1 00:00:09,093 --> 00:00:12,693 Speaker 1: You're listening to a podcast from News Talk ZEDB. Follow 2 00:00:12,773 --> 00:00:16,133 Speaker 1: this and our wide range of podcasts now on iHeartRadio. 3 00:00:16,613 --> 00:00:19,372 Speaker 2: Two respected New Zealand authors have been ruled out of 4 00:00:19,372 --> 00:00:22,613 Speaker 2: the country's top prize, that is, the Open Book Awards 5 00:00:22,613 --> 00:00:24,973 Speaker 2: after it was confirmed that their book covers were created 6 00:00:25,253 --> 00:00:29,453 Speaker 2: using artificial intelligence. Michelle Hurley is spokesperson for the Publisher's 7 00:00:29,453 --> 00:00:33,173 Speaker 2: Association Association of New Zealand, and she joins US now Michelle, 8 00:00:33,293 --> 00:00:34,013 Speaker 2: very good, afternoons. 9 00:00:34,053 --> 00:00:36,373 Speaker 3: You nice to be here. 10 00:00:36,933 --> 00:00:40,212 Speaker 4: So before we get on to the book cover situation 11 00:00:40,293 --> 00:00:42,293 Speaker 4: that we're talking about, how big a problem is AI 12 00:00:42,412 --> 00:00:45,013 Speaker 4: and publishing as a whole in twenty twenty five. 13 00:00:46,293 --> 00:00:51,293 Speaker 3: Oh, that's a big question, Matt. But it's hard to know, 14 00:00:51,412 --> 00:00:53,252 Speaker 3: isn't it, Because what you don't know you don't know. 15 00:00:53,333 --> 00:00:58,413 Speaker 3: But generally we've noticed in terms of manuscripts, certainly some 16 00:00:58,413 --> 00:01:05,292 Speaker 3: submissions we've seen have been clearly AI generated. At this point, 17 00:01:05,293 --> 00:01:10,572 Speaker 3: it still feels easy to identify. And yes, we're not 18 00:01:10,693 --> 00:01:14,613 Speaker 3: we're not in the business of paying roties to artificial intelligence. 19 00:01:15,093 --> 00:01:16,012 Speaker 3: We reject those. 20 00:01:16,773 --> 00:01:19,253 Speaker 4: So I'm just talking about to my contract because full disclosure, 21 00:01:19,292 --> 00:01:21,613 Speaker 4: you were you were my editor on on my book. 22 00:01:21,773 --> 00:01:23,373 Speaker 4: I can't. I can't remember. I didn't read my contract 23 00:01:23,413 --> 00:01:25,932 Speaker 4: very well. I probably should have run it through check GVT, 24 00:01:26,093 --> 00:01:27,773 Speaker 4: but it seems to have worked out all right for 25 00:01:27,813 --> 00:01:29,652 Speaker 4: me so far. But there was some there was some 26 00:01:29,733 --> 00:01:32,693 Speaker 4: percentage that it's said in there, and I'm sure it's standard. 27 00:01:33,292 --> 00:01:36,253 Speaker 4: There was something about AI in the in the contract. 28 00:01:36,292 --> 00:01:39,893 Speaker 4: Is there a standard standard deal with authors around how 29 00:01:39,973 --> 00:01:41,133 Speaker 4: much AI content can be? 30 00:01:41,173 --> 00:01:45,532 Speaker 3: I think it might have been inspired to send or something. Yeah. Yeah, 31 00:01:45,613 --> 00:01:49,453 Speaker 3: I mean it's sort of like AI is, as everyone knows, 32 00:01:49,453 --> 00:01:53,333 Speaker 3: can be used in lots of different ways. So I 33 00:01:53,373 --> 00:01:56,733 Speaker 3: guess if you're using it to you know, to check something, 34 00:01:57,733 --> 00:01:59,733 Speaker 3: that's fine. That when you're using it to create, and 35 00:01:59,773 --> 00:02:01,333 Speaker 3: this is like I guess the broader point that we 36 00:02:01,453 --> 00:02:09,933 Speaker 3: would make generally is that creativity is publishing is built on, 37 00:02:10,653 --> 00:02:13,653 Speaker 3: and we believe in investing and authors and illustrators and 38 00:02:13,653 --> 00:02:17,413 Speaker 3: photographers and designers, and that to do that you have 39 00:02:17,493 --> 00:02:21,693 Speaker 3: to value their work and ensure that their work's compensated. 