WEBVTT - AI Wants Your Life: Tech Boss Meredith Whittaker Says No

0:00:02.880 --> 0:00:08.120
<v Speaker 1>Bloomberg Audio Studios, podcasts, radio news.

0:00:09.720 --> 0:00:13.520
<v Speaker 2>We can't live a full, dignified life with the ability

0:00:13.560 --> 0:00:18.040
<v Speaker 2>to questioned authority, to test ideas, to blow off steam.

0:00:18.440 --> 0:00:22.160
<v Speaker 2>If we don't have private communication, you can't do military strategy,

0:00:22.200 --> 0:00:24.760
<v Speaker 2>you can't do human rights work, you can't do journalism

0:00:24.880 --> 0:00:25.440
<v Speaker 2>full stop.

0:00:26.920 --> 0:00:31.520
<v Speaker 1>Meredith Whittaker, tech boss and defender of privacy. I got

0:00:31.560 --> 0:00:33.080
<v Speaker 1>the impression that by the end of your time at

0:00:33.080 --> 0:00:35.400
<v Speaker 1>Google you were kind of a thorn in management side.

0:00:35.479 --> 0:00:39.640
<v Speaker 2>It wasn't Meredith versus Google. There is a round swell

0:00:39.800 --> 0:00:44.000
<v Speaker 2>of sentiment within tech now that is concerned with the

0:00:44.040 --> 0:00:47.840
<v Speaker 2>fundamental question, how do we make sure that we are

0:00:48.000 --> 0:00:52.960
<v Speaker 2>collectively governing this powerful technology so that those who would

0:00:53.000 --> 0:00:55.640
<v Speaker 2>do harm with it aren't those who are in control.

0:00:57.680 --> 0:01:02.560
<v Speaker 1>From Blueberg Weekend. This is the Michelle Hussein Show. I'm

0:01:02.600 --> 0:01:11.800
<v Speaker 1>Michelle Hussein. Whether it's social media, bands, AI and military strikes,

0:01:11.920 --> 0:01:16.480
<v Speaker 1>the destruction of jobs, debates about tech in our lives

0:01:16.720 --> 0:01:20.680
<v Speaker 1>are all around us every day. I'm old enough to

0:01:20.720 --> 0:01:24.000
<v Speaker 1>remember the nineties and the emergence of what today we

0:01:24.040 --> 0:01:29.360
<v Speaker 1>call big tech. Back then, the companies were perceived as benign.

0:01:29.520 --> 0:01:33.040
<v Speaker 1>They were full of important and exciting features that would

0:01:33.120 --> 0:01:36.080
<v Speaker 1>connect us to people around the world, and they were

0:01:36.200 --> 0:01:40.119
<v Speaker 1>pretty much free to use. Today the story is much

0:01:40.160 --> 0:01:44.600
<v Speaker 1>more complex, But even so, how many of us really

0:01:44.640 --> 0:01:48.760
<v Speaker 1>think hard enough about our choices in tech, the products

0:01:48.800 --> 0:01:53.440
<v Speaker 1>we use, and what the companies get out of us. Well,

0:01:53.480 --> 0:01:57.680
<v Speaker 1>this conversation is with Meredith Whitaker, a tech boss who

0:01:57.880 --> 0:02:02.680
<v Speaker 1>is different from her peers, partly because her product, the

0:02:02.760 --> 0:02:07.960
<v Speaker 1>messaging app Signal, is nonprofit and also it has a

0:02:08.040 --> 0:02:14.120
<v Speaker 1>deep and she says, unshakable promise to barely even collect

0:02:14.160 --> 0:02:18.520
<v Speaker 1>our data, let alone sell it. If you're not a

0:02:18.560 --> 0:02:21.720
<v Speaker 1>Signal user, but you know you've heard about it somewhere.

0:02:22.280 --> 0:02:25.639
<v Speaker 1>It's probably because of the moment last year when the

0:02:25.639 --> 0:02:30.160
<v Speaker 1>then US National Security Advisor added a journalist to what

0:02:30.280 --> 0:02:33.680
<v Speaker 1>should have been a classified group chat about military action

0:02:33.960 --> 0:02:39.160
<v Speaker 1>in Yemen. It became known, of course as Signalgate. So

0:02:39.280 --> 0:02:43.200
<v Speaker 1>that's some of the background. But as Meredith Whittaker came

0:02:43.200 --> 0:02:46.320
<v Speaker 1>into the studio, there was a live issue in the UK,

0:02:46.840 --> 0:02:50.600
<v Speaker 1>the government's plans to make it impossible for under eighteens

0:02:51.080 --> 0:02:55.239
<v Speaker 1>to take or share nude images on their phones. As

0:02:55.280 --> 0:02:58.800
<v Speaker 1>you'll hear, she's got views on that, both the principle

0:02:59.280 --> 0:03:03.200
<v Speaker 1>and whether the idea is even workable, but I had

0:03:03.320 --> 0:03:08.720
<v Speaker 1>questions too on whether everyone has a right to be private,

0:03:09.120 --> 0:03:14.960
<v Speaker 1>what about crime, abuse, terrorism? Ultimately, I hope you find

0:03:15.000 --> 0:03:19.040
<v Speaker 1>this as compelling as I did. Meredith Whittaker is able

0:03:19.080 --> 0:03:22.320
<v Speaker 1>to see the ironies in her world, to spot the

0:03:22.320 --> 0:03:25.880
<v Speaker 1>moments of humor even and at the same time make

0:03:25.960 --> 0:03:30.960
<v Speaker 1>us think. Obviously, I think some people listening to this

0:03:31.080 --> 0:03:33.320
<v Speaker 1>might be pretty new to Signal, but I think anything

0:03:33.360 --> 0:03:35.360
<v Speaker 1>that you can say to let us into your world

0:03:35.480 --> 0:03:38.760
<v Speaker 1>and to tell us about you as well as what

0:03:38.800 --> 0:03:40.480
<v Speaker 1>you do, I think it will be great to.

0:03:41.320 --> 0:03:43.400
<v Speaker 2>I'll do my best. I think Signal is a bit

0:03:43.440 --> 0:03:44.440
<v Speaker 2>more interesting than I am.

0:03:44.480 --> 0:03:49.720
<v Speaker 1>But Meredith, I guess I would just love to start

0:03:49.760 --> 0:03:54.000
<v Speaker 1>by asking about your own journey towards what you do now.

0:03:54.320 --> 0:03:56.800
<v Speaker 1>Because you worked in big tech at Google for more

0:03:56.840 --> 0:04:00.560
<v Speaker 1>than a decade, was there a moment in that period

0:04:00.600 --> 0:04:05.320
<v Speaker 1>where you really started to think differently about privacy and data,

0:04:05.480 --> 0:04:08.320
<v Speaker 1>a moment you can identify that's kind of critical in

0:04:08.360 --> 0:04:09.960
<v Speaker 1>the development of your story.

0:04:10.320 --> 0:04:12.920
<v Speaker 2>I have considered that question. Is there a point where

0:04:12.920 --> 0:04:17.320
<v Speaker 2>I tripped over something or a revelation happened? You know

0:04:17.640 --> 0:04:20.880
<v Speaker 2>Athena with birth from the forehead of Zeus and suddenly

0:04:20.920 --> 0:04:24.960
<v Speaker 2>there's insight. I don't think it was a moment. I

0:04:25.120 --> 0:04:29.120
<v Speaker 2>come out of an non traditional background for somebody working

0:04:29.240 --> 0:04:32.479
<v Speaker 2>in the tech industry. I studied literature, I said, rhetoric.

0:04:32.560 --> 0:04:35.760
<v Speaker 2>I was always an arts kid, and that's still where I,

0:04:35.839 --> 0:04:39.680
<v Speaker 2>as a human being, really live. I love it, and

0:04:39.720 --> 0:04:45.039
<v Speaker 2>so I was calibrated to perhaps ask different questions or

0:04:45.120 --> 0:04:48.880
<v Speaker 2>just you know, pursue different inquiries. But I think it

0:04:49.000 --> 0:04:54.279
<v Speaker 2>was it was over time beginning to understand Google's business models,

0:04:54.440 --> 0:05:00.799
<v Speaker 2>the political economic forces that were animating the company. After

0:05:00.839 --> 0:05:04.240
<v Speaker 2>coming in, as they say, a novice, I did not

0:05:04.600 --> 0:05:08.240
<v Speaker 2>pursue computer science in school. I was not interested in tech.

0:05:08.480 --> 0:05:11.600
<v Speaker 2>I did not get you know, a Commodore sixty four

0:05:11.720 --> 0:05:14.279
<v Speaker 2>when I was a kid, and everything took off from there.

0:05:14.320 --> 0:05:16.599
<v Speaker 2>And I loved computers right, which you know, the classic

0:05:16.680 --> 0:05:19.960
<v Speaker 2>origin story. And so I had to ask really really

0:05:20.000 --> 0:05:22.719
<v Speaker 2>fundamental questions. YEA, what is a server?

0:05:23.680 --> 0:05:25.159
<v Speaker 1>Would I be right in thinking of you as a

0:05:25.160 --> 0:05:28.640
<v Speaker 1>bit of a contrarian or maybe a rebel. I know

0:05:28.920 --> 0:05:30.880
<v Speaker 1>that there was a key moment in your time at

0:05:30.880 --> 0:05:34.800
<v Speaker 1>Google where you organized colleagues against a Pentagon contract that

0:05:34.839 --> 0:05:37.520
<v Speaker 1>you realized was being signed.

0:05:38.279 --> 0:05:42.360
<v Speaker 2>I wouldn't adopt those terms myself. You know, I am

0:05:42.440 --> 0:05:47.920
<v Speaker 2>somebody who thinks it's really important to maintain the dignity

0:05:47.960 --> 0:05:51.760
<v Speaker 2>of my analysis. Like if something is true, if there

0:05:51.839 --> 0:05:55.880
<v Speaker 2>is empirical evidence supporting a conclusion, if we can look

0:05:55.880 --> 0:05:58.839
<v Speaker 2>at the map and clearly see that there's a problem

0:05:58.839 --> 0:06:02.719
<v Speaker 2>on the road ahead, then I am not somebody who's

0:06:02.800 --> 0:06:07.760
<v Speaker 2>going to go along to get along with a convenient delusion.

