WEBVTT - Mustafa Suleyman Isn’t Like Everyone Else in Silicon Valley 0:00:02.720 --> 0:00:07.200 Bloomberg Audio Studios, podcasts, radio news. 0:00:08.600 --> 0:00:11.320 At some point over the next twenty or thirty years, 0:00:11.480 --> 0:00:13.800 machines are going to be more capable than humans are 0:00:13.800 --> 0:00:16.520 doing most work, and so we have to decide as 0:00:16.520 --> 0:00:18.880 a society what our purpose is. 0:00:20.600 --> 0:00:24.400 What's the first solemn? The AI founder who is now 0:00:24.440 --> 0:00:27.040 at the forefront of a global race. 0:00:27.000 --> 0:00:29.240 Getting the balance right is very important. We have to 0:00:29.280 --> 0:00:32.760 deliver in the next few years. We're taking a huge 0:00:32.800 --> 0:00:34.360 bet that we're going to be able to convert this 0:00:34.479 --> 0:00:37.639 into true intelligence, and if we do, then I think 0:00:37.680 --> 0:00:39.600 the world is going to look very, very different. 0:00:39.760 --> 0:00:43.200 We will have abundant intelligence. On tap. 0:00:44.800 --> 0:00:49.840 From Bloomberg Weekend. This is the Michelle Hussein Show. I'm 0:00:49.840 --> 0:00:58.520 Michelle Hussein. Back in twenty sixteen, a computer beat a 0:00:58.600 --> 0:01:03.200 world champion in the ancient game Go. You might remember 0:01:03.240 --> 0:01:05.600 it because it certainly made the news of the time. 0:01:06.080 --> 0:01:10.399 It was a huge milestone artificial intelligence, which few of 0:01:10.480 --> 0:01:14.640 us had ever really heard about, was uncovering strategies that 0:01:14.840 --> 0:01:19.600 hadn't been seen over centuries of human players. One of 0:01:19.640 --> 0:01:23.000 those who made it happen at the startup Deep Mind, 0:01:23.520 --> 0:01:29.160 is now leading AI for Microsoft, but Mustapha Suleiman is 0:01:29.200 --> 0:01:33.679 an unusual figure in big tech. He's British for starters, 0:01:34.200 --> 0:01:38.600 He's interested in technology and public services. He used to 0:01:38.600 --> 0:01:42.240 work in local government in the UK. He's seen climate 0:01:42.319 --> 0:01:46.640 talks up close from the inside. Today though his work 0:01:46.720 --> 0:01:49.640 is at the heart of an AI arms race, the 0:01:49.680 --> 0:01:53.920 big names spending huge sums and making all sorts of 0:01:53.960 --> 0:01:58.280 promises on what it can do and when, which is 0:01:58.360 --> 0:02:02.200 why I'd say he's really worth a listen, not only 0:02:02.240 --> 0:02:06.000 to understand what's happening right now in this crucial field, 0:02:06.560 --> 0:02:10.359 but also to understand one of its most influential players. 0:02:11.280 --> 0:02:16.000 So here's my conversation with mustafas Leman, starting with his 0:02:16.080 --> 0:02:18.480 own everyday use of AI. 0:02:21.360 --> 0:02:25.320 Yesterday night, I stayed up far too late watching a film, 0:02:25.480 --> 0:02:29.800 and afterwards I added to a table that I've made 0:02:29.800 --> 0:02:33.200 in co pilot, which basically records all the films that 0:02:33.280 --> 0:02:36.000 I love, films that I want to watch, lists them 0:02:36.040 --> 0:02:39.560 by date, describes what they were about. I add my 0:02:39.600 --> 0:02:41.760 personal notes to it, gives me a link to the 0:02:41.760 --> 0:02:44.760 film poster, and I can keep just saying, okay, to 0:02:44.880 --> 0:02:46.480 add this film that I want to watch, what would 0:02:46.480 --> 0:02:47.760 be a similar one that I want to watch? 0:02:47.800 --> 0:02:49.880 So, just like you might ask. 0:02:49.680 --> 0:02:53.080 An assistant to organize your life, and I think that 0:02:53.240 --> 0:02:56.720 people will be surprised. The more obscure, the more creative, 0:02:57.240 --> 0:03:00.000 the more challenging you think the task is that you're 0:03:00.080 --> 0:03:02.400 going to ask your AI or co pilot, the better. 