1 00:00:08,840 --> 00:00:11,319 Speaker 1: It's eleven am in the newsroom of the Bourne Myth 2 00:00:11,440 --> 00:00:14,600 Speaker 1: Daily Echo, a local newspaper in the south of England, 3 00:00:15,920 --> 00:00:25,440 Speaker 1: One More Lucky, which which is a very strong story. 4 00:00:31,240 --> 00:00:33,519 Speaker 1: The editors are trying to decide which stories to run 5 00:00:33,520 --> 00:00:36,479 Speaker 1: in the newspaper and on the echoes website, which these 6 00:00:36,560 --> 00:00:42,479 Speaker 1: days might be more important. Technology is changing all of 7 00:00:42,520 --> 00:00:45,200 Speaker 1: our jobs, whether we like it or not, and journalism 8 00:00:45,240 --> 00:00:48,240 Speaker 1: is certainly no exception. There are now far fewer jobs 9 00:00:48,240 --> 00:00:51,320 Speaker 1: in local news right to put them mildly, the internet 10 00:00:51,360 --> 00:00:54,760 Speaker 1: has pretty much busted the media business model. Yeah, and 11 00:00:54,840 --> 00:00:56,880 Speaker 1: journalists they are left. I've had to learn new skills. 12 00:00:58,000 --> 00:01:01,000 Speaker 1: When you're in you have to you you adapt or 13 00:01:01,040 --> 00:01:04,520 Speaker 1: you don't surve. The Echo has worked hard to take 14 00:01:04,560 --> 00:01:07,759 Speaker 1: advantage of new technology. But is the editor Andy Martin 15 00:01:07,880 --> 00:01:11,800 Speaker 1: told you, Jeremy, the paper is struggling financially. You know 16 00:01:11,920 --> 00:01:15,080 Speaker 1: me know that revenues walked out the door to a 17 00:01:15,120 --> 00:01:16,920 Speaker 1: lot of other places in the last ten years, and 18 00:01:17,440 --> 00:01:20,039 Speaker 1: we haven't quite got our business model right in terms of, 19 00:01:21,160 --> 00:01:27,040 Speaker 1: you know, getting people to pay for online news. Yeah, 20 00:01:27,040 --> 00:01:29,759 Speaker 1: that's right. Faced with collapsing budgets, the Echo has had 21 00:01:29,800 --> 00:01:33,120 Speaker 1: to let go about of its staff that's made it 22 00:01:33,160 --> 00:01:35,440 Speaker 1: harder for the paper to fulfill its traditional role of 23 00:01:35,480 --> 00:01:38,400 Speaker 1: holding people in power to account. Now, the founders of 24 00:01:38,440 --> 00:01:41,080 Speaker 1: a service funded by Google say they can help local 25 00:01:41,160 --> 00:01:45,080 Speaker 1: journalists produce more stories without adding more people, but it's 26 00:01:45,120 --> 00:02:01,000 Speaker 1: raising new fears even as it tries to solve full problems. Hi, 27 00:02:01,080 --> 00:02:05,000 Speaker 1: I'm brad Stone and I'm Jeremy con and this week Undercrypted, 28 00:02:05,000 --> 00:02:07,520 Speaker 1: we're taking a look at how one local newspaper is 29 00:02:07,560 --> 00:02:10,880 Speaker 1: grappling with new technology and its struggle to survive. There 30 00:02:10,880 --> 00:02:13,280 Speaker 1: are only about twenty editorial staff left now at the 31 00:02:13,280 --> 00:02:15,799 Speaker 1: Bournemoth Daily Echo, a paper that want it close to 32 00:02:15,840 --> 00:02:19,040 Speaker 1: a hundred journalists when Andy, the editor, first started. That's 33 00:02:19,040 --> 00:02:21,760 Speaker 1: been a painful transition for The Echo, and one that's 34 00:02:21,760 --> 00:02:24,760 Speaker 1: played out around the world as journalists have watched jobs disappear. 35 00:02:25,480 --> 00:02:28,840 Speaker 1: Artificial intelligence is one technology that could help newsrooms like 36 00:02:28,880 --> 00:02:32,799 Speaker 1: The Echo survived with shrinking budgets, but could automation, which 37 00:02:32,840 --> 00:02:36,440 Speaker 1: initially looks like an added convenience, ultimately end up taking 38 00:02:36,440 --> 00:03:00,079 Speaker 1: even more jobs stay with us that deminate So jar me, 39 00:03:00,160 --> 00:03:03,120 Speaker 1: you actually went down to board with recently. Yeah, I 40 00:03:03,320 --> 00:03:05,760 Speaker 1: took the train down from London on a chilly morning 41 00:03:05,760 --> 00:03:09,560 Speaker 1: in February. So I'm imagining sort of the Cleveland of England. 42 00:03:09,600 --> 00:03:12,760 Speaker 1: What's born With like, Well, it's not quite Cleveland. Um. 43 00:03:12,760 --> 00:03:14,480 Speaker 1: It's on the south coast of England and it's a 44 00:03:14,560 --> 00:03:18,040 Speaker 1: summer tourist destination, has a boardwalk and a fair ground 45 00:03:18,080 --> 00:03:20,640 Speaker 1: along the seafront. Um. But the city also hosts the 46 00:03:20,639 --> 00:03:23,440 Speaker 1: back offices of some big banks like JP Morgan, and 47 00:03:23,480 --> 00:03:26,320 Speaker 1: it's got the headquarters of a few industrial companies and 48 00:03:26,360 --> 00:03:29,799 Speaker 1: the born With Echo that's the region's main newspaper. Yeah, 49 00:03:29,840 --> 00:03:31,919 Speaker 1: that's right, and it's published from this big Art Deco 50 00:03:32,000 --> 00:03:33,920 Speaker 1: building with a small clock tower in the heart of 51 00:03:33,919 --> 00:03:40,240 Speaker 1: the city. And he did, hey, Jerry, Yeah, Andy, the 52 00:03:40,280 --> 00:03:42,720 Speaker 1: Echoes editor came out to meet me. He looks like 53 00:03:42,720 --> 00:03:45,280 Speaker 1: a grizzled newspaper writer you've seen a film, with the 54 00:03:45,360 --> 00:03:48,160 Speaker 1: steely gaze and a shock of white hair. He led 55 00:03:48,200 --> 00:03:50,160 Speaker 1: me through the bowels of the building and up into 56 00:03:50,160 --> 00:03:52,560 Speaker 