0:06:08.040 --> 0:06:10.400
<v Speaker 2>Those things, once I see them, I am not going

0:06:10.440 --> 0:06:14.040
<v Speaker 2>to veer away or pretend that they don't exist, even

0:06:14.080 --> 0:06:17.640
<v Speaker 2>if I have to take steps that might be scary

0:06:17.920 --> 0:06:20.480
<v Speaker 2>or against the grain to do that.

0:06:20.600 --> 0:06:22.280
<v Speaker 1>So I got the impression that by the end of

0:06:22.320 --> 0:06:24.040
<v Speaker 1>your time at Google you were kind of a thorn

0:06:24.080 --> 0:06:28.320
<v Speaker 1>in management side. You were organizing walkouts on other issues,

0:06:28.400 --> 0:06:31.880
<v Speaker 1>ethics issues, and they saw you as a problem.

0:06:32.400 --> 0:06:36.760
<v Speaker 2>Well, people who would like to put the goals of

0:06:37.360 --> 0:06:44.039
<v Speaker 2>infinite growth, infinite revenue profit above the reality of the

0:06:44.120 --> 0:06:48.080
<v Speaker 2>harms or the problems or the choppy seas on the

0:06:48.160 --> 0:06:52.000
<v Speaker 2>road ahead don't like people who point that out. But

0:06:52.080 --> 0:06:53.799
<v Speaker 2>I do want to sort of frame that it wasn't

0:06:54.080 --> 0:07:01.200
<v Speaker 2>It wasn't Meredith versus Google. This was a groundswell from

0:07:01.760 --> 0:07:03.360
<v Speaker 2>thousands of people.

0:07:03.880 --> 0:07:06.599
<v Speaker 1>The workouts and the walkouts at the time where people

0:07:06.640 --> 0:07:09.640
<v Speaker 1>came out in big numbers too. We see protested different things.

0:07:09.760 --> 0:07:14.920
<v Speaker 2>Let's go today Deep mind employees just unionized, concerned about

0:07:14.920 --> 0:07:18.360
<v Speaker 2>the application of that technology for harmful purposes. You have

0:07:18.440 --> 0:07:22.520
<v Speaker 2>open AI employees beginning to organize or that's being reported

0:07:23.000 --> 0:07:26.480
<v Speaker 2>in preparation for the IPO. There is a sort of

0:07:26.600 --> 0:07:31.560
<v Speaker 2>groundswell of sentiment within tech now and outside of tech

0:07:31.600 --> 0:07:36.160
<v Speaker 2>that is concerned with the fundamental question that I don't

0:07:36.200 --> 0:07:39.800
<v Speaker 2>think can be avoided by anyone. How do we make

0:07:39.840 --> 0:07:45.920
<v Speaker 2>sure that we are collectively governing this powerful technology so

0:07:45.960 --> 0:07:49.040
<v Speaker 2>that those who would do harm with it aren't those

0:07:49.040 --> 0:07:51.440
<v Speaker 2>who are in control. How do we scrutinize it, how

0:07:51.440 --> 0:07:54.120
<v Speaker 2>do we put guardrails on it, How do we determine

0:07:54.280 --> 0:07:57.120
<v Speaker 2>where it's used, where it's not used, how it's developed,

0:07:57.360 --> 0:07:59.680
<v Speaker 2>and what the goals of that development are. Those are

0:07:59.720 --> 0:08:03.640
<v Speaker 2>fun mental question that every world leader, every government, every

0:08:03.640 --> 0:08:07.160
<v Speaker 2>community considering a data center needs to be looking at

0:08:07.600 --> 0:08:10.840
<v Speaker 2>and where the leverage for governance where that fulcrum exists

0:08:11.440 --> 0:08:14.960
<v Speaker 2>is an open question given the bipolar tech world we

0:08:15.000 --> 0:08:18.080
<v Speaker 2>live in and the volatile geopolitical environment, where.

0:08:18.520 --> 0:08:21.040
<v Speaker 1>The fact that the tech will seems to becoming ever

0:08:21.080 --> 0:08:23.080
<v Speaker 1>more powerful, not least with the amount of interest in

0:08:23.120 --> 0:08:25.720
<v Speaker 1>the IPOs. Absolutely, I want to take a moment to

0:08:25.760 --> 0:08:29.320
<v Speaker 1>really understand what you do at Signal. The core promise

0:08:29.960 --> 0:08:33.520
<v Speaker 1>is end to end encryption, which means only the sender

0:08:33.640 --> 0:08:37.000
<v Speaker 1>or receiver can see the actual content of the messages.

0:08:37.440 --> 0:08:40.880
<v Speaker 1>So how is that different from WhatsApp, which also has

0:08:40.920 --> 0:08:41.720
<v Speaker 1>the same promise.

0:08:42.040 --> 0:08:45.320
<v Speaker 2>Well, the core promise of Signal is that in the

0:08:45.400 --> 0:08:52.280
<v Speaker 2>digital world, where almost every micro action is surveilled, turned

0:08:52.280 --> 0:08:55.360
<v Speaker 2>into data, processed in models, sold to advertisers, sold to

0:08:55.440 --> 0:09:00.240
<v Speaker 2>data brokers, et cetera, et cetera, we preserve the fundamental

0:09:00.320 --> 0:09:04.840
<v Speaker 2>human right to private communication full stop, which has existed

0:09:04.880 --> 0:09:07.120
<v Speaker 2>for hundreds of thousands of years. That was the norm.

0:09:07.160 --> 0:09:09.880
<v Speaker 2>If you and I are talking, that conversation is gone

0:09:09.920 --> 0:09:13.280
<v Speaker 2>once we are done talking and part ways. And we

0:09:13.400 --> 0:09:17.240
<v Speaker 2>have moved into a world where that ground has shifted

0:09:17.320 --> 0:09:20.360
<v Speaker 2>dramatically out from under us, and that is no longer true.

0:09:20.440 --> 0:09:23.160
<v Speaker 2>And we think that it is considered a fundamental human

0:09:23.280 --> 0:09:26.360
<v Speaker 2>right for a very good reason. We can't live a full,

0:09:26.840 --> 0:09:31.360
<v Speaker 2>dignified life with the ability to questioned authority, to have

0:09:31.480 --> 0:09:36.000
<v Speaker 2>different relationships with people to test ideas, to blow off steam.

0:09:36.360 --> 0:09:40.080
<v Speaker 2>If we don't have private communication, you can't do military strategy,

0:09:40.120 --> 0:09:42.679
<v Speaker 2>you can't do human rights work, you can't do journalism

0:09:43.000 --> 0:09:48.520
<v Speaker 2>full stop. So Signal provides that in as robust away

0:09:48.600 --> 0:09:52.400
<v Speaker 2>as we can. We developed the encryption protocol, the Signal protocol,

0:09:52.920 --> 0:09:57.200
<v Speaker 2>that moved the field forward. In twenty thirteen, we introduced

0:09:57.240 --> 0:10:00.520
<v Speaker 2>that to the world, and that is the co of

0:10:00.559 --> 0:10:05.199
<v Speaker 2>our offering that protects the message content. On Signal, what

0:10:05.280 --> 0:10:08.560
<v Speaker 2>you send only goes between sender and receiver. Only the

0:10:08.600 --> 0:10:12.800
<v Speaker 2>people you're talking to know it. WhatsApp licenses that protocol

0:10:12.800 --> 0:10:16.520
<v Speaker 2>from us, and they also apply it to many of

0:10:16.559 --> 0:10:20.199
<v Speaker 2>their messages that are sent. What'sapp for business is excluded

0:10:20.200 --> 0:10:23.840
<v Speaker 2>from this in many cases, but they protect message content,

0:10:23.920 --> 0:10:27.320
<v Speaker 2>so what you type, what you say is to end encrypted.

0:10:27.600 --> 0:10:31.440
<v Speaker 2>They can't see it only the center and receiver. However,

0:10:32.040 --> 0:10:35.319
<v Speaker 2>you have to look at privacy beyond simply the message content,

0:10:35.800 --> 0:10:39.400
<v Speaker 2>and this is where there are very fundamental differences between

0:10:39.440 --> 0:10:43.640
<v Speaker 2>WhatsApp and Signal. Signal also applies encryption to protect what

0:10:43.880 --> 0:10:48.400
<v Speaker 2>is called metadata, but that is extremely intimate. It's extremely

0:10:48.440 --> 0:10:52.280
<v Speaker 2>revealing data. It's your profile, photo, your name, who you

0:10:52.440 --> 0:10:55.480
<v Speaker 2>talk to when you talk to them, who's in your groups,

0:10:55.880 --> 0:10:58.960
<v Speaker 2>how often you chat? When did you text your oncologists?

0:10:58.960 --> 0:11:00.960
<v Speaker 2>When did you text your things therapist, When did you

0:11:01.000 --> 0:11:03.920
<v Speaker 2>text your lawyer? When did you text someone or your

0:11:03.960 --> 0:11:07.920
<v Speaker 2>best friend? All of that information. So WhatsApp does collect that,

0:11:08.760 --> 0:11:11.800
<v Speaker 2>we don't. We go out of our way to collect

0:11:11.880 --> 0:11:14.839
<v Speaker 2>as close to no data as possible, and we make

0:11:14.880 --> 0:11:15.720
<v Speaker 2>that open source.

0:11:16.520 --> 0:11:20.840
<v Speaker 1>Is it really that revealing? The other data, the message data?

0:11:20.880 --> 0:11:24.720
<v Speaker 1>If someone can see that I wake up at six

0:11:24.760 --> 0:11:27.479
<v Speaker 1>in the morning and I send a lot of messages

0:11:27.720 --> 0:11:31.400
<v Speaker 1>at that time, and they might even know who I'm messaging,

0:11:31.480 --> 0:11:35.160
<v Speaker 1>my mother, my colleagues, or people and other time zones,

0:11:35.280 --> 0:11:37.760
<v Speaker 1>is that really so revealing of me?

