Skip to playerSkip to main content
  • 3 months ago
Transcript
00:00Dave Gurra. Yes. What do you Google? That's a great question. We just did sports. I Google
00:05sports scores if I don't watch the end of a game. Recipes. If I'm traveling, I'll Google best hotel
00:10or restaurant in a given city. Very practical. I feel like your Google is very, very, very
00:15practical. Mine is a little chaotic, but yes. All right. Well, Simon Rogers has written a book.
00:20I feel like I have to ask you a follow-up on that. A little chaotic how. You know me.
00:23My brain is
00:24just as chaotic. It's things that I should know, but I'm not sure about. I check on best. It's more
00:29where am I in New York's happening right now, but we'll move on. Sorry to interrupt. All right. So
00:34my friend Simon Rogers has written a book on exactly that. He's Google's data editor. And
00:38before that, he was the data guy at Twitter. He's also the winner of one of the most amazingly hyper
00:43niche awards I've ever heard of, which is the Royal Statistical Society's Award for Statistical
00:48Excellence in Journalism. Simon, you giant nerd. It's so nice to see you. I feel such a nerd today.
00:54Thank you for having me. The book is What We Ask Google. Let's start with the question that
00:58Christina posed to me as she hoists it up here. Broadly speaking, what are the bulk of
01:03queries about on the search engine?
01:05So the bulk of queries are really about the everyday stuff. I mean, I think we tend to focus
01:09on the, you know, the big news events, which, you know, makes sense when something happens,
01:12we want to understand it. But really for most of us, it's just getting through the day.
01:16It's parenting. It's like pets, food, health, grief, you know, these kind of big moments.
01:22And I'm not seeing a kind of real exploration of that everyday nature of search. So that's
01:27what I wanted to look at.
01:28It's really fun. And you break it down into different sections. And there is so much data
01:32in here. I understand why you won that award. Your very first chapter is called Why Am I
01:37So Tired? Which I think is the refrain of modern adulthood. But a lot of it, surprisingly,
01:42is about time. People search like, what time is it? They search, you know, around 2am, you
01:48say you see a spike in parents searching for how to get their kids to sleep. Why do we have
01:52daylight savings time? Why do we have leap year? And then what time is it in Ukraine spiked around
01:57the invasion?
01:58Yeah, exactly. It's like, it's a real mix of the everyday. So yeah, you can't get your baby
02:03to sleep around 2am. You're not alone. You're sharing it with millions of people around the
02:07world. And also there is a kind of rhythm to the way that we search. And it applies during the
02:11year as well. So people search for how to play piano in December, you know, ahead of Christmas,
02:16right? And right now, people are searching for graduation. Month ago, it was prom. Month
02:21before that, it was promposals. And that's a global search now. That's one of these things
02:25which, yeah, I grew up in England, did not grow up with proms. And now it's a global thing.
02:30So you see these searches kind of spread around the world.
02:32Could you talk a bit about your approach to the job? So look, I search on Google and it
02:37auto-populates it sometimes, what other people are asking questions of. But when you're faced
02:41with this gigantic, doesn't even do it justice, trove of data, how do you begin to piece through?
02:46What are you looking for? What are you hoping to glean from all that you can aggregate from
02:49our use of it? I mean, for me, I'm really interested in
02:52the humanity behind the searches. So for instance, whenever there's a hurricane, you know, you'd
02:57expect people to be searching for battering down the hatches and, you know, buying emergency
03:01supplies. But people also search for how to calm a dog in a storm, which I love that search.
03:06Yeah, I mean, it's very human, right? And I think that there is a kind of humanity which
03:12comes across searches, which, you know, maybe we're lacking a little bit in the way we think
03:16about each other. But actually, you can really see it come through. Like searches for help
03:20are higher now than ever before. And they're often about helping other people.
03:24I thought that was interesting, because the premise of this seems like nightmare fuel.
03:28There's a UK game show called Taskmaster. And they recently had a challenge where contestants
03:32had to do a really scary thing. And the guy who won let them read his search history live
03:37on the show, because it seems like it could bring out the worst in us. But to your point,
03:41there's some really cute and kind of reaffirming moments. One of them is in June, people search
03:47like how to save a bee because bees are out and they want to rescue them, which I thought
03:51was really cute. But talk to us about some of the other things that as you're looking through
03:54this data made you find commonalities, not things that separate us.