0:03:02.720 --> 0:03:06.280 But have you used it to do autonomous tasks? So, 0:03:06.440 --> 0:03:10.160 for example, has it booked tickets for you or bought 0:03:10.200 --> 0:03:12.359 a gift for you? Because I know this is one 0:03:12.400 --> 0:03:14.840 of the promises of co pilot, It's just not available 0:03:14.840 --> 0:03:16.480 in my region, so I haven't been able to try 0:03:16.520 --> 0:03:17.560 it out for myself yet. 0:03:18.240 --> 0:03:20.080 We're still experimenting with it. 0:03:20.080 --> 0:03:22.520 It can do it, it doesn't always get it right, 0:03:22.680 --> 0:03:25.520 so it's in like a sort of dev mode, so 0:03:25.560 --> 0:03:29.000 it's not like generally available just yet. But when it 0:03:29.040 --> 0:03:31.720 does work, it is the most magical thing you've ever seen. 0:03:31.840 --> 0:03:35.360 Okay, what are the mistakes it's made that have created 0:03:35.440 --> 0:03:38.480 problems for you? Bord a present for the wrong person, Well. 0:03:38.320 --> 0:03:40.680 It can buy the wrong thing, but obviously you can 0:03:40.720 --> 0:03:43.200 intervene and it will always ask you permission before it 0:03:43.200 --> 0:03:45.760 takes the next action and stuff, so it's it's quite safe. 0:03:46.120 --> 0:03:48.280 It's one of those things with technology, you know it 0:03:48.400 --> 0:03:52.120 promises so much and it's always sort of just around 0:03:52.120 --> 0:03:52.720 the corner. 0:03:52.960 --> 0:03:54.760 I remember, like the allure of. 0:03:54.680 --> 0:03:58.240 Getting a discman back in the nineties, and I had 0:03:58.280 --> 0:04:00.480 this dream that I was going to be roll blading 0:04:00.520 --> 0:04:03.640 down the street, and then I got one, and obviously 0:04:03.720 --> 0:04:06.960 it didn't have any kind of vibration proof shop proof 0:04:07.160 --> 0:04:10.000 technology yet, and so it would skip every time, like 0:04:10.040 --> 0:04:12.360 I turned it on its side or tried to put 0:04:12.400 --> 0:04:13.360 it in my back pocket. 0:04:13.640 --> 0:04:15.880 And so it's a funny thing. Technology. 0:04:15.920 --> 0:04:19.200 It's still magical and amazing, but it's always just got 0:04:19.240 --> 0:04:21.039 a little bit further to go. And I think in 0:04:21.120 --> 0:04:23.560 this case, it's a while yet before it's every day. 0:04:24.360 --> 0:04:27.840 And I guess that given everything you've seen, you're one 0:04:27.880 --> 0:04:30.760 of the founders of deep Mind, and you've worked at 0:04:30.880 --> 0:04:34.599 Google and had your own AI company inflection before you 0:04:34.680 --> 0:04:36.839 came to Microsoft. Does that mean that when you see 0:04:36.880 --> 0:04:40.760 these hiccups or these stages of development, you have faith. 0:04:40.800 --> 0:04:43.799 You don't get frustrated because you remember the moment when 0:04:44.360 --> 0:04:48.479 back in twenty sixteen when your AI model beat the 0:04:48.480 --> 0:04:51.359 world champion at go, which was a huge moment for AI. 0:04:51.960 --> 0:04:54.440 I think I'm very stoic about these things, like I 0:04:54.600 --> 0:04:56.520 know that it's going to work in the next six 0:04:56.600 --> 0:04:59.159 months or twelve months, or maybe the worst case, eighteen months. 0:04:59.200 --> 0:05:01.720 But you know, I think we've got a pretty good 0:05:02.240 --> 0:05:05.640 rhythm now of picking things which are not too far 0:05:05.720 --> 0:05:08.920 away but are genuinely transformative. And also, you know, it's 0:05:08.960 --> 0:05:12.840 important to remember how magical it is today. It can 0:05:12.880 --> 0:05:16.560 generate images, it can generate videos. It is already you know, 0:05:17.000 --> 0:05:20.400 sort of superhuman magical, And I think I'm just one 0:05:20.440 --> 0:05:22.520 of those people that's still inspired by what it can 0:05:22.560 --> 0:05:24.000 do at any given moment. 