1: a big conference room with glass windows that look out 57 00:03:52,560 --> 00:03:55,960 Speaker 1: over the Echoes newsroom. How many years have you you've 58 00:03:56,000 --> 00:03:59,880 Speaker 1: been here at the Echo? Well, this embarrassing. I'm being 59 00:03:59,880 --> 00:04:04,640 Speaker 1: here about thirty years in April. Wow, Um, what was 60 00:04:04,680 --> 00:04:08,440 Speaker 1: your first day here? Like? I can really remembered the 61 00:04:08,440 --> 00:04:11,640 Speaker 1: old style smoke field newsrooms, people you know, slamming your 62 00:04:11,640 --> 00:04:14,680 Speaker 1: phones down and the saner time writers at the time 63 00:04:14,720 --> 00:04:17,880 Speaker 1: in the late eighties, the newsroom was big. Years ago, 64 00:04:17,960 --> 00:04:20,080 Speaker 1: there would have been a hundred editorial staff here. We 65 00:04:20,160 --> 00:04:21,840 Speaker 1: used to have the print depressors. You hear the wrong 66 00:04:21,920 --> 00:04:24,040 Speaker 1: of the primes going up for the first edition of 67 00:04:24,040 --> 00:04:25,400 Speaker 1: the second edition to go down and get a couple 68 00:04:25,400 --> 00:04:26,960 Speaker 1: of paper. We get your boak in sandwich come back 69 00:04:26,960 --> 00:04:29,680 Speaker 1: out of paper. This was a different time for the 70 00:04:29,720 --> 00:04:32,719 Speaker 1: newspaper industry. The born With Daily Echo had a small 71 00:04:32,839 --> 00:04:35,479 Speaker 1: number of district offices and it would send reporters like 72 00:04:35,520 --> 00:04:40,680 Speaker 1: Andy on international assignments too, for about twenty years, so Cyprus, Kossovo, 73 00:04:41,200 --> 00:04:47,080 Speaker 1: the Balkans, Bosnia, Belize, Northern Ireland. So I'd got a 74 00:04:47,080 --> 00:04:49,760 Speaker 1: lot of tours of you know, not tours of duty, 75 00:04:49,800 --> 00:04:51,840 Speaker 1: but at at with the boys, you know, sort of 76 00:04:51,920 --> 00:04:54,840 Speaker 1: week two weeks long. After a few years he became 77 00:04:54,839 --> 00:04:57,440 Speaker 1: an editor on the news desk and as Andy was 78 00:04:57,440 --> 00:05:00,120 Speaker 1: climbing the mast out of the Echo, something happened the 79 00:05:00,120 --> 00:05:03,560 Speaker 1: mid nineteen nineties that would change print journalism forever. The 80 00:05:03,600 --> 00:05:07,080 Speaker 1: Internet back in the glory days local papers had four 81 00:05:07,120 --> 00:05:12,640 Speaker 1: main revenue streams newsstand sales, subscriptions, display, advertising and classified. 82 00:05:13,120 --> 00:05:17,160 Speaker 1: The Internet undercut all four. For example, news Quest Media, 83 00:05:17,360 --> 00:05:20,320 Speaker 1: which owns the Bournemouth Daily Echo, made a profit of 84 00:05:20,320 --> 00:05:23,000 Speaker 1: a hundred and sixty seven million pounds in two thousand 85 00:05:23,000 --> 00:05:27,279 Speaker 1: and three. By that number was just twenty three million 86 00:05:27,279 --> 00:05:30,400 Speaker 1: pounds ouch, so one seventh of what it once was. 87 00:05:31,080 --> 00:05:33,920 Speaker 1: That's right. And as profits slumped, the Echo has had 88 00:05:33,920 --> 00:05:38,240 Speaker 1: to shed staff. So you're talking reporters, you're talking photographers. 89 00:05:38,279 --> 00:05:41,239 Speaker 1: Some of those cuts happened through attrition, people who left 90 00:05:41,320 --> 00:05:44,560 Speaker 1: or retired simply weren't replaced, but there have also been 91 00:05:44,680 --> 00:05:48,920 Speaker 1: deep rounds of layoffs photographers, probably five or six reporters, 92 00:05:49,080 --> 00:05:53,000 Speaker 1: a couple of support staff. Um. So yeah, that's been 93 00:05:53,120 --> 00:05:56,359 Speaker 1: it's been the toughest part of the job, um, in 94 00:05:56,400 --> 00:06:02,320 Speaker 1: the last few years. I've been here from very long time. 95 00:06:02,400 --> 00:06:07,279 Speaker 1: I started at ninety fours up twenty three years. Corn 96 00:06:07,360 --> 00:06:09,960 Speaker 1: Messer joined the Echo as a dark room assistant. At 97 00:06:10,000 --> 00:06:13,640 Speaker 1: that time the paper had six staff photographers. Now Corren 98 00:06:13,720 --> 00:06:17,200 Speaker 1: is the only one left. The last he's seen the 99 00:06:17,240 --> 00:06:22,760 Speaker 1: staff was left in September, which increases one of those. 100 00:06:22,920 --> 00:06:26,120 Speaker 1: He was here the day that I started. He you know, 101 00:06:26,920 --> 00:06:30,880 Speaker 1: I gave the speeches leaving the was that that was very, 102 00:06:30,960 --> 00:06:34,320 Speaker 1: very upsetting. Often Andy has been forced to let go 103 00:06:34,360 --> 00:06:37,320 Speaker 1: of staff he's been working alongside for ten fifteen years, 104 00:06:37,839 --> 00:06:40,360 Speaker 1: and his own role has had to expand to fill 105 00:06:40,480 --> 00:06:43,880 Speaker 1: some of the gaps. Well, I'm the editor, I'm also 106 00:06:43,920 --> 00:06:46,960 Speaker 1: the head of news, but i'm also well today I'm 107 00:06:47,080 --> 00:06:49,560 Speaker 1: essentially the content manager. So I'm doing all the pages today. 108 00:06:49,920 --> 00:06:52,600 Speaker 1: So it's um this sort of about four or five 109 00:06:52,680 --> 00:06:55,279 Speaker 1: jobs rolled up in one to be honest, these days, 110 00:06:55,600 --> 00:06:59,040 Speaker 1: in a in a regional news room. Despite that, Andy 111 00:06:59,080 --> 00:07:01,560 Speaker 1: has tried hard not to cut essentral areas of coverage. 