Latest / Best of Business / Dr Andrew Lensen: AI expert addresses calls that New Zealand needs more AI regulations

Transcript

Dr Andrew Lensen: AI expert addresses calls that New Zealand needs more AI regulations · Best of Business

• 0:09
  Speaker 1: You're listening to a podcast from News Talk sed Be
• 0:12
  Speaker 1: follow this and our wide range of podcasts now on iHeartRadio.
• 0:16
  Speaker 2: So a bit of an eye opening reality this week
• 0:19
  Speaker 2: when it comes to our protectional lack of when it
• 0:22
  Speaker 2: comes to AI ham talking on The Little Things podcast
• 0:25
  Speaker 2: with my co host Louise and II this week, doctor
• 0:27
  Speaker 2: Michelle Dickinson told us New Zealand has no regulation in
• 0:30
  Speaker 2: place to protect us. What does a lack of regulation
• 0:33
  Speaker 2: lead to. Well, here's one example.
• 0:35
  Speaker 3: Meta Facebook just came out recently and said, actually, what
• 0:39
  Speaker 3: we've done is we have been data scraping Australia and
• 0:42
  Speaker 3: New Zealand citizens because you guys didn't make any regulation
• 0:45
  Speaker 3: to protect that. So in the EU you can opt out.
• 0:48
  Speaker 3: In New Zealand, we cannot. We have been using adult
• 0:52
  Speaker 3: Facebook and Instagram pages since two thousand and seven and
• 0:56
  Speaker 3: scraping that data and training all AI models on what
• 0:59
  Speaker 3: you look like. But also if you're holding your child
• 1:02
  Speaker 3: or you're posting a picture of your child and it's
• 1:03
  Speaker 3: not private, we've been using your children's faces and children's
• 1:06
  Speaker 3: images have been created from New Zealand and Australian citizens and
• 1:09
  Speaker 3: we can't. You can never unlearn that.
• 1:12
  Speaker 2: A group of AI experts are also calling from more
• 1:14
  Speaker 2: regulation in the AI space. One of the authors of
• 1:17
  Speaker 2: that letter is Program Director of Artificial Intelligence, doctor Andrew Linsen,
• 1:21
  Speaker 2: and he's with me now. Good morning, Andrew, Good morning.
• 1:24
  Speaker 2: So just how far behind is New Zealand in terms
• 1:27
  Speaker 2: of regulation when it comes to AI.
• 1:30
  Speaker 4: A We're pretty nearly the end, to be honest. I
• 1:33
  Speaker 4: mean we see that Australia is going to do things,
• 1:35
  Speaker 4: Canada and the US, well, the US was doing things,
• 1:38
  Speaker 4: and of course the EU is quite busy as well.
• 1:41
  Speaker 4: So yeah, we're at a back of the pack.
• 1:43
  Speaker 2: Okay, when we look at the countries that are doing something,
• 1:46
  Speaker 2: who's kind of leading the way? Are there examples of
• 1:49
  Speaker 2: countries you know that we should be looking closely at.
• 1:53
  Speaker 4: Yeah, so the example most people turn to will probably
• 1:57
  Speaker 4: be the EU because they have their EUAI Act. It's
• 2:01
  Speaker 4: a very comprehensive piece of legislation and it has its
• 2:04
  Speaker 4: own issues or criticisms, but it's still a really good
• 2:08
  Speaker 4: piece to sort of start looking at as as a
• 2:10
  Speaker 4: starting point. And they have this quite nice risk management
• 2:15
  Speaker 4: based approach, which is sort of what we've advocated for
• 2:17
  Speaker 4: in our letter, And so I think they're a good one.
• 2:19
  Speaker 4: And then of course we can also look at Australia
• 2:21
  Speaker 4: being our neighbors, they also are doing some some quite
• 2:24
  Speaker 4: good stuff as well.
• 2:25
  Speaker 2: What are your concerns if we don't regulate AI.
• 2:29
  Speaker 4: Yeah, I think, like Michelle said, there's there's quite specific
• 2:33
  Speaker 4: concerns like the run strege rays, but there's also you know,
• 2:35
  Speaker 4: many others. I mean, in our leader we see various
• 2:38
  Speaker 4: things like the misuse of intellectual property, right, we see
• 2:42
  Speaker 4: all these all these deep fakes, all these images being created,
• 2:45
  Speaker 4: the use of AI for fraud. We're sort of upheaval
• 2:49
  Speaker 4: the workforce. We know that AI will replace some jobs,
• 2:52
  Speaker 4: but how are we going to manage that transition? As
• 2:56
  Speaker 4: well as even more sort of sinister things like the
• 2:58
  Speaker 4: effect that AI has on democracy itself. I mean a
• 3:01
  Speaker 4: lot of what we're seeing happening in the US. You know,
• 3:03
  Speaker 4: there's a lot of AI lobbying in the AI manipulation
• 3:06
  Speaker 4: behind the scenes there from different parties, so quite wide ranging,
• 3:10
  Speaker 4: but quite scary. But we really can tackle us head
• 3:13
  Speaker 4: on if we choose to.