0:11:38.280 --> 0:11:43.040
<v Speaker 2>Oh? It absolutely can be. Are you messaging your FBI

0:11:43.200 --> 0:11:45.960
<v Speaker 2>contact because you want to be a source. Are you

0:11:46.040 --> 0:11:51.319
<v Speaker 2>messaging a whistleblower attorney because you found corruption in your

0:11:51.320 --> 0:11:54.320
<v Speaker 2>workplace and want to report that. Are you messaging a

0:11:54.360 --> 0:11:58.080
<v Speaker 2>divorce lawyer? Are you messaging an oncologist? Yeah?

0:11:58.080 --> 0:12:01.960
<v Speaker 1>But what I mean is, again it's like WhatsApp, for example,

0:12:02.000 --> 0:12:05.320
<v Speaker 1>could see, oh that other person is an oncologist, or

0:12:05.320 --> 0:12:08.360
<v Speaker 1>that other person is in law enforcement or is a source.

0:12:08.400 --> 0:12:11.240
<v Speaker 1>But again they can't see beyond that what the actual

0:12:11.280 --> 0:12:12.760
<v Speaker 1>content is that I'm saying.

0:12:12.920 --> 0:12:16.440
<v Speaker 2>Well, no, not ye, assuming that's not on what's up

0:12:16.440 --> 0:12:20.640
<v Speaker 2>for business? However, I think we have to expand beyond

0:12:21.240 --> 0:12:26.559
<v Speaker 2>simply the application and look at what'sapp being owned by Meta,

0:12:26.640 --> 0:12:32.040
<v Speaker 2>the parent company Meta is a massive data collecting entity.

0:12:32.320 --> 0:12:37.080
<v Speaker 2>Surveillance advertising is the heart of their business model. So okay,

0:12:37.160 --> 0:12:41.280
<v Speaker 2>you message your oncologists. All right, let's join that data

0:12:41.760 --> 0:12:44.200
<v Speaker 2>with all of the other data we have in hand.

0:12:44.480 --> 0:12:48.520
<v Speaker 2>We buy credit data from data brokers, we have Facebook data,

0:12:48.559 --> 0:12:52.200
<v Speaker 2>we have Instagram data, and we have powerful AI models

0:12:52.240 --> 0:12:54.680
<v Speaker 2>that can find patterns and search across the.

0:12:54.679 --> 0:12:59.760
<v Speaker 1>State see ads for cancer drugs on Instagram? Is that

0:12:59.760 --> 0:13:03.839
<v Speaker 1>what suggesting let's feelings? Advertising is a very inflammatory way

0:13:03.880 --> 0:13:04.360
<v Speaker 1>to put it.

0:13:04.720 --> 0:13:08.360
<v Speaker 2>Well, I mean it's a technical descriptor, right, It's targeted

0:13:08.400 --> 0:13:12.240
<v Speaker 2>advertising based on the collection of data that reveals or

0:13:12.640 --> 0:13:17.440
<v Speaker 2>purportedly reveals certain qualities of you and what you might like.

0:13:17.960 --> 0:13:23.200
<v Speaker 2>Let's take another example. You're messaging a whistleblower attorney and

0:13:23.400 --> 0:13:26.720
<v Speaker 2>it's very high, high risk, and you know the corruption

0:13:26.840 --> 0:13:31.360
<v Speaker 2>that you found is extreme, so they know, you know, okay,

0:13:31.400 --> 0:13:35.000
<v Speaker 2>this is this is whistleblower aid An organization that helps

0:13:35.000 --> 0:13:38.520
<v Speaker 2>with that. Okay, it started messaging. Now okay, this is

0:13:38.559 --> 0:13:42.040
<v Speaker 2>the contact. Okay, well you know what could they do

0:13:42.120 --> 0:13:46.040
<v Speaker 2>with that? What could they reveal? Could that be subpoena?

0:13:46.160 --> 0:13:49.600
<v Speaker 2>Could that be revealed to your employer? The fundamental answer

0:13:49.679 --> 0:13:53.160
<v Speaker 2>is that, yes, this data is extremely revealing. It reveals

0:13:53.160 --> 0:13:58.040
<v Speaker 2>patterns about you. We could hypothesize, but we know already

0:13:58.080 --> 0:14:01.720
<v Speaker 2>that there's a woman who went to jail in the

0:14:01.840 --> 0:14:07.680
<v Speaker 2>US because Meta turned over DMS on Facebook messenger between

0:14:07.679 --> 0:14:10.680
<v Speaker 2>her and her daughter that showed her helping her daughter

0:14:11.320 --> 0:14:14.199
<v Speaker 2>access and deal with reproductive care in the state of Nebraska.

0:14:14.800 --> 0:14:18.000
<v Speaker 1>Do you find it frustrating then, when you see people

0:14:19.480 --> 0:14:24.920
<v Speaker 1>not care about their privacy even like some people wouldn't

0:14:24.960 --> 0:14:27.600
<v Speaker 1>be even be aware, but people who are aware but

0:14:27.720 --> 0:14:30.080
<v Speaker 1>just think this is the price I pay for these

0:14:30.120 --> 0:14:34.600
<v Speaker 1>platforms being free at the point of use, or maybe

0:14:34.640 --> 0:14:37.880
<v Speaker 1>younger people especially who just think that there's nothing really

0:14:37.880 --> 0:14:39.400
<v Speaker 1>in their life that they would mind being in the

0:14:39.440 --> 0:14:40.120
<v Speaker 1>public domain.

0:14:40.480 --> 0:14:44.200
<v Speaker 2>You know, I find that frustrating when the argument that

0:14:44.400 --> 0:14:48.360
<v Speaker 2>no one cares about privacy is leveraged by these firms

0:14:48.400 --> 0:14:52.560
<v Speaker 2>to justify what they're doing. It. Don't actually find it

0:14:52.600 --> 0:14:57.400
<v Speaker 2>frustrating when people who really, in my view, have little

0:14:57.480 --> 0:15:00.920
<v Speaker 2>choice say you know, hey, I can't do anything about this.

0:15:01.120 --> 0:15:04.080
<v Speaker 2>I'm going to delete that from my long list of

0:15:04.120 --> 0:15:07.560
<v Speaker 2>worries because I don't feel that I am empowered to

0:15:07.840 --> 0:15:11.920
<v Speaker 2>act on this. I actually don't believe that people don't

0:15:11.960 --> 0:15:15.120
<v Speaker 2>care about privacy, but I believe that the stakes are

0:15:15.240 --> 0:15:17.920
<v Speaker 2>very obscure, and we're all living in a world where

0:15:18.520 --> 0:15:21.160
<v Speaker 2>we have dishes to do, we have bills to pay,

0:15:21.160 --> 0:15:23.360
<v Speaker 2>we have things that are right on our doorstep to

0:15:23.440 --> 0:15:27.400
<v Speaker 2>deal with, and thinking about data flows and privacy legislation

0:15:27.520 --> 0:15:46.040
<v Speaker 2>is probably pretty far down on our list of concerns.

0:15:47.400 --> 0:15:49.760
<v Speaker 1>You had two big moments in the growth of Signal,

0:15:50.120 --> 0:15:54.000
<v Speaker 1>both of which I remember twenty thirteen when Edward Snowden

0:15:54.240 --> 0:15:59.000
<v Speaker 1>used it and trusted it to talk to journalists about

0:15:59.560 --> 0:16:04.040
<v Speaker 1>US and surveillance. And the second was Signal Gate last year,

0:16:04.200 --> 0:16:07.840
<v Speaker 1>when a journalist was added to a supposedly highly classified

0:16:07.840 --> 0:16:13.040
<v Speaker 1>Trump administration discussion on military action. When the second happened,

0:16:13.080 --> 0:16:16.320
<v Speaker 1>what did you think, this is like a great advertisement

0:16:16.360 --> 0:16:18.560
<v Speaker 1>for Signal because it's secure enough for the US government,

0:16:18.960 --> 0:16:21.400
<v Speaker 1>Or I thought, WHOA.

0:16:21.640 --> 0:16:24.080
<v Speaker 2>What's going on this morning? This is this is an

0:16:24.120 --> 0:16:30.400
<v Speaker 2>interesting story. You know, the article hit my feed the

0:16:30.400 --> 0:16:33.560
<v Speaker 2>same time it hit anyone else's feed, and I think

0:16:34.240 --> 0:16:36.760
<v Speaker 2>I had to read it twice to really let it

0:16:36.800 --> 0:16:40.560
<v Speaker 2>sink in, because it read almost like a telenovela. What

0:16:40.640 --> 0:16:44.520
<v Speaker 2>I didn't anticipate after reading it initially was that Signal

0:16:44.520 --> 0:16:47.280
<v Speaker 2>would become a main character and then it would become

0:16:47.600 --> 0:16:52.480
<v Speaker 2>a story about Signal. And what I was primarily concerned

0:16:52.520 --> 0:16:55.880
<v Speaker 2>about is it you know, suddenly Signal was being reported on,

0:16:56.160 --> 0:17:02.360
<v Speaker 2>spoken about by millions of people who weren't familiar with

0:17:02.440 --> 0:17:07.320
<v Speaker 2>the intricacies of encrypted messaging or the technology itself, and

0:17:07.400 --> 0:17:12.920
<v Speaker 2>I wanted to prevent misinformation that could hinder signal or

0:17:13.160 --> 0:17:16.600
<v Speaker 2>misinform people who you know, often need to choose signal

0:17:16.880 --> 0:17:18.200
<v Speaker 2>in life or death concept.

0:17:17.920 --> 0:17:20.880
<v Speaker 1>Cause this was essentially about human frailty, someone who shouldn't

0:17:20.880 --> 0:17:22.720
<v Speaker 1>have been added to that chat being added, And your

0:17:22.720 --> 0:17:25.040
<v Speaker 1>worry was that it would make it look like Signal

0:17:25.080 --> 0:17:26.119
<v Speaker 1>itself was insecure.