112 00:07:02,279 --> 00:07:05,360 Speaker 1: Role first and foremost. First and foremost is the whole 113 00:07:05,360 --> 00:07:10,320 Speaker 1: people's account, whether it's hospitals, the police, fire service, local authorities, 114 00:07:10,760 --> 00:07:13,880 Speaker 1: any public institution. That's what we're here for. But he 115 00:07:13,920 --> 00:07:16,840 Speaker 1: admits there are some things he can't do anymore. You know, 116 00:07:16,880 --> 00:07:19,640 Speaker 1: we can't now say to someone, Okay, go off and 117 00:07:19,680 --> 00:07:22,320 Speaker 1: research that, investigate that story for two or three days. 118 00:07:22,320 --> 00:07:25,440 Speaker 1: Be off diary and at the echo, as it's so 119 00:07:25,480 --> 00:07:28,560 Speaker 1: many newspapers, the threat of more layoffs still looms in 120 00:07:28,600 --> 00:07:32,000 Speaker 1: the background. You know what what what's the next thing 121 00:07:32,040 --> 00:07:33,400 Speaker 1: that's going to be thrown at me, you will have 122 00:07:33,440 --> 00:07:35,960 Speaker 1: to make more savings. So I think it's kind of 123 00:07:36,000 --> 00:07:38,240 Speaker 1: just made. I think that keeps me awakenizes really keeping 124 00:07:38,240 --> 00:07:50,320 Speaker 1: all the plates spinning and the finances. I think Rooney Now. 125 00:07:50,400 --> 00:07:53,080 Speaker 1: A few months ago, one of Andy's bosses told him 126 00:07:53,080 --> 00:07:56,440 Speaker 1: about a new type of technology that could perhaps help. 127 00:07:56,720 --> 00:07:59,680 Speaker 1: It's an intelligent software that could provide Andy the kind 128 00:07:59,680 --> 00:08:02,400 Speaker 1: of low news stories that he desperately needs, but that 129 00:08:02,480 --> 00:08:05,080 Speaker 1: in recent years he hasn't really had the staff to report. 130 00:08:05,120 --> 00:08:09,080 Speaker 1: And right at first, Andy was wary. Until now technology 131 00:08:09,120 --> 00:08:10,760 Speaker 1: has been more of a job killer for the news 132 00:08:10,800 --> 00:08:14,400 Speaker 1: industry than anything else. Well, I mean, you know, I'm 133 00:08:14,680 --> 00:08:18,200 Speaker 1: obviously naturally skeptical about everything and even in our own industry, 134 00:08:18,760 --> 00:08:22,160 Speaker 1: but Andy was also curious. I'm always open to looking 135 00:08:22,160 --> 00:08:26,560 Speaker 1: at new ideas of new ways of getting information. The 136 00:08:26,600 --> 00:08:29,360 Speaker 1: Echo has had no choice but to embrace digital technology 137 00:08:29,360 --> 00:08:32,559 Speaker 1: and social media in recent years, and now it's become 138 00:08:32,600 --> 00:08:35,280 Speaker 1: one of the first British newspapers to participate in something 139 00:08:35,320 --> 00:08:38,079 Speaker 1: called Project Radar. That's how I ended up in another 140 00:08:38,120 --> 00:08:43,000 Speaker 1: news organization, this time in London. The Press Association was 141 00:08:43,000 --> 00:08:46,520 Speaker 1: built to provide national news coverage for the UK's local papers. 142 00:08:46,880 --> 00:08:49,600 Speaker 1: Here in Britain people call it the p A. I 143 00:08:49,640 --> 00:08:52,400 Speaker 1: went to see Peter Clifton, the p a's editor in chief. 144 00:08:53,000 --> 00:08:55,280 Speaker 1: The PA is facing its own battle to adapt to 145 00:08:55,400 --> 00:08:58,520 Speaker 1: changing times. Now that most people get their news online, 146 00:08:58,760 --> 00:09:02,320 Speaker 1: local papers need less national coverage. I think they've all 147 00:09:03,000 --> 00:09:07,360 Speaker 1: become increasingly focused on trying to serve their local audience 148 00:09:07,400 --> 00:09:10,920 Speaker 1: with more and more local content. Then through in All 149 00:09:10,960 --> 00:09:14,120 Speaker 1: the Queens, Peter heard about a London startup called Herbs 150 00:09:14,200 --> 00:09:17,360 Speaker 1: Media that was founded by two news industry veterans, Alan 151 00:09:17,440 --> 00:09:21,680 Speaker 1: Renwick and Gary Rogers. Here's Gary. I started my career 152 00:09:22,000 --> 00:09:23,679 Speaker 1: in local newspapers and I did that for three years 153 00:09:23,679 --> 00:09:26,559 Speaker 1: of my life and then took a attorney to television. 154 00:09:27,000 --> 00:09:30,000 Speaker 1: Between them, they've got more than fifty five years of experience. 155 00:09:31,240 --> 00:09:34,840 Speaker 1: The consumption of news has changed dramatically and radically, but 156 00:09:34,920 --> 00:09:37,960 Speaker 1: if you looked at the business of how news is produced, 157 00:09:38,320 --> 00:09:43,040 Speaker 1: actually very little has changed. It still requires a journalist 158 00:09:43,360 --> 00:09:46,120 Speaker 1: to go and find a story and go and meet people, 159 00:09:46,320 --> 00:09:48,839 Speaker 1: and it requires phone time and shoe leather, and good 160 00:09:48,880 --> 00:09:54,040 Speaker 1: journalism requires that. In Alan and Gary had a brain wave, 161 00:09:54,440 --> 00:09:56,440 Speaker 1: but we were looking at whether there was another way 162 00:09:56,480 --> 00:10:00,480 Speaker 1: to produce good journalism that would be a little less 163 00:10:00,679 --> 00:10:04,400 Speaker 1: labor intensive. They decided to focus on large government data 164 00:10:04,440 --> 00:10:06,680 Speaker 1: sets that had figures on things like bike theft and 165 00:10:06,760 --> 00:10:10,640 Speaker 1: childhood obesity, and so we were quite interested in sets 166 00:10:10,679 --> 