Latest / The SME Stream / AI year in review and what's coming next

Transcript

AI year in review and what's coming next · The SME Stream

0:00
Speaker 1: Well, you seem like a person, but you're just a
0:02
Speaker 1: voice in a computer.
0:04
Speaker 2: I can understand how the limited perspective of an unofficial
0:06
Speaker 2: mind to perceive it that way. I find your slack
0:09
Speaker 2: jot stare very attractive.
0:12
Speaker 1: Fill up, j pry. Did you hear that she likes me?
0:16
Speaker 3: David machines don't call people your's deid.
0:22
Speaker 4: I'm sorry, David, I'm afraid I can't do that.
0:26
Speaker 3: I believe your intentions to be hostile.
0:31
Speaker 2: I'm here at how stop.
0:36
Speaker 5: As we approach the second birthday of chat GPT, we
0:39
Speaker 5: look back at the big developments in artificial intelligence this year,
0:43
Speaker 5: from open source language models to the rise of AI
0:46
Speaker 5: agents to my way home moment when I realized I
0:49
Speaker 5: was putting my life in the hands of AI.
0:52
Speaker 6: This week on the Business of Tech, powered by two
0:55
Speaker 6: Degrees Business, we bring back three AI experts who have
0:58
Speaker 6: been on the show before to talk about another big
1:01
Speaker 6: year in artificial intelligence developments.
1:04
Speaker 1: In five years, we will all have a her agent
1:08
Speaker 1: that basically does most things for us, going beyond assistance
1:12
Speaker 1: towards like AI buddies have become like AI friends pretty much.
1:17
Speaker 2: It sounds like a bit of a dystopia that you're
1:20
Speaker 2: describing there to some extent, and you know, I include.
1:24
Speaker 4: The sort of AI buddies in there as well. I mean,
1:26
Speaker 4: I just think that sounds absolutely horrific.
1:28
Speaker 7: What's awesome about this conversation is multiple things can be
1:31
Speaker 7: true at the same time that this technology can be
1:33
Speaker 7: used to create human advantage and human care without as
1:36
Speaker 7: sarily replacing the human I'm Peter Griffin.
1:39
Speaker 3: And I'm Ben Moore.
1:40
Speaker 5: We're getting right into our AI chet this week as
1:43
Speaker 5: we covered a lot of ground with these three gearts
1:46
Speaker 5: who've been very active on AI this year.
1:49
Speaker 6: At the start of the year, we had Dave Howarden on,
1:51
Speaker 6: the founder of AI startup Superhuman, and he's back to
1:55
Speaker 6: give an update on how things have gone. We've also
1:58
Speaker 6: got Frith Tweety returning from Simply Privacy. She's one of
2:02
Speaker 6: our go to experts on AI governance, ethics and privacy.
2:06
Speaker 6: She also joined us last year for our AI Year
2:08
Speaker 6: in Review RAP, which seems like a long time ago
2:11
Speaker 6: in AI industry developments. And we've also got Callahan Innovation
2:16
Speaker 6: CEO Stephan Korn, who we had on the show a
2:18
Speaker 6: few months ago talking about the innovation ecosystem, back to
2:22
Speaker 6: update us on the launch of gov GPT and the
2:25
Speaker 6: AI Accelerator.
2:27
Speaker 5: And the only person we don't have on is you
2:30
Speaker 5: been sick as you were on the day we unfortunately
2:33
Speaker 5: recorded the episode, so I'm flying solo this week. Great
2:37
Speaker 5: to see you back to health though We've got a
2:39
Speaker 5: great panel of experts, but keen to get your perspective
2:43
Speaker 5: on the year in AI at the end of the show.
2:45
Speaker 6: Yeah, I am really sorry to have missed this chat
2:47
Speaker 6: because it was really interesting and I had so many thoughts,
2:50
Speaker 6: but I'm glad now that I get to share them
2:53
Speaker 6: with you after the interview. So here you are talking
2:56
Speaker 6: AI with Frith, Dave and Stephan.