0:17:26.359 --> 0:17:29.200
<v Speaker 2>Yeah, it would be framed as a problem with Signal

0:17:29.320 --> 0:17:32.720
<v Speaker 2>or a problem with privacy or you know. That was

0:17:32.760 --> 0:17:36.639
<v Speaker 2>a very volatile, very heightened news cycle, so how that

0:17:36.680 --> 0:17:39.360
<v Speaker 2>would turn out was not clear, But yes, I mean,

0:17:39.400 --> 0:17:42.960
<v Speaker 2>this was ultimately men's thumbs doing things that is not

0:17:43.119 --> 0:17:45.399
<v Speaker 2>something an application developer can take its.

0:17:45.560 --> 0:17:46.639
<v Speaker 1>Was it good for numbers?

0:17:47.000 --> 0:17:49.560
<v Speaker 2>It was, Yeah, it's turned out well for us. We

0:17:49.800 --> 0:17:53.040
<v Speaker 2>saw growth across the US, across the world, and we've

0:17:53.080 --> 0:17:55.200
<v Speaker 2>seen pretty steady growth since then.

0:17:55.560 --> 0:17:58.200
<v Speaker 1>Can I ask you, Meredith, how you reconcile yourself with

0:17:58.560 --> 0:18:01.639
<v Speaker 1>where your product is being used by people to do

0:18:01.760 --> 0:18:06.400
<v Speaker 1>harm to others and particularly the example of child sexual abuse?

0:18:06.600 --> 0:18:10.560
<v Speaker 1>Am I right in thinking that if p defiles abuses

0:18:10.680 --> 0:18:14.400
<v Speaker 1>are using Signal to circulate images of child sexual abuse,

0:18:15.200 --> 0:18:19.120
<v Speaker 1>that can happen without them fearing that you will hand

0:18:19.160 --> 0:18:20.480
<v Speaker 1>over data to law enforcement.

0:18:21.480 --> 0:18:25.840
<v Speaker 2>We go out of our way to not be a

0:18:25.840 --> 0:18:29.879
<v Speaker 2>social media company, so we could pump our numbers fairly

0:18:29.920 --> 0:18:34.160
<v Speaker 2>easily if we added things like broadcast channels or features

0:18:34.160 --> 0:18:37.720
<v Speaker 2>that allowed you to find people with similar interests and

0:18:38.119 --> 0:18:42.760
<v Speaker 2>you meet strangers on Signal. But Signal has made a

0:18:42.920 --> 0:18:47.320
<v Speaker 2>very conscientious choice to stay in our narrow lane as

0:18:47.359 --> 0:18:52.240
<v Speaker 2>an interpersonal communications app, which means you have to encounter

0:18:52.800 --> 0:18:56.760
<v Speaker 2>the person you're talking to in some capacity that is

0:18:56.800 --> 0:19:00.119
<v Speaker 2>not a viral finder friend feed.

0:19:00.320 --> 0:19:03.159
<v Speaker 1>That's outside the app essentially. Yeah, But once you do

0:19:03.240 --> 0:19:05.600
<v Speaker 1>find that person, you can be sure that anything you

0:19:05.640 --> 0:19:07.800
<v Speaker 1>share with them is free.

0:19:07.960 --> 0:19:10.639
<v Speaker 2>That is outside. That is a sign promise. We are

0:19:10.800 --> 0:19:15.080
<v Speaker 2>open source. We are you know, people audit our code regularly.

0:19:15.160 --> 0:19:18.320
<v Speaker 2>We have the open source community looking at our code.

0:19:18.760 --> 0:19:23.080
<v Speaker 2>We cannot collect that data, we cannot turn that data

0:19:23.160 --> 0:19:26.639
<v Speaker 2>over and you know, look, we limit the features that

0:19:26.680 --> 0:19:30.240
<v Speaker 2>allow virality, We limit the features that allow mass distribution,

0:19:31.000 --> 0:19:35.159
<v Speaker 2>and we do that very conscientiously. I also need to

0:19:35.240 --> 0:19:38.640
<v Speaker 2>make a fundamental point, that is, you know, it is

0:19:39.400 --> 0:19:44.840
<v Speaker 2>simply true that encryption either works for everyone or it

0:19:44.880 --> 0:19:47.600
<v Speaker 2>works for no one. There is no way to have

0:19:47.640 --> 0:19:51.880
<v Speaker 2>a golden key or a special back door or access

0:19:52.119 --> 0:19:53.120
<v Speaker 2>just for the good guys.

0:19:53.400 --> 0:19:55.800
<v Speaker 1>And do you have a struggle with that knowing that

0:19:56.119 --> 0:19:58.280
<v Speaker 1>there will be people who are using it for the

0:19:58.320 --> 0:20:02.080
<v Speaker 1>most abusive possible images of children who are on signal?

0:20:02.280 --> 0:20:05.440
<v Speaker 1>Right now, are there moments where you find that hard

0:20:05.840 --> 0:20:08.360
<v Speaker 1>to reconcile or come to terms with yourself?

0:20:09.680 --> 0:20:14.080
<v Speaker 2>The fact of child sexual abuse is horrifying to me, absolutely,

0:20:14.320 --> 0:20:17.359
<v Speaker 2>and I think we need to take that so seriously

0:20:17.920 --> 0:20:20.399
<v Speaker 2>as a society that needs to get much much better

0:20:20.920 --> 0:20:26.000
<v Speaker 2>with looking gnarly problems in the face and actually tackling them.

0:20:26.080 --> 0:20:31.679
<v Speaker 2>But I also struggle with the persistent focus on privacy

0:20:32.160 --> 0:20:37.919
<v Speaker 2>as the purported barrier to solving these issues, when we

0:20:38.040 --> 0:20:42.719
<v Speaker 2>have most of the Epstein list walking around without consequences,

0:20:43.119 --> 0:20:47.399
<v Speaker 2>when we have so few mechanisms to actually help children

0:20:47.480 --> 0:20:50.160
<v Speaker 2>when they report a powerful figure in their community, when

0:20:50.200 --> 0:20:56.280
<v Speaker 2>we are defunding the actual law enforcement officers who take

0:20:56.400 --> 0:21:01.920
<v Speaker 2>this data, act on this data, move to prosecute the uncles,

0:21:02.560 --> 0:21:06.480
<v Speaker 2>the fathers, the pastors who are perpetrating this abuse.

0:21:06.920 --> 0:21:09.159
<v Speaker 1>Can I give you the example in the UK where

0:21:09.840 --> 0:21:12.359
<v Speaker 1>many people would think, you have to use every tool

0:21:12.400 --> 0:21:14.320
<v Speaker 1>in the block, right, So you have to use the

0:21:14.400 --> 0:21:17.119
<v Speaker 1>leavers you're pointing to, and perhaps you have to use

0:21:17.160 --> 0:21:20.080
<v Speaker 1>the leavers in tech as well. So the UK government

0:21:20.440 --> 0:21:26.080
<v Speaker 1>is moving towards making it compulsory for tech companies to

0:21:26.720 --> 0:21:31.800
<v Speaker 1>ensure that children can't take or share nude images through

0:21:31.840 --> 0:21:35.239
<v Speaker 1>their platforms. Is that workable? Do you think? And if

0:21:35.320 --> 0:21:38.840
<v Speaker 1>the UK compelled that, what would it mean for your

0:21:38.880 --> 0:21:39.760
<v Speaker 1>presence here?

0:21:40.240 --> 0:21:43.400
<v Speaker 2>Well, I'll start by saying there's never been more data

0:21:43.440 --> 0:21:47.080
<v Speaker 2>accessible to law enforcement than there is now that were

0:21:47.200 --> 0:21:50.679
<v Speaker 2>law enforcement well funded, we're social services well funded, they

0:21:50.760 --> 0:21:54.359
<v Speaker 2>could act on and often don't, often delay, or often

0:21:54.640 --> 0:21:57.520
<v Speaker 2>simply don't have the capacity. So when we talk about

0:21:57.520 --> 0:22:00.199
<v Speaker 2>every tool in the toolbox. We need to recognize that

0:22:00.240 --> 0:22:02.680
<v Speaker 2>there are some tools that are not being picked up,

0:22:03.080 --> 0:22:06.600
<v Speaker 2>and there are some tools that seem an obsessive focus

0:22:06.680 --> 0:22:10.400
<v Speaker 2>of politicians who like to abstract these very difficult problems

0:22:10.720 --> 0:22:14.400
<v Speaker 2>into the technical domain and pretend there is one technical

0:22:14.440 --> 0:22:17.840
<v Speaker 2>fix that can apply to solve a problem that needs.

0:22:17.800 --> 0:22:19.600
<v Speaker 1>So this is one of many. What do you think

0:22:19.640 --> 0:22:19.920
<v Speaker 1>of it?

0:22:20.400 --> 0:22:25.560
<v Speaker 2>What do I think of mass scanning of devices that

0:22:25.600 --> 0:22:27.959
<v Speaker 2>would be age gated? You know, we put out a

0:22:27.960 --> 0:22:31.639
<v Speaker 2>statement Cure Starmer announced what seems to be a fairly

0:22:31.760 --> 0:22:38.520
<v Speaker 2>hasty proposal to implement nudity scanning as a mandatory feature

0:22:38.560 --> 0:22:42.159
<v Speaker 2>on all devices that would be sold or used in

0:22:42.200 --> 0:22:45.080
<v Speaker 2>the UK, and that you could only opt out of

0:22:45.119 --> 0:22:48.159
<v Speaker 2>this nudity scanning if you provided some sort of verification

0:22:48.320 --> 0:22:52.040
<v Speaker 2>that you were above a certain age threshold. Correct. There

0:22:52.080 --> 0:22:57.240
<v Speaker 2>are very few technical details on that system, but our

0:22:57.320 --> 0:23:00.919
<v Speaker 2>position and our position on the client side scanning debate

0:23:01.040 --> 0:23:04.280
<v Speaker 2>that circled the Online Safety Bill as well now act,

0:23:04.960 --> 0:23:11.440
<v Speaker 2>is that this constitutes very dangerous mass surveillance. That having

0:23:11.720 --> 0:23:15.480
<v Speaker 2>a scanning system on your device that is going to,

0:23:16.000 --> 0:23:20.600
<v Speaker 2>you know, purportedly detect nudity can easily be weaponized by

0:23:20.640 --> 0:23:25.240
<v Speaker 2>governments for political speech, for other content, and that we

0:23:25.359 --> 0:23:29.520
<v Speaker 2>don't understand the technical mechanism by which this would be implemented,

0:23:29.560 --> 0:23:33.320
<v Speaker 2>because my assumption is there is no specification for this,

0:23:34.040 --> 0:23:36.439
<v Speaker 2>but that you know, we would be looking at a

0:23:36.480 --> 0:23:40.520
<v Speaker 2>system that would likely phone home, so that's sending that

0:23:40.640 --> 0:23:45.639
<v Speaker 2>data to governments, to tech companies and ultimately comprise a

0:23:46.200 --> 0:23:51.879
<v Speaker 2>mass surveillance system of permitted speech that would be implemented

0:23:51.920 --> 0:23:56.640
<v Speaker 2>across everyone's device and pose real threats to signal's ability

0:23:56.680 --> 0:24:02.000
<v Speaker 2>to provide the robust privacy guarantees that military is, governments Ukraine,

0:24:02.320 --> 0:24:06.000
<v Speaker 2>human rights workers, journalists, people around the world rely on

0:24:06.119 --> 0:24:07.160
<v Speaker 2>often for life or death.