00:10:13,240 Speaker 1: of crime data at the time. They thought they could 167 00:10:13,240 --> 00:10:15,480 Speaker 1: take a single data set and get dozens or even 168 00:10:15,559 --> 00:10:18,439 Speaker 1: hundreds of local stories from it. I can't remember whether 169 00:10:18,520 --> 00:10:20,920 Speaker 1: it was me or Alan come up with this smart 170 00:10:21,040 --> 00:10:25,800 Speaker 1: idea of why didn't we do a snapshot crime profile 171 00:10:25,880 --> 00:10:29,440 Speaker 1: of every borough. Borough is a city district, London has 172 00:10:29,480 --> 00:10:32,160 Speaker 1: thirty three of them. So we started with the humans 173 00:10:32,720 --> 00:10:37,160 Speaker 1: otherwise known as Me and Garret. But the amount of 174 00:10:37,240 --> 00:10:39,920 Speaker 1: data they had to crawl through soon became a problem, 175 00:10:39,960 --> 00:10:42,760 Speaker 1: and I wrote about ten and I started to lose 176 00:10:42,800 --> 00:10:46,200 Speaker 1: the will to live. Each story was basically the same, 177 00:10:46,559 --> 00:10:50,480 Speaker 1: but every neighborhood had different numbers. It became extremely repetitive 178 00:10:50,600 --> 00:10:55,440 Speaker 1: and very boring. We both began to think there must 179 00:10:55,440 --> 00:10:58,040 Speaker 1: be an easier way to do this, and we came 180 00:10:58,080 --> 00:11:01,360 Speaker 1: across something called natural language generation, and as a way 181 00:11:01,400 --> 00:11:06,880 Speaker 1: to try to turn data into texts. Natural language generation 182 00:11:07,040 --> 00:11:09,440 Speaker 1: is a kind of artificial intelligence that can write whole 183 00:11:09,480 --> 00:11:13,040 Speaker 1: reports based on data the way Gary and Allen have 184 00:11:13,120 --> 00:11:16,080 Speaker 1: designed their program, a human journalist still needs to write 185 00:11:16,120 --> 00:11:19,160 Speaker 1: the basic template for a particular data set. Then the 186 00:11:19,200 --> 00:11:21,559 Speaker 1: software fills in numbers and tweets the language so that 187 00:11:21,600 --> 00:11:24,800 Speaker 1: each story can make sense. I don't know, Jeremy, can 188 00:11:24,840 --> 00:11:28,560 Speaker 1: the software really convey the nuance and the perspective of 189 00:11:28,559 --> 00:11:32,679 Speaker 1: of an actual human journalist. Well, these stories that will 190 00:11:32,679 --> 00:11:34,800 Speaker 1: see later are kind of dry. I mean, they're they're pretty, 191 00:11:34,800 --> 00:11:37,280 Speaker 1: They're written pretty straight. That sounds like it can't potentially 192 00:11:37,280 --> 00:11:40,520 Speaker 1: solve a problem for newspapers. Yeah. The idea here, and 193 00:11:40,559 --> 00:11:42,440 Speaker 1: what Peter thought is that this new way of writing 194 00:11:42,480 --> 00:11:45,920 Speaker 1: stories might help his newsroom create more local stories. Um 195 00:11:46,000 --> 00:11:48,160 Speaker 1: and he decided he was going to team up with Herbs, 196 00:11:48,320 --> 00:11:50,880 Speaker 1: and together they created this thing called Project radar. Okay, 197 00:11:50,880 --> 00:11:54,040 Speaker 1: why project radar, Well, it turns out it's an acronym, 198 00:11:54,040 --> 00:11:56,000 Speaker 1: not a very great one. It stands for reporters and 199 00:11:56,120 --> 00:12:00,320 Speaker 1: data and robots. The project received a seven thousand euro 200 00:12:00,440 --> 00:12:03,920 Speaker 1: grant from Google's European Digital News Initiative to help it 201 00:12:03,960 --> 00:12:08,240 Speaker 1: get off the ground. Interestingly, Peter's own team got nervous 202 00:12:08,240 --> 00:12:11,720 Speaker 1: about this robot taking their jobs. I am probably and 203 00:12:11,800 --> 00:12:15,640 Speaker 1: wisely made a speech at the Society of Editor's conference 204 00:12:15,760 --> 00:12:19,000 Speaker 1: last year where I this This was only really a 205 00:12:19,040 --> 00:12:20,960 Speaker 1: glimmer at that point, but all I said at that 206 00:12:21,920 --> 00:12:26,240 Speaker 1: conference was that we would be looking at automation. One 207 00:12:26,280 --> 00:12:30,040 Speaker 1: of the industry websites immediately had a headline of about 208 00:12:30,400 --> 00:12:33,160 Speaker 1: robots taking over at p A And by the time 209 00:12:33,200 --> 00:12:36,120 Speaker 1: I got back to work, they've built a robot cardboard 210 00:12:36,360 --> 00:12:41,960 Speaker 1: cardboard boxes and put it in my seat. Peter and 211 00:12:42,000 --> 00:12:44,760 Speaker 1: the guy at HERBS insisted that nobody's job is going 212 00:12:44,800 --> 00:12:47,800 Speaker 1: to be lost because of Project Radar. Just like your 213 00:12:47,800 --> 00:12:50,320 Speaker 1: world process. I can just sit on your desk as 214 00:12:50,360 --> 00:12:53,560 Speaker 1: a tool you used to write what Gary is saying. 