0:05:24.240 --> 0:05:26.000 By this time next year, could I be buying my 0:05:26.080 --> 0:05:28.920 Christmas presents using an autonomous AI agent. 0:05:29.400 --> 0:05:31.840 I'm pretty certain that you will be. Yeah, I think 0:05:31.880 --> 0:05:33.160 it's highly highly likely. 0:05:33.520 --> 0:05:37.279 So, speaking of magical things, the recent announcement that you 0:05:37.400 --> 0:05:41.240 made at Microsoft about superintelligence. That's the term which has 0:05:41.320 --> 0:05:44.320 really crept into the public debate thanks to you and 0:05:44.520 --> 0:05:47.800 others in the last few months. When you use the 0:05:47.880 --> 0:05:49.800 term superintelligence, what does it mean to you? 0:05:50.720 --> 0:05:55.080 Super intelligent in the industry today means an AI system 0:05:55.600 --> 0:05:59.320 that can learn any new task and perform better than 0:05:59.360 --> 0:06:02.440 all human combined at all tasks. 0:06:02.839 --> 0:06:04.160 So it is a very. 0:06:04.040 --> 0:06:07.960 High bar and at the moment it comes with a 0:06:08.000 --> 0:06:11.360 great deal of risk. It is not clear. In fact, 0:06:11.400 --> 0:06:14.960 it's very uncertain how we would limit its capabilities and 0:06:15.080 --> 0:06:18.200 align to get a system like that to care a 0:06:18.240 --> 0:06:21.200 lot about humans and want us to succeed and flourish. 0:06:21.400 --> 0:06:23.719 It's very unclear how we would contain and aligned a 0:06:23.760 --> 0:06:26.280 system that is so much more powerful than us. And 0:06:26.320 --> 0:06:29.400 so the framing that I prefer is one of a 0:06:29.480 --> 0:06:34.000 humanist superintelligence, one that is always in our corner, on 0:06:34.040 --> 0:06:37.599 our team, aligned to human interests, and until we can 0:06:37.760 --> 0:06:41.960 prove that it will remain safe, we won't continue to 0:06:42.040 --> 0:06:45.360 develop a system that has the potential to run away 0:06:45.360 --> 0:06:46.640 from us and to get out the box. 0:06:46.640 --> 0:06:48.640 And I think that that's an obvious thing to say. 0:06:49.040 --> 0:06:51.320 Everybody should agree to that, and yet I think it's 0:06:51.400 --> 0:06:53.360 kind of a novel position in the industry. 0:06:53.000 --> 0:06:53.520 At the moment. 0:06:53.760 --> 0:06:55.920 So is that how you're trying to set Microsoft apart? 0:06:56.000 --> 0:06:59.200 A lot of people are working on superintelligence. Are you 0:06:59.360 --> 0:07:02.080 trying to make my SELT distinctive by saying we will 0:07:02.120 --> 0:07:05.160 always use it through a humanist lens? 0:07:05.800 --> 0:07:08.360 Yes, that is our position. You know, Microsoft is a 0:07:08.400 --> 0:07:12.000 company that's been around for fifty years. It is very careful, 0:07:12.120 --> 0:07:15.200 it's highly trusted. Ninety percent of the s and P 0:07:15.320 --> 0:07:19.440 five hundred use us to provide email and operating systems 0:07:19.520 --> 0:07:22.720 and everyday productivity tools which are now among the best 0:07:22.760 --> 0:07:25.360 in the world with co Pilot, And you know, I 0:07:25.400 --> 0:07:28.160 think that we've got that reputation because the company has 0:07:28.160 --> 0:07:30.240 been careful and we're going to continue to be careful, 0:07:30.320 --> 0:07:33.760 and setting out a vision of humanist superintelligences is part 0:07:33.840 --> 0:07:34.600 of that program. 