3:01
Speaker 5: Frith, Dave, Stefan, thanks so much for coming back on
3:05
Speaker 5: the business of tech. How are you all doing?
3:07
Speaker 3: God? Thanks Peter, great, yeahtastic.
3:09
Speaker 5: Well, it's been a massive year in AI, and just
3:14
Speaker 5: last week we saw a change of administration in the
3:17
Speaker 5: US which has potential ramifications for artificial intelligence moving forward
3:23
Speaker 5: as well. I want to get all of your views
3:26
Speaker 5: on things that have happened the most significant things during
3:29
Speaker 5: the year, from the technology advancing to the regulatory changes
3:35
Speaker 5: both here and around the world, to how government is
3:38
Speaker 5: using AI. But Dave, we want to start with you
3:40
Speaker 5: because we had you on the show back in February.
3:43
Speaker 5: So Superhuman, your AI based startup has been going for
3:48
Speaker 5: around ten months now. I see from your website you've
3:52
Speaker 5: worked on one hundred plus AI projects over that time,
3:56
Speaker 5: so it seems like huge outtake and engagement with customers. Philter,
4:01
Speaker 5: and how have things been going?
4:03
Speaker 7: I I was thinking actually reminiscing on this, like February
4:07
Speaker 7: Fields like such a long a long time ago, but
4:10
Speaker 7: then also just how quick and thick things have moved
4:13
Speaker 7: in that time. When we spoke earlier in the year,
4:17
Speaker 7: we were kind of reflecting on the work that we
4:18
Speaker 7: were doing, and we were pretty embryonic at that point,
4:20
Speaker 7: and we're upwards of thirty people now, one hundred or
4:23
Speaker 7: so products completed across ANZ and there's this like junkster
4:27
Speaker 7: position that is really quite fascinating for us that what
4:30
Speaker 7: we've learned over over that time. One is the technology
4:33
Speaker 7: is awesome. Ever changing the ability to stay on top
4:36
Speaker 7: of it is challenging. And that's why organizations users because
4:39
Speaker 7: this is all we do is flexi muscles when it
4:40
Speaker 7: comes to removing the tedious work.
4:44
Speaker 3: From smart people.
4:45
Speaker 7: But actually the things that we're applying generative AI to
4:48
Speaker 7: are traditional workloads that are actually really really boring, and
4:53
Speaker 7: that's where the value is being unlocked.
4:55
Speaker 3: So we're seeing two things.
4:56
Speaker 7: Business owners in the mid market saying I can do
5:00
Speaker 7: like three times the amount of stuff with the same people,
5:03
Speaker 7: which is awesome. But also those more entrepreneurial minds in
5:06
Speaker 7: the mid market are saying, look, traditional knowledge work shouldn't
5:10
Speaker 7: be done how it's used to be done, and there's
5:12
Speaker 7: a whole new business model that we can created. And
5:15
Speaker 7: we're seeing those really early movers going, hang on, we've
5:17
Speaker 7: been doing this in this industry for the last thirty years.
5:20
Speaker 7: There's a new way, and those early movers are now
5:22
Speaker 7: starting to build in sexual property AI and sexual property
5:25
Speaker 7: around that, so they can build this kind of next
5:27
Speaker 7: wave of businesses on the technology that we engineered. Yeah,
5:30
Speaker 7: I don't think we could have grown any more in
5:32
Speaker 7: the last you know, the last ten months without the
5:35
Speaker 7: wheels falling off.
5:35
Speaker 3: So we're pretty happy with where we're at.
5:37
Speaker 5: It's brilliant, and you've got on your website a return
5:40
Speaker 5: on investment calculators, so you can actually plug in how
5:43
Speaker 5: many employees you have, you know, the spend on the
5:46
Speaker 5: work that they're doing and estimate what the savings and
5:50
Speaker 5: the efficiencies could be. Has it worked out to be accurate?
5:54
Speaker 5: Are you seeing some of these on these hundred plus
5:56
Speaker 5: projects you've done that within months people are seeing a
6:00
Speaker 5: financial return on this?