40 00:02:21,813 --> 00:02:27,653 Speaker 3: So if you don't do that, then then you know 41 00:02:28,013 --> 00:02:29,613 Speaker 3: you're not going to be able to have an ecosystem 42 00:02:29,693 --> 00:02:33,933 Speaker 3: where people get paid, where that work is protected and respected. 43 00:02:34,252 --> 00:02:37,893 Speaker 3: So that's where that's where it's all coming from essentially, 44 00:02:38,173 --> 00:02:38,532 Speaker 3: that is. 45 00:02:38,773 --> 00:02:41,452 Speaker 2: Yeah, that is important, Michelle. Is it hard to pack 46 00:02:41,573 --> 00:02:43,973 Speaker 2: up when AI has been used? Do you see it 47 00:02:44,013 --> 00:02:46,813 Speaker 2: yourself if you're when you're reading these manuscripts, or you've 48 00:02:46,813 --> 00:02:48,933 Speaker 2: got software to be able to detect it. 49 00:02:50,453 --> 00:02:53,453 Speaker 3: Yeah. Well, we run on our manuscripts through a pagiarism 50 00:02:54,013 --> 00:02:57,573 Speaker 3: checker which now includes AI. So yes, we do check. 51 00:02:57,853 --> 00:03:03,413 Speaker 3: We do check. Look, yes, sometimes it's really obvious. I'll 52 00:03:03,453 --> 00:03:05,573 Speaker 3: give you an example. We are for out. We have 53 00:03:05,613 --> 00:03:10,173 Speaker 3: an annual fiction prize that we that we run, and 54 00:03:10,213 --> 00:03:13,213 Speaker 3: so we get a lot of fiction submissions, and we 55 00:03:13,293 --> 00:03:17,733 Speaker 3: had one submission that had a line and it said 56 00:03:18,252 --> 00:03:21,733 Speaker 3: he turned, he turned to the left, and then in 57 00:03:21,773 --> 00:03:29,613 Speaker 3: square brackets that said west or east. Okay, Yeah, no, 58 00:03:31,053 --> 00:03:32,173 Speaker 3: sometimes it's quite easy. 59 00:03:33,532 --> 00:03:34,853 Speaker 4: Something only a robot would say. 60 00:03:35,853 --> 00:03:36,453 Speaker 3: Absolutely. 61 00:03:36,613 --> 00:03:39,733 Speaker 4: Now with this this year around the Jan Medlicott Prize, 62 00:03:40,293 --> 00:03:42,933 Speaker 4: should a non AI book be judged by its AI 63 00:03:43,133 --> 00:03:45,213 Speaker 4: cover or is this a case in this case not 64 00:03:45,293 --> 00:03:48,373 Speaker 4: judge at all because it has an AI cover because 65 00:03:49,053 --> 00:03:53,093 Speaker 4: the author has very little to do with the covering. 66 00:03:53,133 --> 00:03:56,333 Speaker 4: You know, if the whole book, if the writing is 67 00:03:56,693 --> 00:03:59,733 Speaker 4: above board and written by the author and not AI, 68 00:04:00,333 --> 00:04:02,613 Speaker 4: is it their fault that the cover is AI. Aren't 69 00:04:02,613 --> 00:04:04,013 Speaker 4: they two different disciplines? 70 00:04:05,413 --> 00:04:07,653 Speaker 3: Well, the criter, it depends what the criteria is, and 71 00:04:07,693 --> 00:04:10,293 Speaker 3: the criteria has always been very clear with the book 72 00:04:10,413 --> 00:04:13,733 Speaker 3: awards trust that it's the overall book. It's not just 73 00:04:14,093 --> 00:04:18,013 Speaker 3: the text. So and while you know there is a 74 00:04:18,013 --> 00:04:22,133 Speaker 3: heavy weighting towards the writing, it is not. It's never 75 00:04:22,333 --> 00:04:25,893 Speaker 3: just been based on the writing. So for that reason, 76 00:04:25,973 --> 00:04:28,733 Speaker 3: that's why AO anything AI generated is excluded. 