The whole business-news world was awash with speculation this week that Apple is building its own search engine (The Telegraph ($/2), Financial Times ($), Forbes, Forbes again, c|net, etc…). These sites all point out that Google is under anti-trust scrutiny, and part of that is the purported $8–12 billion that Google pays Apple to be the default search engine on Apple devices; and indicate that the fact that Apple has recently started showing its own web search results if you search from the home screen on iOS (try it yourself, and you’ll see “Siri suggested websites”) supports the idea. I have mixed feelings about this. On the one hand, Apple absolutely is doing something around search. They hired Google’s then head of search, John Giannandrea, more than two years ago, and have been building search expertise ever since. On the other hand, Apple isn’t necessarily good at this, and … why? As for not necessarily being good at this, have a look at Apple Maps. They’ve put a ton of money and energy into it, and after a decade, it’s now kind of OK — but still can’t compete with Google Maps. But the “why” question is even more important. Even if the justice department ends up quashing that Apple-Google deal and Apple loses out on that $10 billion per year, so what? Google would still almost certainly be the default search engine, because (as much as people malign their search results these days), it’s still head-and-shoulders the best. Google doesn’t break down their expenses by category, but you can bet that they’re paying billions and billions per quarter for search (both in terms of infrastructure, and on-gong R&D). Now, Apple can probably afford that — but even for Apple, it’s non-trivial, and it would involve not investing in something else, so I ask again: why? They’re not going to monetize search, so it will be pure expense. Do they think they’re going to sell more iPhones if they have their own search engine instead of using Google’s? No, this is all nonsense. There’s no way Apple is building a search team (and search engine) to compete with Google. I can’t think of a single reason they would. Even if they’re worried about Google pulling out, it’s not like they’d be left without a search engine. They could easily swap to Bing or Duck-Duck-Go, or … whatever. So, what’s going on here?
In much-less headline-grabbing news last week, it was reported that Apple had acquired Vilynx — a Barcelona based AI company — earlier this year. “[Vilynx] generated rich, motion previews of content offerings, and provided search tools for surfacing content across videos, articles, and images. It also included an AI-powered recommendations platform while also offering tools to see trending topics on websites and social media” (MacRumors). Prima facie, that actually bolsters the “Apple is building a search engine” argument. If you’re going to try to do better than Google, the one place you can get an “easy” win (“easy” because Google is terrible at it; but Google is terrible because it’s not at all easy to do) is in video search. If you could index and search YouTube (and TikTok and Twitch and …) better than Google can, that could be a genuine search-engine win. But I still don’t think it’s the right story — or at least not the whole story. As I discussed back in June, I’m convinced that Apple’s main strategic drive right now is “pulling an Apple” with Siri — that they want to take her to a level that separates her from the pack the way iPhone separated from feature phones; to do something genuinely new. If anything, I think Apple building a search engine is not because they want to compete with Google on search — it’s because having an index of all the human information on the WWW is a strategic asset underpinning an AI future. And while Vilynx had some interesting capabilities around providing deep metadata for videos, they’re far-more importantly a powerhouse AI company. I had spoken with them a few times, prior to the acquisition, and their founder is also a professor on AI at a university in Barcelona. That university has slowly been becoming a global hub for AI research, and she has a direct pipeline of talented engineers that she can recruit into the business. Couple that with the three other AI purchases that Apple has announced this year, and I think you start to get a glimpse of the future — and that future is Siri (and if we get a decent search engine out of the deal, OK, that’s fine too).