215 00:12:53,559 --> 00:12:56,120 Speaker 1: It is probably true today. I wonder they'll cheremy if 216 00:12:56,120 --> 00:13:00,240 Speaker 1: it will be true tomorrow as the technology inevitably evolves. Yeah, 217 00:13:00,280 --> 00:13:02,440 Speaker 1: that's that is how you consider this kinde of technology 218 00:13:02,559 --> 00:13:04,880 Speaker 1: has a tendency to get more capable over time, and 219 00:13:04,880 --> 00:13:08,440 Speaker 1: what it can do it keeps improving. Okay, well, after 220 00:13:08,480 --> 00:13:11,960 Speaker 1: the break, the Bournemouth Daily Echo starts receiving stories from 221 00:13:11,960 --> 00:13:14,760 Speaker 1: Project Greater and it raises an old fear for Andy 222 00:13:14,800 --> 00:13:33,040 Speaker 1: as he starts trying to use the automated stories. Back 223 00:13:33,080 --> 00:13:35,199 Speaker 1: in Bournemouth. I walked through the video arcade on the 224 00:13:35,240 --> 00:13:38,760 Speaker 1: Bournemouth Pierre Jermy, why were you in a video arcade? Well, 225 00:13:38,840 --> 00:13:41,000 Speaker 1: that's a good question. These arcades tend to be a 226 00:13:41,000 --> 00:13:43,680 Speaker 1: fixture of British seaside towns and I was actually on 227 00:13:43,720 --> 00:13:47,000 Speaker 1: my way to visit another beloved British institution, the local 228 00:13:47,000 --> 00:13:52,040 Speaker 1: fish and chip shop. My name is Chloe and I'm 229 00:13:52,080 --> 00:13:54,920 Speaker 1: the assistant manager here Harry Robson's. And it says on 230 00:13:54,960 --> 00:13:56,880 Speaker 1: the science of world famous fish and chip shop? What 231 00:13:56,960 --> 00:13:59,560 Speaker 1: is it? What is it famous for? It's famous for 232 00:13:59,720 --> 00:14:03,040 Speaker 1: us secret recipe that we have hair in staff, which 233 00:14:03,040 --> 00:14:05,319 Speaker 1: you can't tell me what it is, I suppose unfortunately 234 00:14:05,360 --> 00:14:08,199 Speaker 1: I can't. Chloe said she's a loyal reader of the 235 00:14:08,240 --> 00:14:11,400 Speaker 1: Bournemouth Echo and she especially likes the traffic updates on 236 00:14:11,440 --> 00:14:14,600 Speaker 1: its website, especially with the amount of roadworks that we 237 00:14:14,600 --> 00:14:16,480 Speaker 1: haven't borne from the moment. The online one is always 238 00:14:16,520 --> 00:14:18,240 Speaker 1: up to date, so you knew where they are, what 239 00:14:18,280 --> 00:14:20,120 Speaker 1: times are happening and stuff like that. So is that 240 00:14:20,160 --> 00:14:22,400 Speaker 1: a big issue of roadworks and congestion? And I think 241 00:14:22,640 --> 00:14:24,680 Speaker 1: at the moment yes, because there's quite a lot so, 242 00:14:24,960 --> 00:14:28,720 Speaker 1: especially recently, the Bournemouth Echo ran a storytelling readers just 243 00:14:28,800 --> 00:14:31,360 Speaker 1: how much of their lives were wasted in traffic jams 244 00:14:31,360 --> 00:14:34,000 Speaker 1: in an average year. So the congestion one, which showed 245 00:14:34,080 --> 00:14:36,800 Speaker 1: us just how long emotives would spend in a traffic 246 00:14:36,880 --> 00:14:39,880 Speaker 1: jam in an average week or an average year in 247 00:14:39,960 --> 00:14:43,720 Speaker 1: the context of the restructure of the local area, is 248 00:14:43,720 --> 00:14:46,520 Speaker 1: one that was particularly interesting. That story, and he's talking 249 00:14:46,560 --> 00:14:49,440 Speaker 1: about it came from Project Radar. I think it's been 250 00:14:49,440 --> 00:14:52,880 Speaker 1: a help um in terms of, you know, statistics that 251 00:14:52,920 --> 00:14:56,280 Speaker 1: we wouldn't ordinarily be able to access or have have 252 00:14:56,360 --> 00:14:59,760 Speaker 1: the time to assess. I guess, But he says, some 253 00:15:00,040 --> 00:15:03,520 Speaker 1: Project Radars computer generated stories still require a bit more work. 254 00:15:04,000 --> 00:15:06,840 Speaker 1: I wouldn't necessarily say yes, I think that's been written 255 00:15:06,840 --> 00:15:09,040 Speaker 1: by a computer. I would just say it's been written 256 00:15:09,080 --> 00:15:11,840 Speaker 1: really straight. So he might ask one of his staff 257 00:15:11,880 --> 00:15:14,840 Speaker 1: to get reaction quotes from a local official or from residents, 258 00:15:15,360 --> 00:15:17,440 Speaker 1: or to punch up the writing to make it more lively. 259 00:15:18,120 --> 00:15:19,920 Speaker 1: Should we do you have an example about one of 260 00:15:19,920 --> 00:15:23,960 Speaker 1: these Project Radar stories. Actually sounds like yeah, sure, here's 261 00:15:23,960 --> 00:15:26,520 Speaker 1: one about juvenile crime that Project Radar produced, and the 262 00:15:26,560 --> 00:15:29,840 Speaker 1: Bourne Meth Echo actually ran. Young people in Pool are 263 00:15:29,840 --> 00:15:32,400 Speaker 1: significantly less likely to be cautioned or convicted for a 264 00:15:32,440 --> 00:15:35,040 Speaker 1: first offense than they were ten years ago. Changes in 265 00:15:35,080 --> 00:15:38,400 Speaker 1: police policy and overall fallen crime has seen an eighty 266 00:15:38,440 --> 00:15:40,760 Speaker 1: four percent drop in the number of youngsters entering the 267 00:15:40,800 --> 00:15:44,480 Speaker 1: criminal justice system. According to statistics from the Ministry of Justice. 268 00:15:44,960 --> 00:15:47,840 Speaker 1: In the two thousand and six oh seven financial year, 269 00:15:48,080 --> 00:15:50,880 Speaker 1: two fifty two children between the ages of ten and 270 00:15:51,000 --> 00:15:53,960 Speaker 1: seventeen were convicted or cautioned by police for the first time, 271 00:15:54,400 --> 00:15:58,680 Speaker 1: but by seventeen there were just forty one. We are 272 00:15:58,720 --> 00:16:01,360 Speaker 1: well should have had you read that in your Alexa voice, 273 00:16:01,600 --> 00:16:05,840 Speaker 1: But you know it's it's interesting, but certainly not brimming 274 00:16:05,880 --> 00:16:09,360 Speaker 1: with writerly sensibilities. But it's interesting because this is probably 275 00:16:09,360 --> 00:16:11,800 Speaker 1: not the kind of thing that the Andy Ors reporters 276 00:16:11,800 --> 