Can machines think? It’s a question on Sir Tim Berners-Lee’s mind. A man clearly ahead of his time, the scientist, academic and inventor of the World Wide Web, is a visionary, whose innovative work has transformed businesses, governments and the world around us. Eccentric, yet immensely passionate about technology and the future of artificial intelligence (AI), Berners-Lee wants brands and consumers to prepare for what’s coming: a world in which AI touches almost every aspect of our lives.
The conversation between Berners-Lee and Atifa Silk, Campaign Asia-Pacific brand director, takes place in Cannes, where Berners-Lee has just taken to the stage as part of the PHD session at the Lions’ Festival of Creativity. Surrounded by the advertising industry, he is intrigued by the impact of AI and theorist Alan Turing’s investigation into the power of a computer to rival human thought and the exciting potential — as well as the worrying implications — of it on brands and marketing. With more than US$27 billion invested in AI to date, according to artificial intelligence company Quid, there are secret developments worked on that have yet to be made public, he explains. And, the investment in AI will only continue to rise as uptake increases, thanks to the increasing sophistication of algorithms and speed of computers. As companies invest in talent, software and hardware, AI needs to be viewed as not just another piece of technology, but a new era in marketing. Because despite the benefits that AI could bring to the world, Berners-Lee reminds us to question what happens when the technology becomes too powerful, and think about whose best interests it will have in mind — and where we draw the line when (not if) AI develops human traits.
Atifa Silk: Has the definition of ‘Artificial Intelligence’ changed since Alan Turing worked on it over 100 years ago?
Sir Tim Berners-Lee: In a way, I don’t think the definition of AI has changed since Alan Turing said this is what is meant by artificial intelligence. AI is something that compiles the truth. It’s something that can pretend to be a human, or can convince a human being that it’s a human. That was Turing’s arbitrary chosen test. So, the idea of AI has been — and continues to be today — something that behaves like a human. That has stuck with us, even in recent films such as Ex Machina. That film is following the same, old idea we always have: How do you tell the difference between a machine and a human?
Atifa Silk: What are the ways in which AI is active today?
Sir Tim Berners-Lee: The quest to develop a computer to do everything that a person can do has been pushed considerably. It’s moved into vision, for example. So, we have computer vision recognising things or speech, for instance. There are ways that we use that vision for planning how to get across a city. There are lots of different programs, which are being produced as part of that. So while each one of them by itself isn’t an AI in the full sense, it’s still pretty handy. So, for instance, it’s useful to have a computer that can understand your speech. It’s useful that a computer can figure out how to get you across the city. Computers can help you look for patterns. They can use the data they have about you to know the advertising that would most appeal to you.
“It’s reasonable that we should all feel nervous. We should feel nervous about everything happening now, and about what else is coming”
Atifa Silk: How well do machines know us?
Sir Tim Berners-Lee: There are people who are shocked when, for example, they are looking for a hotel in an area and then for two months afterwards when they go onto their computer to search for something completely different, which has nothing to do with holidays or travel, they find there is an ad trying to sell them a holiday in the place they have just been or searched two months ago. People find that frustrating because, obviously, the computer’s got an idea into its head and won’t let it go. I once looked for a GPS online and there are so many people trying to sell these units for cars on the web. I wanted a reliable product so I had a look at a bunch of different brands and, honestly, it took forever. And, of course, nobody knows when you’ve actually bought one. The system is smart enough to know that you’ve looked for a hotel but it isn’t smart enough to know you’ve actually been there and done that. But when the advertising industry suddenly gets the bit between its teeth they want to sell you something, even when it’s something really inappropriate. That can frustrate people. And, they find it frightening because they realise that instead of looking at a magazine as just being a magazine put together by creative people, sticking stuff on a big piece of paper and then printing off lots of copies of it, suddenly it’s not a bunch of people producing a magazine, but a thing that has got ideas about them. And so the knowledge that this machine has an idea in his head about them becomes a bit upsetting.
Atifa Silk: Does the future of AI make you nervous?
Sir Tim Berners-Lee: It’s reasonable that we should all feel nervous. We should feel nervous about everything happening right now and what else is coming next. You should feel nervous about how information about you is used. In a way, advertising is one of the milder, less damaging, or potentially less damaging, ways a machine can use your data against you. I suppose it can be used to block access to things. Suppose it’s used to debar you from a club or supposing, given the things you often search for online, it decides it doesn’t want your sort of person in first class. So it doesn’t offer you first class tickets, not because you can’t afford the first class ticket price, but because it actually thinks you might be loud or just not the kind of person that the other first class passengers would be seen sitting next to. Just think your insurance gets cancelled because you’ve been searching for cancer online too much. But, in fact, you were looking because a friend of a friend had some form of cancer. However, now the system suddenly decides that it’s worth sending you ads about cancer then also it can decide whether it’s worth increasing your insurance premiums, maybe blocking you from taking on a new insurance policy because they’re worried that you might have an existing condition. Cancer is expensive for an insurer and healthcare agents see that as being able to filter out people who have got it. Knowing who might have it would be very valuable. If we look at virtual personal assistants (VPAs), I find the idea of the VPA working for somebody else after you’ve taught your personal assistant everything about yourself, scary.
Atifa Silk: How can we protect our privacy in this world of AI?
Sir Tim Berners-Lee: If you want to stop people being worried about privacy, then you have to go online over the creation of a personal profile, or you put people completely in control of the data. One of the things we’re doing in our lab is building systems where people have their data stored, so when they run an app they have it installed in their own cloud, which they control. If other apps, brands or companies want access to that data they have to ask and they have to explain what they’re using it for and they can be held accountable for using it in that way. I believe we may see a personal revolution in the future where I control my data and I do great things with it. I do very powerful things with it. One of the things I can do is run programs that work out on my behalf what I need to buy and where I need to go on vacation, for example. I can have programs that optimise my point of view; they’re working on my behalf and in that way.