0:07:34.880 --> 0:07:37.880 So what does that mean for your rivals in that field, 0:07:37.960 --> 0:07:41.640 some of whom you work closely with, like open AI, 0:07:41.720 --> 0:07:43.560 which I'm sure we'll talk more about. But does it 0:07:43.640 --> 0:07:46.080 mean that they're the wild West and you're the moral ones. 0:07:46.360 --> 0:07:48.960 I think that everybody has to decide what they stand 0:07:49.000 --> 0:07:51.600 for and how they operate, and I don't I don't 0:07:51.600 --> 0:07:54.800 want to judge how they're operating right now. I don't 0:07:54.800 --> 0:07:58.600 see any evidence of large scale mass harm. I don't 0:07:58.640 --> 0:08:02.680 see any indication that these things are improving themselves or 0:08:02.720 --> 0:08:06.640 operating autonomously. But we all predict a time in the 0:08:06.640 --> 0:08:09.600 next five years, maybe in the next ten years, where 0:08:09.760 --> 0:08:13.520 these capabilities do start to emerge and systems like this 0:08:13.680 --> 0:08:16.640 could set their own goals, or they could improve their 0:08:16.640 --> 0:08:20.840 own code, or they could act autonomously. Those things are 0:08:21.160 --> 0:08:25.840 capabilities that I've clearly outlined as increasing the level of risk, 0:08:26.280 --> 0:08:29.440 and therefore we have to approach them with caution, with 0:08:29.560 --> 0:08:33.679 more transparency and audit, with more government engagement, and make 0:08:33.720 --> 0:08:37.520 proactive declarations about how close we are to those three capabilities. 0:08:37.559 --> 0:08:39.920 And does that mean that you won't be releasing a 0:08:39.960 --> 0:08:43.960 superintelligence tool until you're confident that it can be controlled. 0:08:44.320 --> 0:08:46.120 Yeah, I think that's right. I mean, I think for 0:08:46.200 --> 0:08:51.520 sure the containment and alignment unnecessary prerequisites their red lines, 0:08:51.800 --> 0:08:54.880 and I think everybody in the industry has to sign 0:08:54.960 --> 0:08:56.920 up to that idea. I mean, nobody wants to cause 0:08:56.960 --> 0:09:00.440 mass harmon fundamentally, even though we all disagree. Clearly, everybody 0:09:00.520 --> 0:09:04.040 is committed to the survival of our species and I 0:09:04.040 --> 0:09:06.840 would hope the kind of flourishing and well being of everybody. 0:09:07.200 --> 0:09:09.959 So that's the discussion that we're trying to push now 0:09:10.040 --> 0:09:13.600 and require everybody to ask themselves in the industry, are 0:09:13.640 --> 0:09:16.679 they building a humanist superintelligence? Does it really stand for 0:09:17.480 --> 0:09:18.840 people and for our species? 0:09:18.880 --> 0:09:19.200 First? 0:09:19.679 --> 0:09:22.040 Yeah, I'm just wondering how much confidence we can really 0:09:22.080 --> 0:09:24.679 have in that given that things do go wrong and 0:09:24.720 --> 0:09:28.360 I don't just mean AI, I mean humans do wrong. Things. 0:09:28.360 --> 0:09:30.800 All things just go wrong for other reasons. And the 0:09:30.840 --> 0:09:34.080 recent example that Microsoft had to deal with was having 0:09:34.120 --> 0:09:37.920 to cease and disable some services that were being used 0:09:37.960 --> 0:09:40.600 by the Israeli Ministry of Defense after reporting by The 0:09:40.640 --> 0:09:44.719 Guardian that suggested they could be being used for mass surveillance. 0:09:45.440 --> 0:09:47.240 That was very good reporting by the Guardian. 0:09:47.320 --> 0:09:49.800 We were very grateful for it, and as soon as 0:09:49.840 --> 0:09:52.480 we became aware of it, we actually made all the 0:09:52.520 --> 0:09:57.800 necessary changes, removed the IDF from those servers and I think, 0:09:57.960 --> 0:10:00.840 you know, they were clearly not in compliance with our 0:10:00.920 --> 0:10:04.320 terms of service, and there's an ongoing investigation internally. 0:10:04.800 --> 0:10:08.800 But I guess the broader point is that it's hard 0:10:08.800 --> 0:10:12.720 to have confidence of controls and checks and balances and uses. 0:10:13.559 --> 0:10:14.120 It is hard. 0:10:14.200 --> 0:10:17.120 I mean, these are huge and complicated systems that they 0:10:17.160 --> 0:10:19.960 carry a lot of risk, and the most that we 0:10:20.040 --> 0:10:22.080 can do is make sure that we are auditing them 0:10:22.240 --> 0:10:25.600 and we're removing actors that violate our terms of service 0:10:25.600 --> 0:10:26.640 as quickly as possible. 