0:24:07.280 --> 0:24:10.360
<v Speaker 1>With all of the big brains in tech, you really

0:24:10.359 --> 0:24:13.959
<v Speaker 1>don't think it's possible to refine the product in a

0:24:13.960 --> 0:24:17.760
<v Speaker 1>way that protects children and does the age verification and

0:24:17.800 --> 0:24:20.679
<v Speaker 1>the certain kinds of images without it having these, you know,

0:24:20.800 --> 0:24:24.320
<v Speaker 1>massive repercussions and implications that you're describing.

0:24:25.080 --> 0:24:28.119
<v Speaker 2>Well. Part of being a big brain and tech, I

0:24:28.119 --> 0:24:32.200
<v Speaker 2>think is really understanding the nature and the limits of

0:24:32.200 --> 0:24:36.320
<v Speaker 2>what technology can do. And this isn't a question of

0:24:36.600 --> 0:24:41.879
<v Speaker 2>go think harder nerds, right, there are real limits to

0:24:42.080 --> 0:24:43.800
<v Speaker 2>what tech can and can't do.

0:24:43.720 --> 0:24:46.240
<v Speaker 1>And any texts coming up with new things all the time.

0:24:46.440 --> 0:24:49.000
<v Speaker 2>Their marketing departments are talking about it all the time.

0:24:49.040 --> 0:24:51.160
<v Speaker 2>But you know, if you look at the implementation layer,

0:24:51.200 --> 0:24:53.639
<v Speaker 2>often what you're looking at is sticky tape and twine

0:24:53.720 --> 0:24:56.480
<v Speaker 2>and a handful of legacy libraries that may or may

0:24:56.520 --> 0:25:00.439
<v Speaker 2>not be suitable for purpose. The limits of your the

0:25:00.520 --> 0:25:04.679
<v Speaker 2>limits of the memory that's running on your hardware. You know,

0:25:04.800 --> 0:25:10.520
<v Speaker 2>these are fundamental material limitations and there is no technology

0:25:10.960 --> 0:25:12.920
<v Speaker 2>that has been shown to do this in a way

0:25:12.960 --> 0:25:14.840
<v Speaker 2>that is robust and safe.

0:25:15.040 --> 0:25:18.239
<v Speaker 1>Yeah, I also just wonder that whether how sustainable your

0:25:18.240 --> 0:25:19.959
<v Speaker 1>position is. I hear what you're saying, but it's not

0:25:20.000 --> 0:25:22.600
<v Speaker 1>only the UK government that's really concerned on this front.

0:25:22.640 --> 0:25:26.600
<v Speaker 1>The EU also had temporary legislation we just ran out,

0:25:26.600 --> 0:25:29.760
<v Speaker 1>but it was essentially asking tech companies to voluntarily scan

0:25:29.960 --> 0:25:30.760
<v Speaker 1>their platforms.

0:25:30.800 --> 0:25:32.320
<v Speaker 2>And we've taken the same position on.

0:25:32.280 --> 0:25:35.520
<v Speaker 1>That on that this does seem to be the way

0:25:35.520 --> 0:25:38.240
<v Speaker 1>that governments are going and if it is, can you

0:25:38.320 --> 0:25:41.120
<v Speaker 1>imagine signal not being available in the UK or the EU?

0:25:41.280 --> 0:25:45.520
<v Speaker 1>Like you threaten to walk over the Online Safety Bill

0:25:45.640 --> 0:25:46.840
<v Speaker 1>as it was then in the UK.

0:25:47.480 --> 0:25:52.440
<v Speaker 2>Look, you know, we don't really do pr right. We

0:25:52.680 --> 0:25:55.800
<v Speaker 2>have a position, and our position is we would rather

0:25:56.000 --> 0:26:01.560
<v Speaker 2>exit a market than undermine the technical guarantees that people

0:26:01.600 --> 0:26:06.919
<v Speaker 2>trust for their privacy. That doesn't change based on jurisdiction.

0:26:07.000 --> 0:26:09.679
<v Speaker 2>It doesn't change based on your and that's not something

0:26:09.720 --> 0:26:13.320
<v Speaker 2>we take lightly. We're not just like flexing or being recalcitrant.

0:26:13.800 --> 0:26:18.679
<v Speaker 2>That is, recognizing that encryption requires a network effect, that

0:26:18.800 --> 0:26:23.240
<v Speaker 2>a back door in any market compromises a door into

0:26:23.880 --> 0:26:29.160
<v Speaker 2>Signal that would be weaponized by nation state adversaries. Signal

0:26:29.280 --> 0:26:33.600
<v Speaker 2>is the only mass communications platform that provides the level

0:26:33.600 --> 0:26:36.320
<v Speaker 2>of privacy that we provide, and we know that we

0:26:36.359 --> 0:26:41.800
<v Speaker 2>have adversaries that are consistently searching for vulnerabilities in our code.

0:26:41.880 --> 0:26:42.480
<v Speaker 1>Who do you mean?

0:26:42.760 --> 0:26:45.600
<v Speaker 2>Who do we mean? Any government with a capable intelligence

0:26:45.800 --> 0:26:49.320
<v Speaker 2>agency is spending some of that effort looking.

0:26:49.040 --> 0:26:50.600
<v Speaker 1>At including your own government.

0:26:51.800 --> 0:26:55.480
<v Speaker 2>I don't have clearance to know that for sure, but.

0:26:55.680 --> 0:26:58.560
<v Speaker 1>I'm just wondering if you mean Russia, China, or do

0:26:58.600 --> 0:26:59.800
<v Speaker 1>you mean governments in the.

0:27:00.080 --> 0:27:05.199
<v Speaker 2>We know it has been represented to us that you

0:27:05.240 --> 0:27:10.240
<v Speaker 2>know Russia is what we've seen bounties for the delivery

0:27:10.240 --> 0:27:12.560
<v Speaker 2>of vulnerabilities, meaning that you know they'll pay you if

0:27:12.600 --> 0:27:15.040
<v Speaker 2>you can find a vulnerability they could exploit that have

0:27:15.160 --> 0:27:18.879
<v Speaker 2>been offered by the Russian government. We have had people

0:27:18.960 --> 0:27:22.720
<v Speaker 2>with who would know this or represent that to us.

0:27:23.320 --> 0:27:26.320
<v Speaker 2>But if you look at the people who rely on this,

0:27:26.520 --> 0:27:32.879
<v Speaker 2>heads of state, government officials, military strategists, investigative journalists, people

0:27:32.960 --> 0:27:38.320
<v Speaker 2>reporting corruption, I think we need to just recognize that, yes,

0:27:38.480 --> 0:27:43.960
<v Speaker 2>you have people who have a very vested interest in

0:27:43.960 --> 0:27:50.320
<v Speaker 2>infiltrating those networks, in accessing that communication and accessing that data,

0:27:50.680 --> 0:27:53.600
<v Speaker 2>and we'll spend resources doing it. That's the objective function

0:27:53.800 --> 0:27:57.719
<v Speaker 2>of an intelligence agency. And Signal is the only one

0:27:57.840 --> 0:28:00.399
<v Speaker 2>out there doing this, So it would be very It

0:28:00.400 --> 0:28:03.520
<v Speaker 2>would be an absolute breach of our duty of care

0:28:03.880 --> 0:28:07.520
<v Speaker 2>were we to implement a backdoor in Signal, were we

0:28:07.600 --> 0:28:10.600
<v Speaker 2>to undermine that promise. And you know, we're a nonprofit.

0:28:10.800 --> 0:28:12.720
<v Speaker 2>We are not a tech company that is, you know,

0:28:13.040 --> 0:28:16.399
<v Speaker 2>juicing numbers to IPO and exit.

0:28:16.800 --> 0:28:18.480
<v Speaker 1>Does we take this very SONI That's one of the

0:28:18.520 --> 0:28:22.920
<v Speaker 1>intriguing things about Signal, And I wondered how hard it

0:28:23.160 --> 0:28:27.480
<v Speaker 1>makes it because you have some big donors, but you

0:28:27.600 --> 0:28:31.399
<v Speaker 1>also have users who donate to you. So today, what

0:28:31.440 --> 0:28:33.879
<v Speaker 1>are your annual running costs? What are the costs you

0:28:33.960 --> 0:28:36.560
<v Speaker 1>have to cover. It's about fifty million dollars a year.

0:28:36.640 --> 0:28:40.840
<v Speaker 1>Five zero And do you fall short in donations we're sustaining. Yeah,

0:28:40.880 --> 0:28:44.520
<v Speaker 1>it's about seventy percent large donors, thirty percent people who

0:28:44.600 --> 0:28:46.920
<v Speaker 1>use the app and donate, you know, five dollars a month,

0:28:47.120 --> 0:28:48.080
<v Speaker 1>twenty dollars a month.

0:28:48.480 --> 0:28:49.640
<v Speaker 2>We love them. Thank you.

0:28:49.880 --> 0:28:51.000
<v Speaker 1>Does it feel like a struggle?

