Chapter 23
Zack
Zack was excited. For years, he had been trying to persuade Charles that they needed to move into software, and Charles had always been hesitant, telling him that if he came up with something brilliant he might consider investing in it.
Up to this point, everything Zack had showed him had been pooh-poohed, but he was certain Charles wouldn’t be as dismissive of this. He couldn’t be.
‘I need a photo of a person,’ Zack said. ‘I could use your picture, but you’re well known so it won’t have the same impact. Half the people in the office have pictures of loved ones on their desks. Can you choose one at random? Not a child.’
Begrudgingly, with an air that his time was being wasted, Charles went around the office until he found a photo of a middle-aged woman on the desk of one of the marketing managers. The man’s wife, presumably; an ordinary-looking woman with light brown hair cut in a bob and a gap between her teeth.
‘Perfect.’
Zack opened the app – currently in its developmental phase – and positioned his phone over the photo, waiting for it to scan.
It beeped, and he said, ‘Now watch.’
The word Processing appeared above the woman’s photo. Zack knew Charles wouldn’t have time to ask any annoying questions because the app was so fast, and three seconds later, the results appeared.
‘Abigail Young,’ he read. ‘Fifty-two years old. Works in the planning department of Birmingham Council, married to Steve Young – that’s whose desk you took this from – and has two children, Frank and Grace.
She went to Bartley Green secondary school, and her main interests are, hmm, let me see.
She’s a member of a book group, her favourite band is Take That, and she loves dogs.
There’s a lot more here. I have her address, email and phone number.
I can also tell you that she attended her work Christmas party at the Distillery near Brindleyplace last Tuesday, and that she and Grace are going to see Lana Del Rey at Anfield next summer. ’
He handed the phone to Charles, who put his reading glasses on and scrolled down. As Zack had hoped, he seemed impressed. A little stunned.
‘These are the most recent online photos of her,’ Zack said. ‘Look, there she is at her Christmas party. Her colleague Mandy posted it on Facebook.’
‘I think I need new glasses,’ Charles said. He was acting like Zack had just performed a party trick, not quite getting it yet. ‘I can’t see her.’
Trying not to sound impatient, Zack double-tapped the picture with his finger. ‘There she is.’
She was at the back of a group photo, her head turned away, half in shadow.
‘I can hardly tell that’s her,’ Charles said.
‘I know. But the app can. That’s one of the things that’s most remarkable about it.
It doesn’t need a clear photo. It can identify people if they’re in profile, if they’re wearing sunglasses or face masks, whatever.
It can pick people out of massive crowd shots.
For example, when she goes to that Lana Del Rey concert, if someone posts a photo of the crowd and she’s in it and in focus, even if she’s twenty rows back, it will pick her out. ’
Zack could see Charles thinking. But it wasn’t quite sinking in yet. This was not a magic trick.
‘Choose another photo,’ Zack said.
Charles returned Abigail’s picture to Steve’s desk and, this time, came back with a photo of an elderly woman. Someone’s mum. Zack ran it through the app and the results came back in seconds.
‘Bindiya Patil. Born in Maharashtra, India, in 1950. Moved to the UK when she was twenty-three.’ He showed Charles the rest of the results, including photos of Bindiya at a wedding back in the autumn. ‘Want to try another?’
Charles was silent for a moment. He finally seemed to be grasping it.
‘So this is some kind of search engine.’
‘Exactly.’ Zack had that tingle. The sense that he was on the precipice of something amazing. ‘A search engine for faces. You can show it the face of anyone who has ever appeared online and it will find them.’
Charles was still holding Zack’s phone, staring at the pictures of Bindiya and the screed of information below.
‘I assume you’re going to tell me how it works.’
‘Of course. It searches a huge internal database of photos and information. Every picture a person has posted of him or herself, as well as those they’ve been tagged in, or where their name is on the caption.
It’s incredibly clever. It can even figure out the relationships between people, create networks based on family connections, professional links, friendship groups.
It searches online posts and news stories and any public database, like the Companies House site, or Directory Enquiries. ’
‘The whole internet.’
‘Pretty much. I mean, there are certain walls it can’t go behind.
We’re not stupid enough to try to get past government security barriers, but we don’t need to.
There are a lot of private companies that use face recognition now.
Banks, phone companies, any website that uses facial scanning for age verification.
You know the porn sites are going to have to do that soon.