0:10:26.960 --> 0:10:29.319 So are the first uses of superintelligence going to be 0:10:29.360 --> 0:10:30.320 in the medical field? 0:10:30.760 --> 0:10:31.200 I think so. 0:10:31.559 --> 0:10:34.040 I think that this is probably the most exciting application 0:10:34.120 --> 0:10:37.080 of superintelligence. I think we now have systems that can 0:10:37.200 --> 0:10:41.480 diagnose any rare condition that is found in the literature 0:10:41.920 --> 0:10:45.800 significantly better than human performance, and do so with fewer 0:10:45.840 --> 0:10:48.400 tests and interventions. And that's the key thing. It's like, 0:10:48.520 --> 0:10:53.160 often it's actually quite expensive to run medical intervention tests 0:10:53.200 --> 0:10:55.640 before you get to a diagnosis, and often they can 0:10:55.679 --> 0:10:57.920 be unnecessary to and they take time, and then in 0:10:57.960 --> 0:11:01.160 that process, you know the condition came off wor in 0:11:01.200 --> 0:11:04.720 our case, our system does this significantly more cheaply, with 0:11:04.760 --> 0:11:07.000 fewer tests and with higher accuracy. 0:11:07.360 --> 0:11:08.240 We're just putting it. 0:11:08.200 --> 0:11:11.440 Through independent peer review at the moment, and soon there'll 0:11:11.440 --> 0:11:12.840 be clinical trials of it too. 0:11:12.960 --> 0:11:15.720 So I think this is very, very, very exciting. 0:11:16.160 --> 0:11:20.600 And did you push for that focus yourself, because you've 0:11:20.600 --> 0:11:25.480 been running Microsoft AI for a while, about eighteen months now. 0:11:25.800 --> 0:11:29.040 Super Intelligence was a big recent announcement, and I'm conscious 0:11:29.080 --> 0:11:30.800 as I hear you talk about it that your mother 0:11:31.120 --> 0:11:35.720 was a nurse and you, unusually for senior people in tech, 0:11:35.760 --> 0:11:37.480 you've worked in the public sector as well. 0:11:38.040 --> 0:11:40.280 Yeah, I mean this is an area that's very important 0:11:40.320 --> 0:11:40.480 to me. 0:11:40.520 --> 0:11:43.160 Obviously, my mum was in an HS nurse and you know, 0:11:43.200 --> 0:11:46.320 I'm just a big believer that technology is here to 0:11:46.360 --> 0:11:49.640 serve us. It should make our lives better, should make 0:11:49.720 --> 0:11:52.240 us more comfortable one day. I think it is going 0:11:52.320 --> 0:11:54.679 to help us to live longer, give us the option 0:11:54.760 --> 0:11:57.120 to work less if we choose to. I do think 0:11:57.160 --> 0:11:59.200 it's going to produce abundance, and I think that we 0:11:59.280 --> 0:12:01.880 have to make conscious decisions to use it for those 0:12:01.920 --> 0:12:05.400 applications first, and that that's why, you know, even back 0:12:05.440 --> 0:12:08.400 at Deep Mind, I ran our health unit, and you know, 0:12:08.600 --> 0:12:09.760 I'm very passionate about this. 0:12:10.200 --> 0:12:13.880 Abundance is such a big promise. Tell me what you 0:12:14.240 --> 0:12:16.880 mean by it, because most of the time we hear 0:12:16.880 --> 0:12:20.160 about AI destroying jobs. But are you saying that the 0:12:20.320 --> 0:12:24.840 work will be done by AI increasingly and therefore we 0:12:25.200 --> 0:12:27.240 humans won't have to work as much. 0:12:27.760 --> 0:12:30.440 I mean, I think that it is inevitable that at 0:12:30.440 --> 0:12:33.880 some point over the next twenty or thirty years, machines 0:12:33.920 --> 0:12:36.000 are going to be more capable than humans are doing 0:12:36.080 --> 0:12:40.040 most work, and that might come much sooner. And so 0:12:40.160 --> 0:12:43.840 we have to decide as a society what our purpose is, 0:12:44.320 --> 0:12:46.920 and we have to be very thoughtful about the rate 0:12:46.960 --> 0:12:50.000 of introduction of new machines because we have to make 0:12:50.040 --> 0:12:54.480 sure that that rate of displacement is counterbalanced with a 0:12:54.559 --> 0:12:58.160 mechanism to fund people and to support people through a 0:12:58.200 --> 0:13:01.920 massive transition, which for which there is probably not going 0:13:01.960 --> 0:13:03.719 to be other jobs on the other side that are 0:13:03.760 --> 0:13:06.960 more competitive than what a machine can do. And so 0:13:07.040 --> 0:13:09.120 this is a real big step change in what it 0:13:09.200 --> 0:13:10.760 is to organize society. 