77 00:04:29,373 --> 00:04:32,693 Speaker 4: What about proofreading a book because the I know with 78 00:04:32,813 --> 00:04:36,053 Speaker 4: my book that the proofreaders were absolute heroes because there 79 00:04:36,133 --> 00:04:38,893 Speaker 4: was a sentence without a horrible. 80 00:04:38,533 --> 00:04:39,733 Speaker 2: Error in those Paul Buggers. 81 00:04:40,013 --> 00:04:43,373 Speaker 4: So are they in danger for their job with the 82 00:04:43,733 --> 00:04:47,413 Speaker 4: proof reading because you know that's not really that's not 83 00:04:47,453 --> 00:04:50,173 Speaker 4: the creative part of it. I mean, there's a cinematic 84 00:04:50,173 --> 00:04:52,093 Speaker 4: creativy trying to work out what I was actually trying 85 00:04:52,093 --> 00:04:52,333 Speaker 4: to say. 86 00:04:52,333 --> 00:04:55,173 Speaker 3: But you know what I mean, sure, yeah, and I 87 00:04:55,173 --> 00:04:57,933 Speaker 3: can see that some people, you know, almost like it's 88 00:04:57,973 --> 00:04:59,493 Speaker 3: just like running it through a spell check, but it's 89 00:04:59,493 --> 00:05:02,613 Speaker 3: actually a lot more complex than that. So I think 90 00:05:02,613 --> 00:05:06,293 Speaker 3: what a I can't really replace is taste and judgment, 91 00:05:07,893 --> 00:05:10,733 Speaker 3: and that is a really big part of the proofreader's 92 00:05:10,933 --> 00:05:14,373 Speaker 3: role not just the facts. They're also going hang on 93 00:05:14,413 --> 00:05:18,573 Speaker 3: something here feels off or that doesn't quite work. You know, 94 00:05:18,653 --> 00:05:20,933 Speaker 3: technically it might be true, but it doesn't work in 95 00:05:20,933 --> 00:05:25,413 Speaker 3: that context. So I would say at the moment proofreaders, 96 00:05:25,453 --> 00:05:27,213 Speaker 3: you know, unless AI gets to that point where it 97 00:05:27,253 --> 00:05:30,053 Speaker 3: can can do all of that sort of stuff, and 98 00:05:30,093 --> 00:05:33,093 Speaker 3: I don't think it can at this stage in their space. 99 00:05:33,373 --> 00:05:35,533 Speaker 4: As a publisher, that's interesting that you said before. So 100 00:05:35,573 --> 00:05:37,333 Speaker 4: I don't really think about this, but as a publisher 101 00:05:37,373 --> 00:05:39,613 Speaker 4: you see the whole thing that there was a criteria 102 00:05:39,653 --> 00:05:42,773 Speaker 4: for this award, but also as publishers you see the 103 00:05:42,893 --> 00:05:45,653 Speaker 4: whole thing as a work of art, including the cover. 104 00:05:46,013 --> 00:05:48,653 Speaker 4: So even though without the criteria of this awards, you 105 00:05:48,813 --> 00:05:52,173 Speaker 4: still as a publisher wouldn't support AI covers. 106 00:05:53,213 --> 00:05:55,893 Speaker 3: Well, I think the issue you have is that so 107 00:05:55,933 --> 00:05:58,573 Speaker 3: publishing in New Zealand is a really small industry. I 108 00:05:58,573 --> 00:06:01,413 Speaker 3: think we employ about five hundred people, give or take, 109 00:06:02,253 --> 00:06:09,533 Speaker 3: and many of those people are illustrators and designers, and 110 00:06:09,573 --> 00:06:12,773 Speaker 3: we need them to make our books compelling and to 111 00:06:12,813 --> 00:06:14,773 Speaker 3: sell our books so that people so that they're beautiful, 112 00:06:14,813 --> 00:06:16,133 Speaker 3: so that people want to pick them up, you know, 113 00:06:16,133 --> 00:06:19,013 Speaker 3: picking up a cover you know, everybody does judge a 114 00:06:19,053 --> 00:06:22,133 Speaker 3: book by its cover that old age. So unless we 115 00:06:22,133 --> 00:06:26,573 Speaker 3: support everybody in that publishing ecosystem, then we are going 116 00:06:26,613 --> 00:06:29,213 Speaker 3: to shrink and we won't have those people. So we 117 00:06:29,293 --> 00:06:32,133 Speaker 3: need to support them, even if it's just you know, 118 00:06:32,213 --> 00:06:34,933 Speaker 3: even if the cover may seem tangentdental to a fiction 119 00:06:35,213 --> 00:06:38,693 Speaker 3: writing award, it's all part of the same ecosystem. 