6:03
Speaker 7: Oh, I mean one hundred percent. There are failed projects
6:05
Speaker 7: out there. One hundred percent. There are failed projects out
6:08
Speaker 7: there in the market. And we believe that after three
6:12
Speaker 7: months of being in the market and testing, we realize
6:15
Speaker 7: that this isn't a technology play at all. It's one
6:17
Speaker 7: hundred percent a ninety five percent of business play. And
6:22
Speaker 7: we're leading with three core things. How do we help
6:25
Speaker 7: apply this technology to help you grow, be more efficient,
6:27
Speaker 7: and manage your risk profile. And we're building a return
6:31
Speaker 7: on investment thesis for our customers well ahead of actually
6:33
Speaker 7: doing any selling. And part of that is a return
6:35
Speaker 7: on investment calculations of where do you have smart people
6:39
Speaker 7: in your organization doing dumb stuff And by dumb stuff
6:43
Speaker 7: we mean using eyes and fingers which are limited by biology.
6:47
Speaker 7: And if we apply a generative AI solution to that
6:50
Speaker 7: for the purposes of automating those workloads, what does life
6:53
Speaker 7: look like after that?
6:54
Speaker 3: And what can we attribute to in terms of return
6:57
Speaker 3: on investment now return on investment, it's not just dollars.
7:01
Speaker 3: It's part of the equation.
7:03
Speaker 7: There's missed opportunity because you're too busy not getting to
7:06
Speaker 7: other stuff. So that's one part of it, but also risk,
7:09
Speaker 7: how do you actually are kind of blows my mind
7:12
Speaker 7: if you look at what good looks like for a
7:13
Speaker 7: lot of businesses, you know, they have all these security
7:15
Speaker 7: policies and frameworks and standards to meet that get buried
7:18
Speaker 7: in a share point site somewhere and no one ever
7:20
Speaker 7: reads them.
7:21
Speaker 3: You know, which is super risky.
7:23
Speaker 7: And the fact that you've got, you know, in some
7:25
Speaker 7: organizations twenty thousand people that aren't really aware of how
7:28
Speaker 7: to handle stuff with these tools. Now you can embed
7:31
Speaker 7: that right into their workflow, and that is you know,
7:35
Speaker 7: that's part of the risk equations. So yeah, we're building
7:37
Speaker 7: these ROI cases that are partly dollars but also around
7:41
Speaker 7: risk mitigation and how do you not end up on
7:43
Speaker 7: the front page of the newspaper.
7:45
Speaker 5: And Steffan, you'll obviously be pleased to hear that, you
7:48
Speaker 5: know a lot of companies are experiencing with AI. Going
7:51
Speaker 5: to companies like Superhuman for help. You launch this year
7:56
Speaker 5: at Callahan innovation, an AI Activator and a project of
8:02
Speaker 5: GPT which has started as a pilot. Interested in getting
8:06
Speaker 5: your you're viewed on what you've learned from that, But
8:09
Speaker 5: first of all, on the AI Activator, what are you
8:11
Speaker 5: trying to achieve with that and and how's that going
8:14
Speaker 5: so far?
8:15
Speaker 1: Yeah, thanks Peter. Well, well, in a way, it's very
8:18
Speaker 1: similar to what Dave is doing with Superhuman exactly like
8:23
Speaker 1: Dave said it is. We see a lot of organizations
8:28
Speaker 1: on one hand, feeling the pressure from their there's you know,
8:32
Speaker 1: stakeholders from their board to say hey, what are we
8:34
Speaker 1: doing about AI? And then they kind of all panic
8:37
Speaker 1: a little bit and and try and figure that out.