0:13:11.240 --> 0:13:13.160 But do you believe in the idea of a universal 0:13:13.200 --> 0:13:17.400 basic income that that's what AI being very productive making 0:13:17.400 --> 0:13:20.080 the economy more productive could unlock. 0:13:20.679 --> 0:13:22.480 Yeah, I mean, I think I've long been on record 0:13:22.600 --> 0:13:26.880 saying that I think that is both inevitable and very desirable. 0:13:27.280 --> 0:13:29.319 We already live in a world of abundance, it's just 0:13:29.400 --> 0:13:32.760 poorly distributed. We already have more calories to consume, we 0:13:32.800 --> 0:13:34.559 have an incredible amount of food wastage. 0:13:34.600 --> 0:13:36.160 It's just unevenly distributed. 0:13:36.600 --> 0:13:39.880 And now that value isn't just manifested in atoms like 0:13:40.040 --> 0:13:44.400 food and cars and physical things. It's manifested in digital 0:13:44.440 --> 0:13:49.360 goods ideas, knowledge intelligence. And that's actually great news because 0:13:49.400 --> 0:13:53.240 that can proliferate, it can spread extremely quickly around the 0:13:53.400 --> 0:13:56.120 entire world. I mean, these llms and chatbots have been 0:13:56.160 --> 0:13:59.760 the fastest spreading technology in history, basically two billion and 0:13:59.760 --> 0:14:03.560 a users in the space of three years, and that's 0:14:03.559 --> 0:14:07.640 actually amazing because it's delivering knowledge to every single person 0:14:07.640 --> 0:14:09.880 who wants to use it on their phone in the 0:14:09.920 --> 0:14:13.120 current sort of format and device they already use. So 0:14:13.360 --> 0:14:16.360 I think that we have no problem with getting technology 0:14:16.400 --> 0:14:18.560 out and available to everybody because there's going to be 0:14:18.640 --> 0:14:22.760 massive competitive forces to reduce the cost of experiencing an AI. 0:14:23.360 --> 0:14:25.600 The challenge we're going to have to figure out is 0:14:25.840 --> 0:14:29.520 how we tax and redistribute so that the transition is 0:14:29.560 --> 0:14:30.440 a healthy one. 0:14:31.000 --> 0:14:34.240 Tell me more about where these ideas and your beliefs 0:14:34.720 --> 0:14:38.160 come from. Your mother was a nurse. But the family 0:14:38.240 --> 0:14:41.560 you grew up in in London in the nineteen eighties 0:14:41.640 --> 0:14:43.720 and nineteen nineties, how would you describe it? 0:14:44.280 --> 0:14:47.040 Preally working class there. My dad was a mini cab driver, 0:14:47.480 --> 0:14:48.240 my mom works in. 0:14:48.400 --> 0:14:52.520 HS and yeah, I mean we were fairly regular, like 0:14:52.600 --> 0:14:57.120 kind of unremarkable. My parents didn't supervalue education stuff. They 0:14:57.160 --> 0:14:59.360 always tho actually go get trade and my mom would 0:14:59.360 --> 0:15:00.680 often say to me, you know, you should be a 0:15:00.680 --> 0:15:04.400 carpenter electrician, you know, leave scot sixteen and stuff. So 0:15:04.800 --> 0:15:06.920 I don't know, it comes from a place of just 0:15:07.480 --> 0:15:10.240 experiencing the kind of rougher end of things a little 0:15:10.280 --> 0:15:13.520 bit and then having just a desire to try to 0:15:14.520 --> 0:15:16.080 do the best that we can do with the short 0:15:16.120 --> 0:15:16.640 life that we have. 0:15:17.080 --> 0:15:20.120 Your dad had come to the UK from Syria. Did 0:15:20.120 --> 0:15:21.920 your parents meet in England? 0:15:22.720 --> 0:15:23.840 Yes, yeah, that's right. 