0:28:52.400 --> 0:28:56.280
<v Speaker 2>I would, It's something I think about a lot. One

0:28:56.320 --> 0:28:58.320
<v Speaker 2>of the real issues we're working with is that the

0:28:58.360 --> 0:29:02.280
<v Speaker 2>economic engine of the tech industry as it's shaped now

0:29:03.000 --> 0:29:07.520
<v Speaker 2>is surveillance. It's the collection and monetization of data that's

0:29:07.560 --> 0:29:12.520
<v Speaker 2>not alarmist, that is technically true, and so existing within

0:29:12.640 --> 0:29:17.600
<v Speaker 2>a framework where all of the pressures for revenue, for growth,

0:29:17.760 --> 0:29:21.840
<v Speaker 2>for profit for IPO are pointing in the direction of

0:29:22.040 --> 0:29:25.880
<v Speaker 2>collect and monetize data means that we have to choose

0:29:25.880 --> 0:29:28.360
<v Speaker 2>a different model, one where we don't have a VC

0:29:29.120 --> 0:29:32.560
<v Speaker 2>saying hey, you're not going to meet your quarterly goals,

0:29:32.760 --> 0:29:34.840
<v Speaker 2>you're not going to get your Series B if you

0:29:34.880 --> 0:29:37.440
<v Speaker 2>don't figure out how to monetize, where we don't have

0:29:37.640 --> 0:29:41.560
<v Speaker 2>a board and shareholders saying figure out how to carve

0:29:41.600 --> 0:29:44.280
<v Speaker 2>out some data collection so you could build an AI

0:29:44.400 --> 0:29:48.200
<v Speaker 2>model so you can continue to meet and exceed market consensus.

0:29:48.440 --> 0:29:51.080
<v Speaker 1>Well, that made me wonder about the llms and chat,

0:29:51.120 --> 0:29:53.600
<v Speaker 1>GPT and claud, Like, what would you like us to

0:29:53.600 --> 0:29:56.600
<v Speaker 1>be aware of in privacy terms when we use those,

0:29:56.640 --> 0:30:00.200
<v Speaker 1>when we ask questions, when we put prompts into the individ.

0:30:00.560 --> 0:30:04.240
<v Speaker 2>Well, I mean flatly like, these are not your friends,

0:30:04.280 --> 0:30:09.920
<v Speaker 2>these are not conscious beings, these are not sentient interlock cuters.

0:30:10.240 --> 0:30:14.240
<v Speaker 2>They're very useful, sure mean, things are very.

0:30:13.960 --> 0:30:17.280
<v Speaker 1>Chats And you can say you don't want your chance

0:30:17.320 --> 0:30:18.320
<v Speaker 1>to be shared.

0:30:18.520 --> 0:30:21.440
<v Speaker 2>You can say all sorts of things, but ultimately the

0:30:21.480 --> 0:30:24.600
<v Speaker 2>power to determine whether that happens or not is in

0:30:24.640 --> 0:30:29.040
<v Speaker 2>the hands of the entity that is running that service.

0:30:29.080 --> 0:30:32.040
<v Speaker 2>And those can be stored, those can be logged, those

0:30:32.080 --> 0:30:35.680
<v Speaker 2>can be mined for more data. Those can be tweaked

0:30:36.080 --> 0:30:40.880
<v Speaker 2>slightly to calibrate the responses to your purported preference on

0:30:40.960 --> 0:30:45.920
<v Speaker 2>behalf of say advertisers. You know, maybe there's an airline

0:30:46.360 --> 0:30:49.960
<v Speaker 2>that has paid to advertise and instead of showing ads.

0:30:50.440 --> 0:30:53.120
<v Speaker 2>Perhaps and I'm working with a hypothetical here. We don't

0:30:53.120 --> 0:30:55.280
<v Speaker 2>have evidence of this, but this is very much in

0:30:55.320 --> 0:30:58.160
<v Speaker 2>the realm of the possible. You say, hey, chat GPT,

0:30:58.320 --> 0:31:01.520
<v Speaker 2>find me a very cheap flight, and instead of finding

0:31:01.560 --> 0:31:04.920
<v Speaker 2>you the cheapest flight, it finds you one with their

0:31:04.960 --> 0:31:08.800
<v Speaker 2>preferred advertising partner that is not cheaper. Do you use them?

0:31:09.160 --> 0:31:11.720
<v Speaker 2>I mean, I asked them to form out a document

0:31:11.760 --> 0:31:15.120
<v Speaker 2>here and there, but no, I have no relationship with them.

0:31:15.120 --> 0:31:17.920
<v Speaker 2>I don't ask them questions. You know, I'm very serious

0:31:17.920 --> 0:31:22.200
<v Speaker 2>about my thinking and my writing, and I don't want

0:31:22.600 --> 0:31:27.360
<v Speaker 2>the process of working through an idea, the struggle of

0:31:27.560 --> 0:31:32.080
<v Speaker 2>not knowing and then figuring it out, to be foreclosed

0:31:32.160 --> 0:31:35.880
<v Speaker 2>or eclipsed by the response of a system that's kind

0:31:35.880 --> 0:31:37.680
<v Speaker 2>of averaging what's already out there.

0:31:37.800 --> 0:31:40.080
<v Speaker 1>Yeah, I guess that's a different argument. That's also about

0:31:40.120 --> 0:31:45.800
<v Speaker 1>safeguarding your headspace or your thinking time. Beyond privacy, the

0:31:45.880 --> 0:31:51.040
<v Speaker 1>next generation, or the next layer of all these models

0:31:51.280 --> 0:31:55.400
<v Speaker 1>is the more autonomous personal assistant type thing. You know,

0:31:55.440 --> 0:31:58.200
<v Speaker 1>Microsoft Copilot most of Usilmano was on this program and

0:31:58.720 --> 0:32:01.080
<v Speaker 1>he said to me, you know, by next Christmas, you

0:32:01.080 --> 0:32:04.800
<v Speaker 1>can be buying your Christmas presents through Microsoft Copilot, working

0:32:04.840 --> 0:32:08.200
<v Speaker 1>out what you need to buy for whom. What do

0:32:08.240 --> 0:32:11.640
<v Speaker 1>you think of that kind of autonomous agent being in

0:32:11.760 --> 0:32:13.960
<v Speaker 1>people's lives? What would it mean in privacy terms?

0:32:14.320 --> 0:32:17.080
<v Speaker 2>Yeah, I mean, that's a great question. And let's let's

0:32:17.120 --> 0:32:22.240
<v Speaker 2>take myself as little vision of you know, AI Christmas.

0:32:22.560 --> 0:32:27.240
<v Speaker 2>Who could I can't wait? Yeah, everyone's getting a give

0:32:27.360 --> 0:32:34.120
<v Speaker 2>certificate to Microsoft Copilot. But what would what would what

0:32:34.160 --> 0:32:36.840
<v Speaker 2>would that mean? Right? So we have a Windows laptop.

0:32:37.040 --> 0:32:41.760
<v Speaker 2>It's running co Pilot, and I say, hey, Copilot, I

0:32:41.800 --> 0:32:43.440
<v Speaker 2>don't want to live my life. I would like you

0:32:43.480 --> 0:32:45.600
<v Speaker 2>to live it for me. And I don't really want

0:32:45.640 --> 0:32:48.040
<v Speaker 2>to think about the warm feelings I have for those

0:32:48.120 --> 0:32:50.280
<v Speaker 2>that I'm buying gifts for, or think about what they

0:32:50.320 --> 0:32:53.520
<v Speaker 2>might actually love or you know, engage in that like

0:32:53.800 --> 0:32:57.640
<v Speaker 2>generous interpersonal communal feeling that maybe should be at the

0:32:57.640 --> 0:32:59.320
<v Speaker 2>heart of Christmas. Who knows, But I don't have the

0:32:59.360 --> 0:33:02.240
<v Speaker 2>time for that. I'm going to play pickleball or whatever

0:33:02.280 --> 0:33:04.560
<v Speaker 2>the thing I am doing, and I would like you

0:33:04.600 --> 0:33:06.200
<v Speaker 2>to do that for me. And I would like you

0:33:06.280 --> 0:33:08.520
<v Speaker 2>to figure out what gifts to buy. I would like

0:33:08.560 --> 0:33:12.840
<v Speaker 2>you to consult with my siblings group chat and see,

0:33:12.880 --> 0:33:16.400
<v Speaker 2>you know what our family might like. I would like

0:33:16.400 --> 0:33:18.240
<v Speaker 2>you to buy those gifts, and I would like you

0:33:18.320 --> 0:33:22.360
<v Speaker 2>to figure out how to get them to the right place. Well, okay,

0:33:22.360 --> 0:33:25.320
<v Speaker 2>that would need access to my credit card. That would

0:33:25.360 --> 0:33:29.400
<v Speaker 2>need access to my browser to pretend it is me,

0:33:29.680 --> 0:33:34.040
<v Speaker 2>to drive through that browsing history to find those things.

0:33:34.840 --> 0:33:37.200
<v Speaker 2>If it's access to my siblings group chat, that would

0:33:37.200 --> 0:33:39.480
<v Speaker 2>need access to my signal, that would need the ability

0:33:39.520 --> 0:33:42.440
<v Speaker 2>to message my siblings on my behalf, and then it

0:33:42.480 --> 0:33:46.240
<v Speaker 2>would need my home address. Say, to ship those gifts,

0:33:46.920 --> 0:33:50.239
<v Speaker 2>it would need maybe my calendar to make sure they

0:33:50.280 --> 0:33:52.600
<v Speaker 2>get there on time, et cetera, et cetera. So what

0:33:52.640 --> 0:33:56.960
<v Speaker 2>you've just described is a system with very very pervasive

0:33:57.040 --> 0:34:03.520
<v Speaker 2>access across multiple apps, applications, and services, and importantly, in

0:34:03.560 --> 0:34:07.120
<v Speaker 2>the context of signal, it would constitute a kind of

0:34:07.200 --> 0:34:09.680
<v Speaker 2>a backdoor in the signal. It would be able to

0:34:09.719 --> 0:34:12.280
<v Speaker 2>access signal pull data out of protection.