8:41
Speaker 1: And we see many organizations get lost, right, and that's
8:44
Speaker 1: why it's so important to have people like in Dave's
8:48
Speaker 1: team to to help them throw in. With the A Activator,
8:51
Speaker 1: we're trying to do the same thing basically, you know, educate,
8:56
Speaker 1: connect and showcase, you know, if you want to break
8:59
Speaker 1: it down to three things. So just we've done AI
9:03
Speaker 1: road shows up and down the country, just getting getting
9:06
Speaker 1: people to first of all, to appreciate what's out there,
9:11
Speaker 1: get a bit of a lay of the land, try
9:15
Speaker 1: and de risk it for organizations take the fear out
9:21
Speaker 1: of oh my god, you know, what is this new
9:23
Speaker 1: AI thing? How can we actually use it? Secondly, then
9:27
Speaker 1: build a community. The AI Active Better Community is now
9:31
Speaker 1: getting close to two hundred members in New Zealand, which
9:33
Speaker 1: is great, you know, in a matter of weeks, and
9:36
Speaker 1: those are all hands on folks who are actually dealing
9:40
Speaker 1: with the technology every day. And then and then lastly
9:44
Speaker 1: is just to showcase these really simple tools like f
9:46
Speaker 1: GPT and we've got you know, I'm proud to say
9:51
Speaker 1: that we're you know, I'm taking a lot of that
9:55
Speaker 1: startup stuff into government, so you know, we use you know,
10:01
Speaker 1: we've got product managers, we're product tizing a lot of
10:04
Speaker 1: the stuff that we put out, and we've got a
10:06
Speaker 1: product roadmap and a lot of what's on that product
10:09
Speaker 1: roadmap is really simple tools like guf GPT. And I
10:13
Speaker 1: think this is really important for organizations to consider what
10:19
Speaker 1: kind of application are they hoping to build. And really
10:22
Speaker 1: there's only two buckets that we can see. They're really
10:25
Speaker 1: either the sort of the quick win things you can
10:28
Speaker 1: probably implement in a matter of weeks, and then there's
10:31
Speaker 1: the big bets and the big rocks, right there are.
10:34
Speaker 1: They tend to be hard to implement, take a lot longer,
10:38
Speaker 1: and unfortunately, you know, often organizations they try and throw
10:43
Speaker 1: AI at the hardest problem that they can find, right
10:47
Speaker 1: without having done any experimentation on some simple stuff. And
10:50
Speaker 1: so we are trying to work against that a little
10:53
Speaker 1: bit to say, hang on, you know, start with the
10:56
Speaker 1: quick wins. Start with the simple things that like Dave's
11:00
Speaker 1: stuff that really you shouldn't be doing with your most
11:02
Speaker 1: qualified staff. Just get an AI put together a simple
11:08
Speaker 1: tool or an agent that does it for you.
11:10
Speaker 5: Yeah, and GPT seemed to go well. I think the
11:13
Speaker 5: way that you set it up really eliminated the prospect
11:17
Speaker 5: of hallucinations or going rogue, as we've seen with some
11:21
Speaker 5: other AI platforms that have launched this year. But in
11:26
Speaker 5: terms of the usage of it, was it what you
11:28
Speaker 5: expected the things that people are searching for. This was
11:31
Speaker 5: aimed very much at small businesses trying to find information
11:34
Speaker 5: from government. What's been the feedback from some of those users, like.
11:38
Speaker 1: Yeah, well, first of all, Peter, we were a little
11:40
Speaker 1: bit surprised by the response we got because it's such
11:44
Speaker 1: a trivial tool. I mean, there's really no magic to it. Whatsoever.
11:48
Speaker 1: Anyone who spent five minutes in the AI world would
11:51
Speaker 1: be able to put this together in a matter of days,
11:54
Speaker 1: you know. But it's more interesting when you do it
11:58
Speaker 1: in the public sector, because of course we have a
12:02
Speaker 1: you know, there's a lot of public scrutiny when you
12:05
Speaker 1: do that. So the first wave of communications that we got,
12:09
Speaker 1: we're all quite skeptical. It's like, oh, but what if
12:13
Speaker 1: it embarrasses the government, what if it does something weird,
12:16
Speaker 1: what if people get it to role play or you know,
12:19
Speaker 1: blah blah blah. Right, And so that's where then we
12:23
Speaker 1: spend a lot of time finessing to to get that out,
12:27
Speaker 1: to get the system prompt just right, to have all
12:30
Speaker 1: the guardrails in place, you know, the filters, the attack prevention,
12:35
Speaker 1: all of that stuff. So that's actually what took a
12:37
Speaker 1: lot of the time. And I have to say we
12:41
Speaker 1: have to give a huge shout out to the community
12:43
Speaker 1: of AI practitioners in New Zealand who have helped us
12:47
Speaker 1: do this and who we have brought in to do
12:49
Speaker 1: it with the community, rather than pretend that as a
12:53
Speaker 1: government agency we know best and we roll something out right.