0:15:24.440 --> 0:15:27.280 And what happened when you were sixteen must have had 0:15:27.320 --> 0:15:30.440 because I read that when your parents split up, you 0:15:30.920 --> 0:15:34.720 and your brother, your younger brother, were pretty much left 0:15:34.720 --> 0:15:36.000 to fend for themselves. 0:15:36.960 --> 0:15:37.960 Ah, that's true. 0:15:38.000 --> 0:15:40.320 Yeah, I don't know where you've read that, but yeah, 0:15:40.800 --> 0:15:42.720 me and my younger brother did live on our own 0:15:42.760 --> 0:15:43.440 for a few years. 0:15:43.520 --> 0:15:43.760 Yeah. 0:15:43.920 --> 0:15:46.600 Yeah, I think you mentioned this in a Guardian interview 0:15:46.600 --> 0:15:48.520 a couple of years ago when your book was coming out, 0:15:48.520 --> 0:15:51.360 But I'm really curious to know more about what it 0:15:51.520 --> 0:15:53.560 taught you and how you dealt with it at the time. 0:15:54.560 --> 0:15:56.640 I think that when you're that age, you're just sort 0:15:56.640 --> 0:16:00.200 of precocious and overconfident and fearless. We had everything we 0:16:00.280 --> 0:16:02.520 needed and it worked out pretty well. 0:16:02.640 --> 0:16:04.640 And this isn't a crucial time the last couple of 0:16:04.680 --> 0:16:08.400 years of school, and you did do well in those 0:16:08.480 --> 0:16:10.360 last couple of years of school because you ended up 0:16:10.680 --> 0:16:11.880 getting to Oxford. 0:16:12.240 --> 0:16:12.440 Yeah. 0:16:12.480 --> 0:16:15.680 I was very lucky because I had great teachers and 0:16:16.000 --> 0:16:18.280 good mentors, and I went to a very good school 0:16:18.600 --> 0:16:21.760 which was far from my home, but you could do 0:16:21.960 --> 0:16:24.280 exams to get in, so I studied really hard for 0:16:24.280 --> 0:16:26.800 when I was ten years old for the entrance exams, 0:16:26.800 --> 0:16:29.600 and it was essentially like going to a private school. 0:16:29.760 --> 0:16:32.880 Basically everyone was like very studious and got great grades 0:16:32.920 --> 0:16:36.000 and stuff. And yeah, I was lucky enough to get 0:16:36.000 --> 0:16:38.720 into Oxford, and you know, that was an amazing experience. 0:16:39.400 --> 0:16:41.080 Have you thought about what your life would have been 0:16:41.400 --> 0:16:43.760 like in that period in your mid teens had it 0:16:43.800 --> 0:16:45.080 not been for that school. 0:16:45.960 --> 0:16:48.440 Yeah, I haven't thought much about that. It would have 0:16:48.480 --> 0:16:49.680 been pretty different. 0:16:50.120 --> 0:16:53.800 And then you got to Oxford and you did actually 0:16:53.800 --> 0:16:55.840 subject pretty close to what my son is doing there 0:16:55.920 --> 0:16:58.720 right now. He's doing theology. You did philosophy and theology, 0:16:59.040 --> 0:17:01.840 and then tragedy cle to my ears as a parent, 0:17:01.880 --> 0:17:02.680 you dropped out. 0:17:03.240 --> 0:17:06.600 Yeah, well, I was very frustrated and eager to change 0:17:06.600 --> 0:17:09.800 the world and get stuff done, and so I actually 0:17:09.880 --> 0:17:13.400 dropped out. And I helped to start a telephone counseling 0:17:13.440 --> 0:17:17.080 service for young British Muslims called Muslim Youth Helpline, which 0:17:17.119 --> 0:17:19.920 I worked on for two or three years as a volunteer. 0:17:20.520 --> 0:17:24.040 And it was a secular peer to peer listing service 0:17:24.160 --> 0:17:26.800 for young British Muslims that post nine to eleven were 0:17:27.119 --> 0:17:30.120 dealing with a lot of I guess identity crisis would 0:17:30.160 --> 0:17:31.960 be the way that we would describe it in hindsight, 0:17:32.000 --> 0:17:34.080 but at the time it was just like, you know, 0:17:34.240 --> 0:17:37.639 lack of connection to community, family, parents, a lot of 0:17:37.680 --> 0:17:40.520 bullying and so yeah, I worked on that pretty much 0:17:40.560 --> 0:17:42.200 full time for two or three years. 