03:58Yeah, I mean, I knew I wanted to write the book. And I started doing it a couple of months
04:02after my mum had died. And, you know, anybody who's joined that club, you know, you feel
04:06incredibly alone, you're the only person this has ever happened to. And when you start looking
04:10at the data, you realise that not only are you not the only person this has ever happened
04:14to, there are lots of people in the same situation. But people are also searching for how to help
04:18others in that situation, like the right thing to say or the right bit of poetry to read
04:23or whatever it is. And that kind of gave me a couple of themes of the book, really, because
04:27this idea that you're not alone, even if you feel lonely, you're one of a community
04:31of people who are doing similar things, whether or not you know it. And also that other people
04:36are trying to help you, which is kind of reassuring and reaffirming. And also really
04:42that we know nothing, you know, that everyday life, whether it's how to boil an egg or how
04:46to tie a tie, you know, these things. And we're not ashamed to ask for it.
04:51How is your life being transformed by AI and the way that people are using it to ask some
04:56of these questions? And I think about the rigor of thinking, I think a criticism of AI is
04:59somebody might search a query, like you just asked, what time is it someplace? And this
05:03is something that we could answer very basically, if we obtained a wristwatch or looked at the
05:07clock across the room. And it's using all of this computing power to find this very basic
05:11question. I'm curious how you think about that, just the way in which we're using search and
05:15how that's kind of transforming our approach to finding out what, in some cases, is very
05:18basic information.
05:19Yeah, my favorite is how to tell the time drunk, which is most searched in London, which
05:25is no surprise.
05:26Sure.
05:27There is a change and a revolution happening in the way people search. And we have some
05:31data out this week around the way that people search with AI mode in Google, which is, you
05:36know, we've had for a year now. So obviously we've got over 20 years of regular search history
05:41and now we've got a year of AI search history. So it's really early days. But what you're seeing
05:45is that queries are getting longer. They're much more specific now than they used to be.
05:50So you'll see a car in a movie and you want to know what brand it is. Or you'll hear
05:55a
05:56song and you want to know what instrument they played on it. And they're much more
05:59conversational. So people will go back and forth and kind of return to it because...
06:03That's so interesting to me because as a kid who grew up in the 2000s, there was this
06:07thing called Ask Jeeves where you could, like, before people knew how to query, you could
06:11put what is, you know, blank and it would find it to you in Google.
06:15Search terms were so primed to just enter keywords. It's interesting to me now that
06:20we're going back to searching in natural language models.
06:22It's a whole new set of skills, isn't it? And I think it's basically people are just
06:25asking how they want to ask. So me and my son, we're doing up an old BMW at the moment,
06:31stupidly complicated. And so we're doing a lot of these very kind of weird diagnostic
06:35things where we'll come back and we'll go back to it. So you're really seeing it in that
06:39way. And also searches are becoming much more kind of multimodal. And by that you mean
06:43that people are searching not just on the phone or on a laptop, but they're searching
06:46with a voice or a home device. But I, my sense is, and it's early, is that the things we're
06:52searching for are staying the same. We're still searching for love and relationships
06:56and, you know, health and hope and that sort of thing.
06:59Donna, I liked the, towards the end, you had, you end on a nice note and you, you have a
07:04graph
07:04that shows the number of people searching for a job that makes money and the number of people
07:09searching for a job that helps people. And over time, then the line for helping people
07:13far surpasses the one for people wanting to make money. What do you think that is? And what's your
07:18biggest takeaway?
07:19I mean, I think you're seeing these social changes happening and you can see the data
07:24through search. So for instance, searches for take care of parents have overtaken searches for
07:29take care of kids, you know, with a sandwich generation, maybe you're looking after parents
07:33and looking after children. And a lot of these kind of big social trends, you can see showing up
07:38in search.
07:39Simon, thank you so much. The book is what we Google. Thank you for coming in.
07:43And if you need to feel better about humanity, I highly endorse it. All right. For more reviews
07:47and recommended books, be sure to subscribe to the new Bloomberg Weekend Newsletter on books.
07:51You can sign up at Bloomberg.com slash newsletters.
Comments

Recommended