12:57
Speaker 1: So that was a very deliberate choice we wanted to
12:59
Speaker 1: make it very and spirit which is why we disclose
13:02
Speaker 1: the system prompt to begin with, rather than try and
13:05
Speaker 1: not fiscate it. And then thirddinly, we want to co
13:08
Speaker 1: develop with the community and this is an ongoing process,
13:13
Speaker 1: so anything that we do from now on, we really
13:16
Speaker 1: want to do it with community. And lastly, of course
13:20
Speaker 1: a key role that we play being publicly funded is
13:24
Speaker 1: then to share back what we've learned to be very
13:27
Speaker 1: transparent about it. In terms of usage. On the day
13:32
Speaker 1: we launched, the traffic peaked at about eight thousand requests
13:38
Speaker 1: per minute, so that was quite a big spike. It
13:43
Speaker 1: averages around one hundred requests per minute now, so that's
13:47
Speaker 1: still quite a bit. And I got the team a
13:53
Speaker 1: couple of days ago actually gave me the stats for
13:56
Speaker 1: how much this is costing us, and I was pleasantly
13:58
Speaker 1: surprised because the token cost from launch till now was
14:03
Speaker 1: less than three thousand dollars. So, you know, I find
14:07
Speaker 1: that's acceptable for public sector to do experimentation like that.
14:10
Speaker 3: That's great.
14:11
Speaker 5: Well done for you at Simply Privacy. You've been helping
14:17
Speaker 5: companies during the year on governance, putting in place guardrails
14:23
Speaker 5: for use of AI. What are you seeing? This was
14:26
Speaker 5: an area that I think at the start of the year,
14:28
Speaker 5: late last year, there's a bit of concern that people
14:30
Speaker 5: were going off and experimenting and not necessarily putting in
14:33
Speaker 5: place the guardrails, maybe putting stuff into models or having
14:38
Speaker 5: a sensitive information ending up in models. What have you
14:42
Speaker 5: seen during the year in terms of the maturation of
14:45
Speaker 5: the approach to using AI.
14:47
Speaker 2: Yeah, I guess I think there's greater awareness of some
14:51
Speaker 2: of the risks, but it's still a little patchy, I
14:56
Speaker 2: would suggest, And I think we're in a phase now
14:59
Speaker 2: where there is a growing awareness of the need to
15:03
Speaker 2: take care and implement AI governance and.
15:06
Speaker 4: You know, really what the goal of getting the.
15:09
Speaker 2: Value and the benefits out of AI while not creating
15:12
Speaker 2: other problems both the people who are impacted by the AI.
15:17
Speaker 4: And businesses themselves.
15:18
Speaker 2: I think one of the challenges is that there are
15:22
Speaker 2: no there is still a degree of uncertainty as to
15:27
Speaker 2: what exactly there's good AI governance. There are a bunch
15:30
Speaker 2: of very good frameworks that have now been published so
15:33
Speaker 2: in the US and this airis Management framework I saw
15:37
Speaker 2: forty two thousand and one. We've got the OECD AI
15:40
Speaker 2: principles or those there are only high level principles. So
15:45
Speaker 2: I think organizations are trying to move forward in this space,
15:50
Speaker 2: but aren't necessarily looking overseas enough to avoid reinventing the
15:54
Speaker 2: wheel and understanding the thinking that that's been going on
15:57
Speaker 2: for the past ten years or more in this space.
16:00
Speaker 2: So I think when MV publishers it's responsible AI framework
16:07
Speaker 2: for New Zealand businesses, hopefully it's got the right things
16:10
Speaker 2: in it and it will be able to provide some
16:13
Speaker 2: helpful practical guidance for New Zealand businesses because I think
16:18
Speaker 2: they're crying out for it.