0:34:12.000 --> 0:34:15.080
<v Speaker 1>Isn't just privacy and access to our data. It's actually

0:34:15.080 --> 0:34:17.360
<v Speaker 1>that you think, once something like that ends up on

0:34:17.960 --> 0:34:21.959
<v Speaker 1>a phone because we've given it permission to, it could

0:34:22.000 --> 0:34:26.440
<v Speaker 1>actually be a threat to signals, ethos and promise.

0:34:26.880 --> 0:34:30.200
<v Speaker 2>We have described it metaphorically as breaking the blood brain

0:34:30.280 --> 0:34:35.600
<v Speaker 2>barrier between the operating system and the application layer, so.

0:34:35.840 --> 0:34:38.120
<v Speaker 1>You can make signals secure enough to protect against that.

0:34:38.600 --> 0:34:41.680
<v Speaker 2>Well, what we are seeing in the case of these

0:34:41.680 --> 0:34:46.360
<v Speaker 2>agents or promises like this, right the access of data

0:34:46.400 --> 0:34:50.160
<v Speaker 2>at the application layer, pulling that data into a context window.

0:34:50.160 --> 0:34:52.640
<v Speaker 2>But it's sort of a big data slurry that the

0:34:52.920 --> 0:34:58.280
<v Speaker 2>agent sort of acts on and remembers, and then implementations

0:34:58.440 --> 0:35:02.440
<v Speaker 2>architectures for this agent that are very unsafe from a

0:35:02.440 --> 0:35:04.320
<v Speaker 2>cybersecurity perspective.

0:35:03.840 --> 0:35:06.279
<v Speaker 1>Because the agent is essentially becoming the person or the

0:35:06.320 --> 0:35:09.000
<v Speaker 1>agent has been given the license to be. You have

0:35:09.080 --> 0:35:11.560
<v Speaker 1>to be me, well not you, because I'm sure you

0:35:11.600 --> 0:35:13.040
<v Speaker 1>would never go down this route, but.

0:35:13.400 --> 0:35:15.600
<v Speaker 2>I try to ask you. So that is my struggle.

0:35:16.560 --> 0:35:22.080
<v Speaker 2>But you've given it, in technical terms like route access

0:35:22.560 --> 0:35:25.600
<v Speaker 2>to your life. You've given it the keys to the kingdom,

0:35:25.680 --> 0:35:28.239
<v Speaker 2>because to act on your behalf, it needs to be

0:35:28.320 --> 0:35:30.400
<v Speaker 2>able to do everything you would do.

0:35:30.400 --> 0:35:32.839
<v Speaker 1>Do you worry then about this as an existential threat

0:35:32.840 --> 0:35:35.560
<v Speaker 1>to signal? It's clearly not that easy to run the foundation.

0:35:35.800 --> 0:35:38.880
<v Speaker 1>I do worry because it seems quite likely that it

0:35:38.920 --> 0:35:41.000
<v Speaker 1>will come to pass, just if you look at how

0:35:41.719 --> 0:35:43.920
<v Speaker 1>much bigger these companies are getting by the day, and

0:35:43.960 --> 0:35:46.200
<v Speaker 1>they're going to need to make more promises about what

0:35:46.239 --> 0:35:49.040
<v Speaker 1>they can offer. I use the example of co Pilot,

0:35:49.080 --> 0:35:49.760
<v Speaker 1>but there are others.

0:35:49.800 --> 0:35:54.000
<v Speaker 2>Clearly, any one of the operating system vendors that decides

0:35:54.080 --> 0:35:59.960
<v Speaker 2>to implement an agent with these capacities is ultimately hog

0:36:00.000 --> 0:36:04.000
<v Speaker 2>blowing out the ability for Signal and others to guarantee

0:36:04.000 --> 0:36:09.080
<v Speaker 2>privacy or data protection at the application layer. And this

0:36:09.239 --> 0:36:13.719
<v Speaker 2>is a quiet but profound shift that we are witnessing.

0:36:13.800 --> 0:36:14.520
<v Speaker 1>Can you fight it?

0:36:15.160 --> 0:36:19.320
<v Speaker 2>Oh, I don't believe in inevitability, and I am willing

0:36:19.560 --> 0:36:22.560
<v Speaker 2>to fight on my principles. I don't know if I

0:36:22.560 --> 0:36:25.399
<v Speaker 2>can win it, but I will do everything I can

0:36:25.480 --> 0:36:28.560
<v Speaker 2>because I think this is fundamental. I am concerned that

0:36:28.640 --> 0:36:34.040
<v Speaker 2>we are seeing extremely hasty decisions made by very powerful actors.

0:36:34.120 --> 0:36:37.319
<v Speaker 2>You know, we have three companies in the world with

0:36:37.480 --> 0:36:41.560
<v Speaker 2>control over our core operating systems. That's Google with Android,

0:36:41.800 --> 0:36:47.360
<v Speaker 2>Microsoft with Windows, and Apple with iOS mac os. We

0:36:47.600 --> 0:36:50.640
<v Speaker 2>all are relying on one or another of those, and

0:36:50.960 --> 0:36:55.440
<v Speaker 2>they can make decisions by FIAT that fundamentally harm our

0:36:55.480 --> 0:37:00.400
<v Speaker 2>collective cybersecurity, that take away any choice we might have

0:37:00.480 --> 0:37:05.719
<v Speaker 2>for privacy, and that completely hamstring developers at the application

0:37:05.840 --> 0:37:10.440
<v Speaker 2>layer who've spent so much labor, so much money, so

0:37:10.600 --> 0:37:16.360
<v Speaker 2>much intention on building different services that protect or do x,

0:37:16.520 --> 0:37:20.600
<v Speaker 2>y and z. So we are seeing, ironically, almost in

0:37:20.640 --> 0:37:25.520
<v Speaker 2>the name of AI agents, a move of agency away

0:37:26.280 --> 0:37:31.000
<v Speaker 2>from the application layer, away from the user, away from

0:37:31.040 --> 0:37:36.400
<v Speaker 2>the consumer who uses these devices, back into these centralized companies.

0:37:36.600 --> 0:37:39.320
<v Speaker 1>Give us some advice, then, because I think I could

0:37:39.360 --> 0:37:41.879
<v Speaker 1>take away okay, wherever you have the option to take

0:37:41.920 --> 0:37:44.960
<v Speaker 1>the box, saying to limit the use of your day

0:37:45.000 --> 0:37:46.560
<v Speaker 1>to do it. But I think you're also telling me

0:37:47.320 --> 0:37:51.240
<v Speaker 1>that you don't really trust in what you get back

0:37:51.320 --> 0:37:54.279
<v Speaker 1>for saying don't use my data, I don't want it shared.

0:37:54.120 --> 0:37:56.759
<v Speaker 2>Or you know. Look, I think we have evidence that

0:37:57.040 --> 0:38:02.120
<v Speaker 2>you know, simply trusting is and not a sound conclusion.

0:38:02.480 --> 0:38:04.520
<v Speaker 2>I do come from maybe an old school where I

0:38:04.520 --> 0:38:07.200
<v Speaker 2>think you should verify then trust. What is the advice?

0:38:07.480 --> 0:38:12.160
<v Speaker 2>I think healthy skepticism and a recognition that this technology

0:38:12.200 --> 0:38:15.120
<v Speaker 2>is not magic. It can be extremely useful. It can

0:38:15.160 --> 0:38:19.840
<v Speaker 2>have profound applications, particularly where finding patterns and large amounts

0:38:19.840 --> 0:38:25.200
<v Speaker 2>of data is useful for XYZ. But it runs on hardware.

0:38:25.640 --> 0:38:28.520
<v Speaker 2>It runs on software. When we talk about AI, we're

0:38:28.520 --> 0:38:32.240
<v Speaker 2>not talking about a magical sky god that can suddenly

0:38:32.400 --> 0:38:35.680
<v Speaker 2>solve all our problems, answer all our questions. We are

0:38:35.719 --> 0:38:40.080
<v Speaker 2>talking about a statistical model that is then embedded in

0:38:40.239 --> 0:38:45.600
<v Speaker 2>a hardware software system, using software libraries, implementing choices on

0:38:45.640 --> 0:38:48.680
<v Speaker 2>the user interface level that are made by people that

0:38:48.719 --> 0:38:50.080
<v Speaker 2>are susceptible to vulnerable.

0:38:50.120 --> 0:38:51.640
<v Speaker 1>Tell me what you do in your own life, because

0:38:52.000 --> 0:38:55.560
<v Speaker 1>signal for messaging? What else? What is common in the

0:38:55.600 --> 0:38:57.680
<v Speaker 1>rest of our lives that you never use and you

0:38:57.719 --> 0:38:59.239
<v Speaker 1>think we should. Like when you go for a run,

0:38:59.280 --> 0:39:01.920
<v Speaker 1>do you put it on st Rava? Do you you

0:39:01.960 --> 0:39:02.640
<v Speaker 1>would not do that?

0:39:03.239 --> 0:39:06.200
<v Speaker 2>No? I don't. I mean, I'm not a I'm not

0:39:06.560 --> 0:39:09.400
<v Speaker 2>like the full crumb of leverage for me is not

0:39:09.560 --> 0:39:12.200
<v Speaker 2>individual consumer choice. And I am not trying to be

0:39:13.440 --> 0:39:17.440
<v Speaker 2>holier than thou purest right. You know, I have a

0:39:17.760 --> 0:39:20.960
<v Speaker 2>tracking app I used for different health metrics. It's been

0:39:20.960 --> 0:39:23.319
<v Speaker 2>helpful for me. I travel a lot, and it's been

0:39:23.440 --> 0:39:24.560
<v Speaker 2>helpful for me seeing that.

0:39:24.400 --> 0:39:25.160
<v Speaker 1>That's health data.

