DeepMind boss ‘would pay thousands of dollars per month’ to get rid of his email, so Google is working on a next-gen Gmail AI that will answer them in ‘your style—and maybe make some of the easier decisions’ for you

My Gmail inbox currently has 84,685 unread emails. Were there an AI tool that could clear my backlog automatically, let’s just say you could sign me up and take my money. It’s the only way I’m going to achieve inbox hygiene (there’s always the trashcan button – Ed.). Rejoice, then, because the head of Google DeepMind says they’re working on a next-gen Gmail that can answer emails for you.

Yep, Demis Hassabis has been speaking at the SXSW London festival (via the Guardian) about all things AI. And apart from predicting the arrival of AGI or artificial general intelligence within five to 10 years, Hassabis revealed Google’s plans for clearing the world’s email backlog.

“I would love to get rid of my email. I would pay thousands of dollars per month to get rid of that,” Hassabis said. More specifically, he said Google was planning, “something that would just understand what are the bread-and-butter emails, and answer in your style—and maybe make some of the easier decisions.”

Such a technology raises all kinds of immediate questions. Such as, why not pay a human thousands of dollars per month to manage that for you. Like an executive assistant maybe? And exactly what kind of email will it answer? Will it actually save you much time if in practice you need to know it’s answered for you and what exactly it said?

I mean, if your Gmail bot replies, “sure, I’m on it” or “thanks, understood,” well, you’d need to be actually on it or have understood it, right? Likewise, do you trust any AI with your personal communications?

Slightly more dystopianly (if that’s a word, and it ought to be, I sense we’re going to need it, though it might need another ‘n’ or perhaps an extra ‘l’), it also raises the prospect of bots spiralling off into endless back-and-forth comms.

After all, if you are a Gmail user who emails another Gmail user, you’d presumably have situations where you fire off an email, their Gmail bot replies, following which your Gmail bot decides to weigh in and they’re off to the races, chatting away.

Somewhat ironically, Hassabis also said AI could be used to protect you from attention-grabbing AI algorithms. The idea is that an AI assistant might give you, “more time and maybe protect your attention from other algorithms trying to gain your attention. I think we can actually use AI in service of the individual.”

So, that’s Google’s bots protecting you from the attention-grabbing efforts of other companies’ bots. What a time to be alive that will be. Of course, we could just cut out the toxic attention grabbing nonsense in the first place.

Sorry, silly idea. Without the toxic algorithms, Google couldn’t sell you a bot to protect you from its competitors, while those competitors presumably sell you another bot to protect you from Google’s algorithms. This, people, is the future. And it’s beautiful. Kinda.