when there are irrelevant reviews due to product change/hijack.