’ He smiled. ‘It’s astonishing how much information people put online about themselves.
All the social media posts, the forum chats, the employee sites – then add all the news stories and review sites and payment platforms. Almost everyone can be found online, either through stuff they’ve posted themselves or things other people have posted about them. ’
Charles was beginning to look irritated. ‘You’re talking to me as if the internet is a new invention. Gravitas had one of the first e-commerce sites in the UK. I was a dotcom pioneer when people thought the worldwide web was an international network of spiders.’
How many times had Zack had to pretend to laugh at that joke? He cleared his throat.
‘Sorry. I’m just so excited and want to drive home how this works.
What we have here is a tool that can scrape all of these sites, all this data, creating a massive database of billions of pictures, along with all the biographical information that goes with those pictures.
The genius part is that when you scan a photo of a face, the system scans the database for a match, and it is insanely accurate.
You saw how it worked even when the face was partially hidden.
It’s able to build a full picture of someone’s face from the part that’s visible.
I’ve tested it using pictures of everyone I know, and it always gets it right. ’
Charles appeared to be deep in thought. Eventually, he said, ‘Is there nothing else like this out there?’
Zack knew he needed to respond carefully.
‘There is. There’s a tool that’s already being used by some police forces and government agencies in the US, but it’s purely photo recognition.
It’s able to identify people who commit crimes.
For example, if CCTV catches someone but the police don’t know who they are, they can use this tool to identify them.
It doesn’t do what this tool does, though, providing a biography as well as all the connections.
Ours is also far easier to use. The interface is beautiful. People are going to go crazy for it.’
Charles’s eyebrows went up. ‘You see this as a consumer product?’
‘I’m not sure. It depends on whether we could overcome the legal issues. Concerns about privacy. But I think if any company was going to be able to get past that, it’s us. We’re not a bunch of upstart tech bros. We’re well respected, and you have a lot of contacts.’
Charles was good friends with several government ministers. He had also sat on numerous advisory committees over the years and had already been awarded a CBE for his services to the British economy.
‘If we’re unable to do that, I think we can still build a very lucrative income stream selling this to law enforcement, security companies, and so on. Or even high-net-worth individuals who might be willing to pay large sums to access a tool like this.’
‘What for?’
‘I can think of a hundred reasons. You might want to identify a face at a protest, a troublemaker on your property. Or maybe—’ He laughed. ‘You spot a pretty girl on the street and want to know who she is. All you have to do is take her photo.’
Charles stared at him, and for a moment Zack thought he might have gone too far. After all, when you said it out loud it did sound a little creepy. But then Charles said, ‘Interesting.’
Charles scrolled through the results for Bindiya Patil again. He paused at the bottom of the screen.
‘There’s a list of similar matches, too?’
‘Yes. People who we might think of as Bindiya’s lookalikes.
I’ve been reading up on this because when I saw these “similar matches” it made me wonder if I might have a double out there myself.
Apparently, you only have something like a one in a trillion chance of finding two exact matches because there are so many variables in the human face.
They have to be the same age as you, too, which lengthens the odds.
But most of us have someone out there, of the same age, who is almost identical.
As close to a doppelg?nger as you’re ever going to get.
I found mine. He lives in Sydney. If I ever need to swap identities with someone, I reckon we could fool pretty much everyone. Except this app, of course.’
‘Have you got a name for it?’
‘I was thinking Fase.’ He spelled out the acronym. ‘Facial Analysis Search Engine.’ He paused. ‘You don’t like it? I can get one of the creatives to work on it.’
He waited for Charles’s reaction. Again, Charles seemed lost in thought. He was beginning to think his boss had started spontaneously meditating when Charles said, ‘This is all fascinating. Extremely fascinating.’
That didn’t sound like a full commitment. Hurriedly, Zack said, ‘I’ll do all the costings and a forecast, write up a full report and get it to you to look at over the break.’
‘Yes.’ Charles’s voice had taken on an uncharacteristically dreamy quality. ‘You do that.’
He left the office without another word, leaving Zack feeling oddly deflated.
Was Charles worried about the privacy concerns?
Was it his usual conservative scepticism about moving the company into software?
Whatever it was, Zack vowed to do everything he could to convince Charles this had to be pursued.
He needn’t have worried. Because the next time he met with Charles, in early January, Charles had come to see the full potential of Fase.
And there was something he wanted to try.