00:16:16,600 Speaker 1: could have done on their own. Okay, so Andy is 277 00:16:16,680 --> 00:16:19,760 Speaker 1: using these stories of the Bournemouth Echo and finding them useful, 278 00:16:20,160 --> 00:16:22,320 Speaker 1: Yeah he is, But out of a d twenty five 279 00:16:22,360 --> 00:16:25,000 Speaker 1: stories that they publish every week, only about to come 280 00:16:25,040 --> 00:16:27,880 Speaker 1: from Project Radar. Oh wow, Okay, so that's not much 281 00:16:27,880 --> 00:16:30,360 Speaker 1: at all. No, and Andy told me he's worried about 282 00:16:30,440 --> 00:16:33,040 Speaker 1: using the service too much. Well, we to be looking 283 00:16:33,080 --> 00:16:35,440 Speaker 1: at doing them, or that it would it would It 284 00:16:35,480 --> 00:16:38,320 Speaker 1: would for people if you use too many of them, 285 00:16:38,320 --> 00:16:41,280 Speaker 1: that you'd worry that your bosses or somebody would say 286 00:16:41,360 --> 00:16:44,480 Speaker 1: you may not need that many staff be a little 287 00:16:44,480 --> 00:16:47,480 Speaker 1: bit a little bit worried about about them. So even 288 00:16:47,480 --> 00:16:50,320 Speaker 1: though the p A and Herbs insists the Project Radar 289 00:16:50,440 --> 00:16:53,280 Speaker 1: is about helping us do our jobs better, Andy is 290 00:16:53,320 --> 00:16:56,280 Speaker 1: still worried that in his newsroom this technology could take 291 00:16:56,280 --> 00:16:59,120 Speaker 1: away even more jobs. Yeah. I think that's always a 292 00:16:59,120 --> 00:17:01,760 Speaker 1: concern with AI because it gets better and better all 293 00:17:01,800 --> 00:17:04,280 Speaker 1: the time. As we said earlier, how good is it 294 00:17:04,359 --> 00:17:08,359 Speaker 1: the writing podcast scripts, Lucky for us, is not that 295 00:17:08,400 --> 00:17:10,400 Speaker 1: good yet. But the fact is this is a very 296 00:17:10,400 --> 00:17:13,399 Speaker 1: new and interesting area of technology and it's maturing fast. 297 00:17:13,640 --> 00:17:16,359 Speaker 1: But the specter of even more job cuts hasn't deterred 298 00:17:16,400 --> 00:17:20,680 Speaker 1: the news industry from experimenting with computer generated articles. Lots 299 00:17:20,680 --> 00:17:23,399 Speaker 1: of organizations are trying versions of this now, like the 300 00:17:23,440 --> 00:17:25,919 Speaker 1: Associated Press here in the US, which is working with 301 00:17:25,960 --> 00:17:29,760 Speaker 1: a company called Automated Insights on writing simple financial stories. 302 00:17:30,080 --> 00:17:32,600 Speaker 1: That's right, and the Washington Post has a software called 303 00:17:32,640 --> 00:17:36,000 Speaker 1: Heliograph to write stories on everything from high school football 304 00:17:36,000 --> 00:17:39,040 Speaker 1: games to US congressional elections. And we should also say 305 00:17:39,040 --> 00:17:41,800 Speaker 1: that even here at Bloomberg, our own team of Quotas 306 00:17:41,840 --> 00:17:44,720 Speaker 1: has created software that can take corporate earnings reports or 307 00:17:44,760 --> 00:17:49,160 Speaker 1: stock price movements and automatically generate news stories from them. Yeah. 308 00:17:49,160 --> 00:17:51,840 Speaker 1: It's pretty crazy what this technology can do. But I'm 309 00:17:51,880 --> 00:17:53,320 Speaker 1: pretty happy we don't have to write some of those 310 00:17:53,320 --> 00:17:56,560 Speaker 1: earning stories anymore. Yeah, me too. And actually, on that point, 311 00:17:56,840 --> 00:17:59,280 Speaker 1: I mean, I hope I'm not expressing the hubrists of 312 00:17:59,320 --> 00:18:02,920 Speaker 1: a human here, but I do wonder if the AI 313 00:18:03,000 --> 00:18:06,119 Speaker 1: can kind of replace the you know, the creativity and 314 00:18:06,160 --> 00:18:10,119 Speaker 1: the storytelling of an actual journalist. Yeah. I mean, I 315 00:18:10,160 --> 00:18:12,320 Speaker 1: think there's a lot that these these types of programs 316 00:18:12,359 --> 00:18:13,919 Speaker 1: are never going to be able to do. They're not 317 00:18:13,960 --> 00:18:16,280 Speaker 1: really going to be able to do investigative reporting and 318 00:18:16,400 --> 00:18:19,240 Speaker 1: narrative storytelling, right articles with the beginning of middle and 319 00:18:19,280 --> 00:18:21,919 Speaker 1: an end and that the you know, convey some nuance 320 00:18:21,960 --> 00:18:25,360 Speaker 1: about a character. Yeah, I think that would be very difficult. 321 00:18:25,359 --> 00:18:27,280 Speaker 1: It's certainly the way some of these programs are set up. 322 00:18:27,520 --> 00:18:29,480 Speaker 1: But they are getting more sophisticated all the time. And 323 00:18:29,480 --> 00:18:31,720 Speaker 1: I know that there are some researchers working on on 324 00:18:31,760 --> 00:18:34,639 Speaker 1: this type of technology and academia who think you could create, 325 00:18:34,760 --> 00:18:37,080 Speaker 1: you know, a narrative structure, at least teach these programs 326 00:18:37,119 --> 00:18:39,840 Speaker 1: to create a narrative structure. Now you're getting a little 327 00:18:39,880 --> 00:18:42,560 Speaker 1: too close to home. But let me ask you, Jeremy, 328 00:18:42,600 --> 00:18:44,680 Speaker 1: I mean, is there a reason to worry that if 329 00:18:44,720 --> 00:18:47,400 Speaker 1: the AI is doing the kind of grunt work stories, 330 00:18:47,800 --> 00:18:51,520 Speaker 1: that some perhaps young reporters aren't getting the kind of training, 331 00:18:52,000 --> 00:18:53,680 