"If I have a virtual personal assistant...helping me do my shopping...then as a brand, you are not selling to me. You are essentially selling to the machine or my agent. Suddenly, that means you need to be good at data."
Atifa Silk: What should brands be doing today to prepare for this future?
Sir Tim Berners-Lee: If I have a virtual personal assistant and have somebody who is helping me do my shopping, for example, and they are working for me then they will be able to look at the web or data out there about different products and they will be able to compare them and prices and decide what I should choose. Now, if that’s the reality, then as a brand, you are not selling to me (the consumer). You are essentially selling to the machine or my agent. Suddenly, that means you need to be good at data. It means that you need to make sure that you have all your products and all the scripts are described in the data that the machine understands. Most of the work today is completely under wraps. We don’t see what companies are working on. AI development is pretty secretive.
Atifa Silk: What is your vision of the future?
Sir Tim Berners-Lee: Pixels. They are getting cheaper and so every time you go to a conference the stage has got more pixels on it and in an auditorium they may have 36 different video feeds. Bit by bit everything’s going to be covered by pixels. This is a good and a bad thing. It means that you won’t have to choose the wallpaper because you can just program it. So, that’s cool. In your house it means that you’ll be able to sit and the video will be on the wall surrounding you. It will be so good that your eyes won’t be able to tell that you’re not actually sitting in your house.
Video quality, I believe, will get to the point where it gets past the retina level, as Apple calls it. It will get to a point where the video graph isn’t perfect but it’s better than your eyes can see anyway. It’s like sound stereo. Do you remember when hi-fi came out? It was cool and everybody got stereo sets and they competed for the best frequency response. Then it suddenly got to the point where actually everybody’s stereo collection was just better, on the human ear anyway. So, people stopped arguing about it because the ear couldn’t tell that well. Video will get to a point when we can’t tell that well and it will be wraparound. That I believe will be a really interesting medium.
Atifa Silk: How will pixels transform the world?
Sir Tim Berners-Lee: We’re already starting to see hoardings become programmable. Paper hoardings are becoming rare because what’s the point of having to hire people to change the paper over? Potentially, it could be very intrusive. Imagine if an entire wall or hoarding in the street is covered with pixels, which are programmed by some outside company. In this world, it might actually be difficult to see because everything on that street could look like a jungle with lions and tigers rushing through it or something. So, you can imagine that you may want to say, for the point of view of drivers driving around, that actually your ability to display stuff on the outside of your car should be limited to only the side or only the roof of the car, but you can imagine pixels everywhere. So every surface will be pixels and you’ll be able to use it for anything and people will be able to rent out space with pixel pieces on it. Just think of the bandwidth, the pixels and the videos that could be created for each of us. It will be fun and a challenge I suppose.
As computers become ever more powerful, the capabilities of artificial intelligence are accelerating. And as AI turbo-charges our information gathering and processing capabilities, marketing will naturally be at the forefront of the AI revolution.
We can already see glimpses of what the future may hold for marketing in today’s “virtual personal assistants”, such as Apple Siri, Google Now and Microsoft Cortana. These services already help consumers find information about brands and products, and soon they will automate even more of our lives. That will create new challenges and opportunities for brands.
Sentient Virtual Personal Assistants
Although you can use Siri to add an appointment to your calendar, or ask for directions to a specific location, more complex tasks are beyond the capabilities of today’s virtual personal assistants (or VPAs).
But new VPA systems from companies like Viv are being designed to handle many of the things we’d expect a human assistant to do, such as rescheduling flights or coordinating the best time for a meeting based on several different people’s schedules.
Many of the smallest details of our lives will be taken over by our VPAs. For example, your VPA will order more tooth paste for you when the tube is running low. It will also adapt by watching your behaviour. If it notices that you’ve been making more environmentally conscious decisions in general, it might recommend a new brand of laundry detergent, based on the ingredients, the next time you need more.
Adapting to the New World
With sentient VPAs making many of our purchasing choices for us, brands will have to adapt. Instead of marketing directly to consumers, brands will have to market to algorithms. There is, of course, already a precedent for this to some degree in the world of search engine optimisation.
For example, all the products in a grocery shop will have “tagged” information including not just name and price, but also the essential ingredients, where they were produced and under what environmental and ethical conditions, and how far it was shipped, all so that our VPAs can make informed decisions on our behalf.
This ‘tagged-up world’, will be a vital part of how brands optimise their offerings for the age of the sentient VPA. Marketers will need to think hard about what types of product data to tag-up, and what types of offers make the most sense in different contexts.
Matching offers to the demands of consumers will become akin to high frequency trading, as brand algorithms race to provide consumers with the right offer at the right time, based on real-time data.
Maintaining positive brand sentiment will also be crucial. Feedback from other customers will be one of the most important signals that sentient VPAs look at when deciding whether to recommend a product to you.
Of course, traditional direct to consumer advertising won’t go away. We’ll still be able to ask our VPAs to book a stay at a hotel that we read about, or to stock our refrigerator with a new brand of soda that we want to try.
AI will radically change what we do, and create huge new challenges in reaching consumers. But even if our messages will reach fewer people, they will, in the future, reach only the right people. Marketing will become more effective for brands, and less intrusive for customers.
PHD Worldwide explores the impact of AI in ‘Sentience: The Coming AI Revolution and the Implications for Marketing’
This article is part of the Campaign Innovate series, a collection of articles that examine the way innovation, startups and technology are affecting the advertising and marketing industry.
Campaign Asia-Pacific has also launched the Campaign Innovate competition, an event that aims to provide a platform for Asia-Pacific's startups to pitch to some of the world's biggest brands.