0:17:42.480 --> 0:17:46.800 And had you experienced anti Muslim sentiment and hatred yourself 0:17:46.840 --> 0:17:49.640 at that time given the period we're talking about after 0:17:49.720 --> 0:17:50.640 nine to eleven. 0:17:51.080 --> 0:17:53.600 A little bit, Yeah, I think it was more just 0:17:53.920 --> 0:17:55.360 not so much for me, but I think a lot 0:17:55.359 --> 0:17:58.040 of people felt that, you know, we weren't sort of 0:17:58.040 --> 0:18:01.359 British enough, and people were figuring out how to both 0:18:01.560 --> 0:18:06.560 live their cultural religious identities in the context of families 0:18:06.560 --> 0:18:09.639 that often were first generation, didn't really speak the language, 0:18:09.680 --> 0:18:12.399 didn't know how to navigate the system, and then this 0:18:12.600 --> 0:18:15.840 increased kind of skepticism that we were sort of mostly 0:18:15.960 --> 0:18:17.000 terrorists or whatever. 0:18:17.119 --> 0:18:19.040 And I think that was just this general. 0:18:18.720 --> 0:18:22.280 Kind of fear and exclusion, and actually most of the 0:18:22.320 --> 0:18:25.760 time that was dealt with by just having kind and 0:18:25.800 --> 0:18:29.600 supportive conversation and having somebody be available on the other 0:18:29.720 --> 0:18:30.480 end of a telephone. 0:18:30.640 --> 0:18:32.240 Yeah, someone to talk to. I know the work of 0:18:32.280 --> 0:18:34.879 the helpline actually, so I know exactly the world that 0:18:34.920 --> 0:18:37.679 you're talking about. Did you think hard about dropping out 0:18:37.720 --> 0:18:41.920 of university and have you regretted it at any point? 0:18:41.960 --> 0:18:45.399 Since I haven't regretted it, and I probably could have 0:18:45.480 --> 0:18:48.439 thought harder about it. My tutor did encourage me to 0:18:48.480 --> 0:18:51.960 like switch subjects or change college and actually had me 0:18:52.040 --> 0:18:54.840 go meet one of his friends in a different college 0:18:54.880 --> 0:18:57.399 to go and talk about doing history or PP. But 0:18:57.640 --> 0:19:00.679 that point I was just so full of energy and 0:19:00.760 --> 0:19:03.119 like eager to start things. I started a business at 0:19:03.160 --> 0:19:07.040 the same time, sort of PDAs into restaurants to try 0:19:07.040 --> 0:19:10.439 and do restaurant booking and stuff and doing their networking. 0:19:09.960 --> 0:19:10.879 And so on. 0:19:10.920 --> 0:19:12.560 It was our very early days of the internet, so 0:19:12.600 --> 0:19:14.080 it was a bit early, but I did that for 0:19:14.400 --> 0:19:16.680 about a year or two as well. I had too 0:19:16.720 --> 0:19:19.440 much energy, I think for sitting around thinking in the 0:19:19.520 --> 0:19:20.600 dreamy spires. 0:19:37.680 --> 0:19:41.480 Now Here you are in this circle of power, right, 0:19:41.640 --> 0:19:43.960 one of a small group of people who are making 0:19:44.000 --> 0:19:47.280 decisions on a technology that is already changing all our 0:19:47.320 --> 0:19:50.280 lives and will continue to do. So, how conscious are 0:19:50.320 --> 0:19:51.159 you of that power? 0:19:52.240 --> 0:19:52.840 Very conscious? 0:19:52.880 --> 0:19:56.240 I take it very seriously and it's a great responsibility. 0:19:56.600 --> 0:19:59.080 And you know, I think that this is a moment 0:19:59.200 --> 0:20:03.560 when decisions that we make may have very lasting consequences. 0:20:03.600 --> 0:20:06.840 And I read history and I look back at times, 0:20:07.520 --> 0:20:11.000 for example, in the social media revolution, where I think 0:20:11.040 --> 0:20:13.199 it was very clear that there were potential harms that 0:20:13.280 --> 0:20:15.800 maybe fell on deaf fears for too long. Or I 0:20:15.840 --> 0:20:18.520 look back at you know, the electric car revolution that 0:20:18.560 --> 0:20:21.560 could have taken off way back in the twenties, or 0:20:21.680 --> 0:20:24.639