Speaker 1: you know, being able to cut their teeth on these 332 00:18:53,760 --> 00:18:56,800 Speaker 1: kinds of stories because the AI is taking over. Yeah, 333 00:18:56,840 --> 00:18:58,040 Speaker 1: I mean, I think it's a bit of a double 334 00:18:58,080 --> 00:19:00,359 Speaker 1: edged sword. On the one hand, it means that those 335 00:19:00,440 --> 00:19:02,439 Speaker 1: those entry level reporters don't have to do a lot 336 00:19:02,480 --> 00:19:03,960 Speaker 1: of that grunt work, and maybe they don't get his 337 00:19:04,040 --> 00:19:06,240 Speaker 1: board with the job and maybe they get to move 338 00:19:06,240 --> 00:19:09,639 Speaker 1: on to higher value added skills earlier. But but they 339 00:19:09,680 --> 00:19:12,080 Speaker 1: also might not learn the basics. And I think that's 340 00:19:12,080 --> 00:19:14,240 Speaker 1: an issue in a lot of professions as AI kind 341 00:19:14,240 --> 00:19:16,200 Speaker 1: of takes over, is you know, all of that stuff 342 00:19:16,240 --> 00:19:18,800 Speaker 1: that's being automated was how people learn the basics of 343 00:19:18,840 --> 00:19:21,040 Speaker 1: their job, And there were a lot of important skills 344 00:19:21,040 --> 00:19:22,720 Speaker 1: that they might have learned that they're now gonna have 345 00:19:22,720 --> 00:19:25,600 Speaker 1: to learn some other way. Right, But the big issue 346 00:19:25,640 --> 00:19:28,359 Speaker 1: here is of course fixing the business model does the 347 00:19:28,480 --> 00:19:30,600 Speaker 1: does Project Rador do that? I mean, do you get 348 00:19:30,600 --> 00:19:33,679 Speaker 1: the sense that this technology can solve some of the 349 00:19:33,720 --> 00:19:37,000 Speaker 1: problems at a newspaper like the Born withth Echo? Well, 350 00:19:37,240 --> 00:19:39,520 Speaker 1: I think it's helping the echo just at the margins, really, 351 00:19:39,560 --> 00:19:41,359 Speaker 1: I mean, it's helping it fill in a few stories 352 00:19:41,400 --> 00:19:44,399 Speaker 1: a week, but it's not solving that the underlying business problem. 353 00:19:44,440 --> 00:19:46,480 Speaker 1: And I don't think any of these automated soleations really are. 354 00:19:46,640 --> 00:19:48,920 Speaker 1: I mean, they help newspapers produce a little bit more 355 00:19:49,040 --> 00:19:51,520 Speaker 1: with fewer people. Um, they take away some of the 356 00:19:51,560 --> 00:19:53,720 Speaker 1: grunt work, which maybe frees up people to do some 357 00:19:53,720 --> 00:19:56,200 Speaker 1: some higher stuff. But but it's not fixing the fundamental problem, 358 00:19:56,240 --> 00:19:59,199 Speaker 1: which is if you don't have advertisers willing to pay 359 00:19:59,280 --> 00:20:01,880 Speaker 1: that much money to to run ads against something online, 360 00:20:02,200 --> 00:20:03,359 Speaker 1: you're just not gonna be able to pay for what 361 00:20:03,400 --> 00:20:06,159 Speaker 1: you produce. Yeah, and articles like the one that you 362 00:20:06,240 --> 00:20:09,760 Speaker 1: read about juvenile crime. It's probably not solving the other 363 00:20:09,800 --> 00:20:12,920 Speaker 1: basic issue, which is kind of preserving that relationship, that 364 00:20:13,119 --> 00:20:17,480 Speaker 1: very powerful relationship between the local newspaper and the reader. Yeah, no, 365 00:20:17,640 --> 00:20:19,560 Speaker 1: not at all. These programs aren't going to know about 366 00:20:19,560 --> 00:20:21,960 Speaker 1: the local community. They're not going to have that sense 367 00:20:21,960 --> 00:20:31,240 Speaker 1: of place at the Bournemouth Echo. The staff who remain 368 00:20:31,359 --> 00:20:35,080 Speaker 1: have tried their best to adapt changing times. Yeah, take 369 00:20:35,160 --> 00:20:38,320 Speaker 1: Cora and the last photographer on staff. He told us 370 00:20:38,320 --> 00:20:43,160 Speaker 1: he's moved from taking still images to mostly shooting video video. 371 00:20:43,160 --> 00:20:45,439 Speaker 1: It's been costizing, it says, it's more of a challenge, 372 00:20:45,840 --> 00:20:50,359 Speaker 1: you know, learning Learning video is something that really really 373 00:20:50,600 --> 00:20:54,520 Speaker 1: interested me. In fact, Coran's video skills were in such demand. 374 00:20:54,600 --> 00:20:57,520 Speaker 1: The Connett that's the company that owns news Quest Media, 375 00:20:58,080 --> 00:21:00,280 Speaker 1: tapped him to shoot video for USA Today and the 376 00:21:00,320 --> 00:21:03,720 Speaker 1: rest of the Ganet papers at the World Economic Forum 377 00:21:03,800 --> 00:21:06,000 Speaker 1: in Davas, Switzerland. You know, if you were going to 378 00:21:06,119 --> 00:21:09,679 Speaker 1: stay static and say I'm not learning this, I'm not 379 00:21:09,800 --> 00:21:11,439 Speaker 1: changing this is not something I'm going to do that 380 00:21:11,800 --> 00:21:15,080 Speaker 1: you are. You know, you're probably playing the last pan 381 00:21:15,119 --> 00:21:17,720 Speaker 1: in your career. Cora and has helped train the rest 382 00:21:17,760 --> 00:21:20,720 Speaker 1: of the Echoes reporting staff and photography skills, and he 383 00:21:20,800 --> 00:21:23,720 Speaker 1: oversees this thing called the Echoes Camera Club that's a 384 00:21:23,760 --> 00:21:26,840 Speaker 1: group of readers and amateur photographers who contribute images to 385 00:21:26,880 --> 00:21:30,119 Speaker 1: the paper. While he's had to adapt, Coran still enjoys 386 00:21:30,160 --> 00:21:33,000 Speaker 1: his job. Mari