This year’s I/O developer conference felt a little repetitive. Almost everything announced was an iteration on what Google showed off last year. But there was one theme that was particularly clear, especially on day one: Google is becoming more and more confident with its bet on AI.
The consumer version of this is Google Assistant. Last year, Google Assistant was merely introduced, not even launched. This year, there were so many features, integrations, and partnerships announced that we wrote a separate post just to summarize it all.
Google is still looking for its next big revenue driver. Like how Microsoft had Windows and Office, or how Amazon has ecommerce and AWS, Google wants to find that number two. Search and ads are its bread and butter, but nothing lasts forever. The company needs another hit, and it knows that to find one will take years of investment.
Google has settled on AI. The custom tensor processing unit (TPU) chips unveiled last year also got an update. TPUs are chips developed specifically for machine learning. Make no mistake: When a tech giant invests in infrastructure, it’s not messing around.
Nobody at the company knows how the AI bet will play out, and that’s fine. The huge advantage of investing in AI is that it’s not a siloed vertical — you can use it to improve your products across the board, not unlike Google did by building search right into products like Gmail and Chrome.
Assistant could be the search successor Google is looking for. I mean, you can now type to it just like into a Google Search box, not just use your voice. There are still many cases where a standard page of search results is preferable, but it’s easy to see how Google could consistently cut that down by improving Assistant.
Or, Assistant could fizzle out like Google Now and Now on Tap before it. So far it’s got great momentum, but it’s nascent and unfinished. But that doesn’t really matter, because Google is playing the long game. Google isn’t throwing money into Assistant — it’s throwing money into AI and watching how that pays off across its current product line, its new projects, and beyond.
If ads are what make search viable, AI is what makes Assistant viable.
ProBeat is a column in which Emil rants about whatever crosses him that week.