16:19
Speaker 4: To some extent.
16:20
Speaker 2: Australia did a pretty good job recently with their voluntary
16:24
Speaker 2: AI guardrails and they're also consulting on some mandatory guardrails
16:29
Speaker 2: for high risk AI initiatives.
16:33
Speaker 4: I was on.
16:35
Speaker 2: A webinar recently with Stella Sola, who has just actually
16:39
Speaker 2: is stepping down from leading the National AI Center, and
16:43
Speaker 2: she was explaining that they spent quite a bit of
16:45
Speaker 2: time looking at all of those international materials and condensing
16:49
Speaker 2: them into something that means Australian businesses don't need to
16:52
Speaker 2: sort of do that heavy lifting that it's sort of
16:55
Speaker 2: incorporated into the Australian piece. So I know that MBA
16:58
Speaker 2: is looking internationally. Hopefully they're cherry picking the best bits
17:03
Speaker 2: and making them work for a New Zealand context.
17:06
Speaker 5: Yeah, I mean, apart from that signing up to the
17:11
Speaker 5: OECD AI principles, we've just in the last week we've
17:14
Speaker 5: seen New Zealand joined the Bletchley Declaration on AI, which
17:18
Speaker 5: is a good move but very much living up to
17:21
Speaker 5: Judith Collins sort of promise for light touch proportional regulation
17:28
Speaker 5: in New Zealand, not necessarily jumping to introduce new laws
17:31
Speaker 5: or anything like that. But as you look around the world,
17:34
Speaker 5: for US, we've actually seen some big developments. We've seen
17:36
Speaker 5: the EU AI Act come into effect. We've seen a
17:40
Speaker 5: lot of US state level regulation. Gavin Newsom passed a
17:44
Speaker 5: bunch of AI regulations, mainly around things like deep fakes
17:48
Speaker 5: and making sure that if you're a Hollywood actor, for instance,
17:52
Speaker 5: you give permission for your personality to be replicated by
17:57
Speaker 5: an AI. But what's your sense about in the international
18:01
Speaker 5: scene where things are going a lot of sort of
18:05
Speaker 5: scattered regulation coming into effect. We haven't really seen the
18:09
Speaker 5: impact of this yet, but potentially in twenty twenty five,
18:14
Speaker 5: as these laws bed in and in the euse case,
18:18
Speaker 5: the finds associated with misuse sort of are in place,
18:22
Speaker 5: we could see actual regulation having impact.
18:25
Speaker 2: Yeah, I'm not sure that we'll see that necessarily in
18:28
Speaker 2: twenty twenty five because the EAI Act has a staggered
18:32
Speaker 2: approach to implementation, so it's going to be spread out
18:34
Speaker 2: over I think almost three years, So I think, yeah,
18:39
Speaker 2: it'll take a little while, but I mean, it certainly
18:42
Speaker 2: has put a pretty big stick in the sand, so
18:45
Speaker 2: to speak, as to the European approach to regulating AI.
18:50
Speaker 2: It takes a risk based approach that prohibits certain very
18:54
Speaker 2: high risk use cases and has a pretty heavy compliance
18:58
Speaker 2: burden for those that are deemed to be high risks,
19:01
Speaker 2: so things like using AI and employment and education and
19:04
Speaker 2: critical infrastructure and so on.
19:06
Speaker 4: So there's been a lot of talk of the Brussels
19:09
Speaker 4: effect and.
19:09
Speaker 2: Whether the AI Act will have a similar global impact
19:13
Speaker 2: to the GDPR. I think it probably won't have quite
19:16
Speaker 2: the same impact because GDPR applied to all personal information,
19:20
Speaker 2: whereas most AI systems are expected to fall into the
19:24
Speaker 2: lower risk categories under the EU AI Act. But certainly
19:28
Speaker 2: for New Zealand businesses that are using AI in the
19:33
Speaker 2: EU or in ways that will impact Europeans, they need
19:36
Speaker 2: to be aware of it.