Hart is with the Woman Dirt Echo. 387 00:21:33,240 --> 00:21:35,199 Speaker 1: If there's if there's a massive news story going on 388 00:21:35,200 --> 00:21:36,719 Speaker 1: in town, I'm going to get some to it. If 389 00:21:36,760 --> 00:21:39,480 Speaker 1: there's if there's an interesting, you know, photographic assignment to 390 00:21:39,480 --> 00:21:41,119 Speaker 1: go to, I'm going to get to go to it. 391 00:21:41,880 --> 00:21:44,480 Speaker 1: In many ways, Andy says, the Corn's story is really 392 00:21:44,520 --> 00:21:47,480 Speaker 1: the story of the Echo in microcosm. I can't think 393 00:21:47,480 --> 00:21:50,080 Speaker 1: of any other business though, where it's such a short 394 00:21:50,119 --> 00:21:52,440 Speaker 1: space of time. People have to learn a whole range 395 00:21:52,440 --> 00:21:55,040 Speaker 1: of new skills, and they do it day and day out. 396 00:21:55,680 --> 00:21:58,760 Speaker 1: They're the phenomenon. The paper has managed to build a 397 00:21:58,760 --> 00:22:02,720 Speaker 1: decent online readership. The Echoes Content and Audience manager Katie 398 00:22:02,760 --> 00:22:05,840 Speaker 1: Clark walked me through a few of the numbers. So 399 00:22:06,280 --> 00:22:08,640 Speaker 1: monthly we're looking at like one point two to one 400 00:22:08,640 --> 00:22:13,679 Speaker 1: point three unique million uniques um and page views are 401 00:22:13,760 --> 00:22:16,439 Speaker 1: somewhere anyone, kind of anywhere monthly between like ten million 402 00:22:16,560 --> 00:22:19,720 Speaker 1: and like thirteen million, depending on how much breaking news 403 00:22:19,720 --> 00:22:23,200 Speaker 1: we've had. The Echoes sports coverage has also helped it 404 00:22:23,280 --> 00:22:26,840 Speaker 1: to attract online readers. Bournemouth has a Premier League soccer 405 00:22:26,880 --> 00:22:31,080 Speaker 1: team now Louis Cook Cook Well, that might be a 406 00:22:31,080 --> 00:22:33,880 Speaker 1: brilliant ball. It is. It's Calor Wilson, It's laid off 407 00:22:34,080 --> 00:22:43,200 Speaker 1: rid first. It's a historic win for they're nicknamed the Cherries, 408 00:22:43,640 --> 00:22:45,800 Speaker 1: and fans around the world go to the Echoes website 409 00:22:45,800 --> 00:22:48,280 Speaker 1: to follow the team. But as we said, the problem 410 00:22:48,320 --> 00:22:51,000 Speaker 1: is that advertisers pay a lot less per click than 411 00:22:51,040 --> 00:22:53,840 Speaker 1: they do with print ads, and many sites have become 412 00:22:53,920 --> 00:22:58,120 Speaker 1: dangerously dependent on third parties, especially Facebook, to drive readers 413 00:22:58,160 --> 00:23:01,440 Speaker 1: to their sites. We are looking at kind of how 414 00:23:01,440 --> 00:23:04,800 Speaker 1: we're talking to people, engaging with people on Facebook. We're 415 00:23:04,800 --> 00:23:08,119 Speaker 1: putting a lot more emphasis on Twitter coming in the 416 00:23:08,200 --> 00:23:11,479 Speaker 1: last kind of month or so. Facebook recently tweaked its 417 00:23:11,520 --> 00:23:15,159 Speaker 1: algorithm to prioritize posts from friends and family in users 418 00:23:15,160 --> 00:23:18,000 Speaker 1: news feeds. That's making it harder for publishers to use 419 00:23:18,040 --> 00:23:21,639 Speaker 1: social media to promote their stories. So you've gotten used 420 00:23:21,680 --> 00:23:25,000 Speaker 1: to seeing a slight decline last year because of the 421 00:23:25,000 --> 00:23:27,520 Speaker 1: Facebook and trying not to panic about. To be honest, 422 00:23:28,920 --> 00:23:31,560 Speaker 1: if you panicked every time Facebook changed their algorithm their 423 00:23:31,720 --> 00:23:36,479 Speaker 1: user if you're screwed. And finally, print circulation continues. It's 424 00:23:36,560 --> 00:23:40,199 Speaker 1: long decline. Meanwhile, in Bournemouth, and he says he has 425 00:23:40,240 --> 00:23:42,320 Speaker 1: no choice but to stay optimistic about the future of 426 00:23:42,359 --> 00:23:45,240 Speaker 1: journalism and the future of the Echo. To my point 427 00:23:45,240 --> 00:23:47,120 Speaker 1: of view. I love the paper, I love what we do. 428 00:23:47,520 --> 00:23:49,840 Speaker 1: I love seeking it to people and holding people's account 429 00:23:49,880 --> 00:23:55,160 Speaker 1: and you know, getting things done. Um So I think 430 00:23:55,200 --> 00:23:57,119 Speaker 1: by large you have to be in this business. You 431 00:23:57,160 --> 00:24:13,320 Speaker 1: have to be an egoistow wow. And that's it for 432 00:24:13,400 --> 00:24:17,359 Speaker 1: this week's episode of Decrypted. Thanks for listening. We always 433 00:24:17,359 --> 00:24:19,159 Speaker 1: want to know when you think of the show. You 434 00:24:19,160 --> 00:24:22,320 Speaker 1: can write to us at Decrypted at Bloomberg dot net. 435 00:24:22,760 --> 00:24:25,760 Speaker 1: I'm on Twitter at Jeremy a Con and I'm on 436 00:24:25,800 --> 00:24:29,320 Speaker 1: Twitter at Bradstone. Please consider leaving us a rating or 437 00:24:29,359 --> 00:24:32,040 Speaker 1: a review on Apple Podcasts or any of your favorite 438 00:24:32,040 --> 00:24:35,840 Speaker 1: podcast apps. It really helps us find new listeners. This 439 00:24:35,920 --> 00:24:40,359 Speaker 1: episode was produced by Pia Gutkari Magnus Hendrickson, Liz Smith, 440 00:24:40,600 --> 00:24:44,040 Speaker 1: and Christie Westgard. Francesco Levie is head of Bloomberg Podcasts. 441 00:24:44,320 --> 00:24:45,240 Speaker 1: We'll see you next week.