0:39:25.400 --> 0:39:28.960
<v Speaker 2>Yeah, that is health data used. Yeah, and I benefit

0:39:29.000 --> 0:39:32.640
<v Speaker 2>from that. I think the lever we need to be

0:39:32.680 --> 0:39:36.440
<v Speaker 2>pushing on is much more systemic change. I am not

0:39:36.560 --> 0:39:38.799
<v Speaker 2>going to opt out of life. I am not going

0:39:38.880 --> 0:39:40.839
<v Speaker 2>to go live in a cabin in the woods and

0:39:40.840 --> 0:39:43.520
<v Speaker 2>like pay someone to print out my emails at the

0:39:43.520 --> 0:39:46.359
<v Speaker 2>post office in town or whatever we're talking about, you know,

0:39:46.480 --> 0:39:49.680
<v Speaker 2>so that I can feel private, right, Like, if we

0:39:49.719 --> 0:39:52.640
<v Speaker 2>look even at signals, privacy promises. We're not talking about

0:39:52.800 --> 0:39:57.319
<v Speaker 2>sort of isolation from our social, economic, political lives, living

0:39:57.360 --> 0:40:01.000
<v Speaker 2>in a vacuum of like holier than now CyberSecure, purity, purity.

0:40:01.360 --> 0:40:04.279
<v Speaker 2>We're talking about the ability to talk to other people,

0:40:04.360 --> 0:40:07.279
<v Speaker 2>to share with them, right, and so, Yeah, I have

0:40:07.320 --> 0:40:11.279
<v Speaker 2>an Instagram. I post mainly work events or whatever on that, right,

0:40:11.320 --> 0:40:14.520
<v Speaker 2>I use social media. I have the nts app. It's

0:40:14.560 --> 0:40:18.000
<v Speaker 2>a really good web radio that has great mixes. Right, Like,

0:40:18.239 --> 0:40:20.279
<v Speaker 2>I listen to that. Maybe they collect data, you know,

0:40:20.320 --> 0:40:22.640
<v Speaker 2>I push reject all, but I don't always trust that

0:40:22.640 --> 0:40:26.759
<v Speaker 2>that's actually rejecting cookies. I am a person living in

0:40:26.800 --> 0:40:29.400
<v Speaker 2>twenty twenty six and I want to participate in the world.

0:40:29.960 --> 0:40:32.319
<v Speaker 2>What I don't believe is that participation should be on

0:40:32.360 --> 0:40:34.680
<v Speaker 2>the terms of a handful of tech companies located in

0:40:34.719 --> 0:40:37.840
<v Speaker 2>the US whose choices are made based on the desire

0:40:37.880 --> 0:40:39.000
<v Speaker 2>to maximize revenue.

0:40:39.040 --> 0:40:41.560
<v Speaker 1>That you, too, are American tech.

0:40:41.560 --> 0:40:46.040
<v Speaker 2>I the irony abounds. Look at this a complex individual.

0:40:46.440 --> 0:40:48.600
<v Speaker 2>I am, and I worked in tech for twenty years.

0:40:48.600 --> 0:40:53.200
<v Speaker 2>I'm of the tech industry. So those two things are true.

0:40:53.360 --> 0:40:55.640
<v Speaker 1>But I think you're saying choice, make better choices, or

0:40:55.719 --> 0:40:57.960
<v Speaker 1>just think more carefully about your choices like, even if

0:40:58.000 --> 0:41:00.000
<v Speaker 1>you get that much from people, will you consider it

0:41:00.600 --> 0:41:01.880
<v Speaker 1>valuable or even a victory?

0:41:02.200 --> 0:41:06.839
<v Speaker 2>I would say those things, and that you be skeptical

0:41:06.920 --> 0:41:10.200
<v Speaker 2>of these claims. You're not stupid. You know the choice

0:41:10.280 --> 0:41:13.200
<v Speaker 2>is offered, yes, no or maybe maybe? Aren't the choices

0:41:13.280 --> 0:41:16.160
<v Speaker 2>you want as a dream a little bigger? What do

0:41:16.200 --> 0:41:19.000
<v Speaker 2>you want from these companies? What world do you want

0:41:19.040 --> 0:41:22.840
<v Speaker 2>to live in? What world would be the thing you

0:41:22.880 --> 0:41:26.319
<v Speaker 2>want to help build? And then how do you think

0:41:26.360 --> 0:41:31.560
<v Speaker 2>about demanding from the people who have commandeered, who have created,

0:41:31.640 --> 0:41:35.719
<v Speaker 2>who now hold so much power over our social and

0:41:35.760 --> 0:41:42.359
<v Speaker 2>political and economic institutions via the tech infrastructures, the information ecosystem,

0:41:42.480 --> 0:41:45.560
<v Speaker 2>on the tech platforms, the AI decision making that is

0:41:45.880 --> 0:41:49.360
<v Speaker 2>inoculating so many functions in our lives. What do we

0:41:49.480 --> 0:41:54.080
<v Speaker 2>need from them to help create that world? Not simply

0:41:54.120 --> 0:41:56.839
<v Speaker 2>if I'm given the choice yes or no, which do

0:41:56.920 --> 0:41:59.920
<v Speaker 2>I choose? And that's the limit of my ability to

0:42:00.560 --> 0:42:03.399
<v Speaker 2>enact any change in my life? But do we want

0:42:03.440 --> 0:42:07.239
<v Speaker 2>tech in our education system? And how? Yes? We do? Okay, well,

0:42:07.280 --> 0:42:09.440
<v Speaker 2>then what does it look like? And how do we

0:42:09.560 --> 0:42:12.360
<v Speaker 2>work to guarantee that we're actually using it to shape

0:42:12.680 --> 0:42:16.200
<v Speaker 2>that better world? And not simply being shaped by decisions

0:42:16.200 --> 0:42:18.520
<v Speaker 2>made on a whiteboard in Mountain View by a handful

0:42:18.560 --> 0:42:22.040
<v Speaker 2>of product managers who really have to ship AI this

0:42:22.160 --> 0:42:24.120
<v Speaker 2>quarter or they're not going to get promoted.

0:42:23.760 --> 0:42:25.560
<v Speaker 1>Who you probably know because you once used to work

0:42:25.560 --> 0:42:25.879
<v Speaker 1>with them.

0:42:25.960 --> 0:42:33.080
<v Speaker 2>Yeah, they're all named Mike. Even now is like that,

0:42:33.719 --> 0:42:35.840
<v Speaker 2>even now, even now I've heard, I've.

0:42:35.719 --> 0:42:37.920
<v Speaker 1>Heard you never missed that world. I mean it's a

0:42:38.000 --> 0:42:41.279
<v Speaker 1>very big Bucks world for other things.

0:42:41.360 --> 0:42:46.120
<v Speaker 2>Am I said, I'm not like rich rich? Yeah, no, no,

0:42:46.520 --> 0:42:49.359
<v Speaker 2>like I'm I'm hood rich. I'm happy being hood rich.

0:42:49.800 --> 0:42:53.000
<v Speaker 2>I don't need to be. It means rich in the

0:42:53.040 --> 0:42:55.759
<v Speaker 2>context of the neighborhood of you, not rich in the

0:42:55.800 --> 0:42:58.719
<v Speaker 2>context of complaining about the Champagne on your private flight.

0:42:59.600 --> 0:43:02.400
<v Speaker 1>You'll and your herds are both New York and Paris.

0:43:02.040 --> 0:43:07.879
<v Speaker 2>So yeah, you'll get the like. Don't cry it from me,

0:43:07.880 --> 0:43:09.560
<v Speaker 2>Meredith Whitika, thank you so much.

0:43:09.680 --> 0:43:18.080
<v Speaker 1>That was great, Thank you, and that is where we

0:43:18.160 --> 0:43:22.200
<v Speaker 1>said goodbye. Unknown to me. Later that day, Meredith was

0:43:22.280 --> 0:43:26.080
<v Speaker 1>visiting Downing Street to talk to Kirs Starmer's team about

0:43:26.200 --> 0:43:30.040
<v Speaker 1>signals concern at their plans. You can see a photograph

0:43:30.120 --> 0:43:32.319
<v Speaker 1>of her in front of the famous black door and

0:43:32.400 --> 0:43:35.480
<v Speaker 1>get all the wider context if you read the written

0:43:35.560 --> 0:43:39.160
<v Speaker 1>version of this at Bloomberg dot com forward slash Michelle

0:43:40.760 --> 0:43:44.520
<v Speaker 1>a few other notes. We asked Meta to comment on

0:43:44.600 --> 0:43:48.440
<v Speaker 1>her characterization of its business model and approach to privacy.

0:43:49.000 --> 0:43:53.600
<v Speaker 1>A spokesperson for WhatsApp responded by saying this, we also

0:43:53.680 --> 0:43:57.440
<v Speaker 1>look at privacy beyond encrypting messages. We don't keep logs

0:43:57.440 --> 0:44:00.520
<v Speaker 1>of who everyone is messaging or calling. We do not

0:44:00.640 --> 0:44:03.760
<v Speaker 1>track the messages people are sending one another for ads,

0:44:04.120 --> 0:44:08.640
<v Speaker 1>we do not sell user data. We approached Google too

0:44:08.680 --> 0:44:12.000
<v Speaker 1>on Meredith's assessment of their business model, but we've had

0:44:12.000 --> 0:44:19.040
<v Speaker 1>no response, and that Ai Christmas thought. The conversation with

0:44:19.160 --> 0:44:23.200
<v Speaker 1>Mustapha Suliman of Microsoft is in the feed. Just scroll

0:44:23.280 --> 0:44:27.880
<v Speaker 1>back to last December to find it. Finally, the team.

0:44:28.040 --> 0:44:31.120
<v Speaker 1>The producers are Jessica Beck and Chris Martlou. The video

0:44:31.239 --> 0:44:35.720
<v Speaker 1>producer is Andy Haywood. Social media's by Alex Morgan, audio

0:44:35.760 --> 0:44:39.160
<v Speaker 1>mixing by Richard Ward, our music is by Bart Warshaw,

0:44:39.600 --> 0:44:44.120
<v Speaker 1>and the executive producer is Louisa Lewis. At Bloomberg Weekend

0:44:44.280 --> 0:44:48.560
<v Speaker 1>thanks to Brendan Francis Newnham and our executive editor Catherine

0:44:48.600 --> 0:44:54.320
<v Speaker 1>Bell and until next time, Goodbye,