• 3:15
  Speaker 2: We need to be clear, this isn't about not having AI,
• 3:18
  Speaker 2: is it. I mean AI can be really useful that
• 3:20
  Speaker 2: we want to be innovative here in New Zealand. But
• 3:22
  Speaker 2: it's about deciding in New Zealand, deciding what role were
• 3:25
  Speaker 2: plays in our lives, isn't it.
• 3:27
  Speaker 4: Yeah, that's exactly right. I mean we see really really
• 3:31
  Speaker 4: cool and helpful uses of AI. I mean I have
• 3:34
  Speaker 4: a research project working in the medical sector to use
• 3:37
  Speaker 4: AI as a way to enhance cancer treatment. And you
• 3:40
  Speaker 4: know that's not something we want to we want to stop.
• 3:42
  Speaker 4: That is the sort of thing we should be encouraging.
• 3:44
  Speaker 4: But it's when we think about these sort of third
• 3:47
  Speaker 4: party interests and as Michelle said, places like Facebook and
• 3:51
  Speaker 4: Meta coming into New Zealand and doing whatever they like
• 3:54
  Speaker 4: because we're not saying no. And so I think it's
• 3:57
  Speaker 4: actually really dangerous that the government has basically said we're
• 3:59
  Speaker 4: not going to regulate AI. You know, you're welcome to
• 4:02
  Speaker 4: come here and use however you like, because we will
• 4:05
  Speaker 4: get manipulated and we will see these big players misuse it,
• 4:09
  Speaker 4: rather than getting to set our own course and say hey,
• 4:12
  Speaker 4: this is what we as a country want to use
• 4:15
  Speaker 4: AI for and going from that position.
• 4:17
  Speaker 2: Because if you were a Facebook or Instagram user, you
• 4:19
  Speaker 2: would have to probably trawl through very small terms and
• 4:23
  Speaker 2: conditions in fine print to find where it said that
• 4:25
  Speaker 2: they were going to scrape your public material, right. I mean,
• 4:29
  Speaker 2: I don't think a lot of New Zealanders is probably
• 4:30
  Speaker 2: known that's been going on for such a long time.
• 4:33
  Speaker 4: Well no, And I mean that's just one of many
• 4:35
  Speaker 4: examples you could find, right, And I think it's a
• 4:38
  Speaker 4: kind of indicative of how the hold and we are
• 4:41
  Speaker 4: to these companies, and that people aren't going to say no,
• 4:44
  Speaker 4: I'm not going to use Facebook, or no I'm not
• 4:45
  Speaker 4: going to use Instagram, because that's your whole social network.
• 4:48
  Speaker 4: That's how you talk to your friends, your family, that's
• 4:50
  Speaker 4: how you know, have part of your social group in
• 4:54
  Speaker 4: this modern world. And so the fact they can sort
• 4:56
  Speaker 4: of make these rules and then technically we could say
• 4:59
  Speaker 4: no and not use them. But at the same time,
• 5:01
  Speaker 4: I'm not going to stop talking to my relatives overseas
• 5:04
  Speaker 4: on Facebook, am I very much? It's an unfair transaction.
• 5:10
  Speaker 2: We mentioned sort of what other countries are doing and
• 5:12
  Speaker 2: things we probably need to be creating our own regulation
• 5:16
  Speaker 2: that's specific to New Zealand, though, don't we. I Mean,
• 5:18
  Speaker 2: it's good to look at what other people are doing,
• 5:20
  Speaker 2: but we probably need to be creating what's right for
• 5:22
  Speaker 2: us here, is that correct?
• 5:24
  Speaker 4: Yeah, for sure. And I mentioned the EUAI Act as
• 5:27
  Speaker 4: an example, but obviously that has been designed first of
• 5:30
  Speaker 4: all for European market and also for the culture and
• 5:34
  Speaker 4: sort of the social systems of the EU. And we
• 5:37
  Speaker 4: know in New Zealand we obviously have quite different demographics,
• 5:40
  Speaker 4: we have different value systems, and we have different concerns
• 5:43
  Speaker 4: around the use of AI. So I would like to
• 5:46
  Speaker 4: see us look at these international examples as sort of
• 5:49
  Speaker 4: starting points or as almost case studies, but then make
• 5:53
  Speaker 4: a version for ourselves that is fit for purpose. It
• 5:55
  Speaker 4: doesn't mean starting for scratch. It just means, as you said,
• 5:57
  Speaker 4: tayloring it to our needs.
• 5:58
  Speaker 2: Okay, So we want the benefits, but we want to
• 6:00
  Speaker 2: minimize the risks. So what kind of regulation do we need?