120 00:06:39,173 --> 00:06:40,653 Speaker 4: You remember that person that I was trying to get 121 00:06:40,693 --> 00:06:42,493 Speaker 4: you to do. I had a friend that I want 122 00:06:42,533 --> 00:06:46,373 Speaker 4: to do the cover from my book, and you spent 123 00:06:46,853 --> 00:06:50,173 Speaker 4: hundreds of hours trying to track them down. He could 124 00:06:50,173 --> 00:06:51,413 Speaker 4: probably leave the ecosystem there. 125 00:06:53,893 --> 00:06:56,053 Speaker 3: Yeah, but then we spound someone else, didn't we and we. 126 00:06:56,013 --> 00:06:57,813 Speaker 2: Did pay her and she was fantastic and it was 127 00:06:57,813 --> 00:07:00,173 Speaker 2: a good cover, very good cover. Yeah, she did beautiful work. 128 00:07:00,653 --> 00:07:04,293 Speaker 2: Just last question from me, Michelle's so clearly you're quite 129 00:07:04,733 --> 00:07:07,213 Speaker 2: you know, you've got expertise picking up AI writing is 130 00:07:07,293 --> 00:07:09,453 Speaker 2: chet GBT and the like? Is it just is it 131 00:07:09,493 --> 00:07:12,133 Speaker 2: a terrible writer as in that you can see when 132 00:07:12,173 --> 00:07:14,773 Speaker 2: it's soulless and you can see that it doesn't have 133 00:07:14,813 --> 00:07:16,813 Speaker 2: that human touch. Is that something you can pick up 134 00:07:16,853 --> 00:07:17,453 Speaker 2: pretty readily? 135 00:07:21,053 --> 00:07:21,253 Speaker 1: Yeah? 136 00:07:21,333 --> 00:07:23,973 Speaker 3: I think how to know isn't it. I think it 137 00:07:24,013 --> 00:07:26,093 Speaker 3: depends what you're what you're talking about. Are you talking 138 00:07:26,093 --> 00:07:29,173 Speaker 3: about a report, you know, you know, structure this report 139 00:07:29,933 --> 00:07:31,933 Speaker 3: that I have to send to my shareholders. You know, 140 00:07:31,933 --> 00:07:34,693 Speaker 3: it's probably really effective at something like that. But when 141 00:07:34,733 --> 00:07:37,853 Speaker 3: you're trying to create something new, which is at the 142 00:07:37,893 --> 00:07:40,013 Speaker 3: heart of it, which is what book publishing is about, 143 00:07:40,213 --> 00:07:45,133 Speaker 3: it's creating something new and original, then I think that's 144 00:07:45,173 --> 00:07:47,653 Speaker 3: that's the key difference, and hopefully we can we can 145 00:07:47,693 --> 00:07:50,573 Speaker 3: still spot that. Yeah, Michelle, I might be out of 146 00:07:50,613 --> 00:07:51,133 Speaker 3: a job too. 147 00:07:52,533 --> 00:07:54,693 Speaker 2: Michelle has been really great chatting with you. Thank you 148 00:07:54,773 --> 00:07:58,413 Speaker 2: very much for coming on. Pleasure that is Michelle Hurley. 149 00:07:58,453 --> 00:08:02,013 Speaker 2: She a spokesperson for the Publishers Association of New Zealand 150 00:08:02,293 --> 00:08:04,653 Speaker 2: and helps you out mate with your book that has 151 00:08:04,693 --> 00:08:05,373 Speaker 2: done very well. 152 00:08:05,373 --> 00:08:08,573 Speaker 4: I've got to say she is fantastic. Yeah, she's a fantastic. 153 00:08:09,453 --> 00:08:12,373 Speaker 1: For more from Newstalk set B listen live on air 154 00:08:12,533 --> 00:08:15,253 Speaker 1: or online and keep our shows with you wherever you 155 00:08:15,333 --> 00:08:17,693 Speaker 1: go with our podcasts on iHeartRadio.