• 6:05
  Speaker 4: Yeah, And so this is why we've been asking for
• 6:09
  Speaker 4: a risk based approach, where we have, for example, some
• 6:12
  Speaker 4: forms of AI use such as state surveillance, facial tracking,
• 6:17
  Speaker 4: you know, those really sort of dystopian things that we
• 6:19
  Speaker 4: say no, no, no, those are not allowed at all.
• 6:21
  Speaker 4: Those are banned. And then we also have a lot
• 6:23
  Speaker 4: of uses of AI that are not concerning. So if
• 6:27
  Speaker 4: people are using co pilot to summarize their emails. You know,
• 6:31
  Speaker 4: we don't really want to put red tape around that.
• 6:33
  Speaker 4: But it's the sort of the stuff in the middle
• 6:35
  Speaker 4: where we look at things like the use of AI
• 6:38
  Speaker 4: and social media, or the use of AI and healthcare,
• 6:40
  Speaker 4: where we can potentially see some benefits or that is
• 6:44
  Speaker 4: needed perhaps, but we want some more oversight as to
• 6:47
  Speaker 4: how that's done, and we want some more guardrails in place.
• 6:50
  Speaker 4: And so this is this idea of having a risk
• 6:52
  Speaker 4: based approach. You're saying, Okay, what's the potential harm here,
• 6:55
  Speaker 4: and based on that, how do we need to what
• 6:58
  Speaker 4: rules do we need to put in place to make
• 6:59
  Speaker 4: sure those those harms are minimized and those benefits are maximized.
• 7:03
  Speaker 2: Are you seeing any interest or effort from politicians when
• 7:06
  Speaker 2: it comes to regulating AI?
• 7:10
  Speaker 4: Yes and no. I mean I had a couple of conversations.
• 7:13
  Speaker 4: We put this letter out, but it's not going anywhere
• 7:17
  Speaker 4: particularly fast, and we haven't really had much response from
• 7:20
  Speaker 4: the government, of course, And so yeah, I'm a little
• 7:24
  Speaker 4: bit I'm not surprised, but I'm still disappointed.
• 7:27
  Speaker 2: Visual how urgently do we need to get something done?
• 7:30
  Speaker 2: I mean, tech development moves very fast, and we just
• 7:34
  Speaker 2: generally always seem to be behind the ABEL, but we
• 7:36
  Speaker 2: are really far behind in New Zealand. He someone just
• 7:38
  Speaker 2: put it in a two hard basket.
• 7:41
  Speaker 4: I think so, and I think so. I think it
• 7:43
  Speaker 4: might become a bit of an election issue next year,
• 7:45
  Speaker 4: but of course even next year is still a long
• 7:48
  Speaker 4: time away. When you think to AI, think about AI,
• 7:51
  Speaker 4: and I mean New Zealand is really concerned about this too, right.
• 7:54
  Speaker 4: I think we're third to last in a global study
• 7:57
  Speaker 4: in terms of trusting AI, and about eight percent of
• 8:00
  Speaker 4: New Zealanders want AI to be regulated. So there's definitely
• 8:03
  Speaker 4: a political mandate there. It's just the government is for
• 8:06
  Speaker 4: whatever reason, not wanting to bother with it. And so
• 8:09
  Speaker 4: I would be very interested to hear why they ignoying
• 8:12
  Speaker 4: it and why do they think it's not an issue,
• 8:14
  Speaker 4: because as your raised, you know, there's very real concerns,
• 8:18
  Speaker 4: not just from what you've said with Michelle, but from
• 8:20
  Speaker 4: when you talk to everyday people as well.
• 8:22
  Speaker 2: Absolutely, thank you so much, doctor Andrew Linsen for talking
• 8:25
  Speaker 2: us through that. The episode of the Little Things is
• 8:28
  Speaker 2: out now you can get it at iHeart. Michelle is
• 8:30
  Speaker 2: really interesting. Andrew mentioned there, of course, the upheaval of
• 8:33
  Speaker 2: the workforce, and Michelle talks a lot about that, and
• 8:37
  Speaker 2: you know about how Singapore put an AI strategy in
• 8:39
  Speaker 2: place ten years ago in order to deal make sure
• 8:41
  Speaker 2: that you know, if you're over forty, you can deal
• 8:43
  Speaker 2: with AI. But so we really are behind the eight ball.
• 8:46
  Speaker 2: But look, if you want to understand a little bit
• 8:48
  Speaker 2: better about what it is and what it does, and
• 8:50
  Speaker 2: also how it might affect you, how it might affect
• 8:52
  Speaker 2: your children, what kind of heart's going to affect our
• 8:54
  Speaker 2: jobs and things. Michelle covers all that in the podcast
• 8:58
  Speaker 2: The Little Things, So I Heart where you get your podcasts,
• 9:01
  Speaker 2: have a listen.
• 9:02
  Speaker 1: Check it out for more from news talks that'd be
• 9:04
  Speaker 1: listen live on air or online, and keep our shows
• 9:08
  Speaker 1: view wherever you go with our podcasts on iHeartRadio