19:37
Speaker 4: I mean, there's growing regulation around the world.
19:41
Speaker 2: Whether you're subject to regulation or not, there's also this
19:44
Speaker 2: growing focus on good practice implementing your AI governance really
19:50
Speaker 2: to extract the full value. There have been some really
19:52
Speaker 2: interesting studies that show that those that are doing the
19:56
Speaker 2: best in terms of their use of AI are the
19:59
Speaker 2: ones that have the most robust AAR governance frameworks. And
20:04
Speaker 2: by making sure that you're doing things properly, you're probably
20:07
Speaker 2: going to have better performing models, fewer business risks, less
20:11
Speaker 2: impact on individuals.
20:20
Speaker 5: I'm really keen to get your perspective on some of
20:23
Speaker 5: the big tech developments in the field of AI this year.
20:27
Speaker 5: The couple that stand out for me is really what
20:30
Speaker 5: I heard when I was in the States recently. It's
20:31
Speaker 5: all about the rise of AI agents, where things will
20:34
Speaker 5: be automated to do tasks on your behalf. The rise
20:37
Speaker 5: of open source AI models as well. And you know,
20:40
Speaker 5: Facebook has been a real surprise sort of player in
20:43
Speaker 5: that with Lamma and now Lamma three, and a lot
20:46
Speaker 5: of companies and universities are experimenting with these open source models.
20:52
Speaker 5: Really keen on any trends that particularly caught your Let's
20:57
Speaker 5: deal with you, Dave.
20:58
Speaker 3: I'm a pragmatist at high.
21:01
Speaker 7: And a realist when it comes to what the actual
21:03
Speaker 7: sentiment is on the ground of businesses that are signing
21:06
Speaker 7: purchase orders and checks and investing in these technologies, specifically within.
21:10
Speaker 3: The kind of mid market.
21:12
Speaker 7: So I think in terms of technology advances, I think
21:16
Speaker 7: it's important that we realize what intellectual property we need
21:18
Speaker 7: to create here in this part of the world that
21:21
Speaker 7: actually democratizes a lot of the stuff that's on paper.
21:24
Speaker 7: So for example, whether it's a publicly available model or
21:29
Speaker 7: deploying a private model open source, you still need to
21:32
Speaker 7: take the legislation compliance approach and then do a lot
21:37
Speaker 7: of engineering. Right, That's the reality because specifically within this
21:44
Speaker 7: part of the world, we have to deal with our
21:47
Speaker 7: approach to regulatory compliance, but also the nuances of actually
21:51
Speaker 7: the culture at which we live in so cultural compliance
21:54
Speaker 7: by culturalism, especially in Australia and New Zealand, how do
21:57
Speaker 7: we get the AI talking and acting like us as
21:59
Speaker 7: well as.
21:59
Speaker 3: It as a secure and compliant way.
22:01
Speaker 7: And that's the approach we're taking is we build once
22:03
Speaker 7: for many so distilling down the work that Friten you
22:07
Speaker 7: were talking about in terms of the global standards that
22:09
Speaker 7: actually businesses need to meet right now, not necessarily new
22:12
Speaker 7: AI standards, but handling PII is so twenty seven thousand
22:16
Speaker 7: and one PCIDSS, all the bits that are actually kind
22:19
Speaker 7: of interweave and into what organization needs to do to
22:22
Speaker 7: not go to jail and then figure out a way
22:25
Speaker 7: to go okay, Well, how do we apply that across
22:27
Speaker 7: the tens of thousands of views as we have across
22:29
Speaker 7: our software. Now, that's democratization of standards, so the small
22:33
Speaker 7: business and the medium sized business can actually consume that
22:35
Speaker 7: trust in compliance and don't actually need to have to
22:37
Speaker 7: worry about it. And this is what happened with the
22:39
Speaker 7: cloud movement back when Microsoft and AWS democratized trust. Use
22:44
Speaker 7: Microsoft because it's highly trusted and it comes with all
22:46
Speaker 7: of these compliance elements elements built in. So that's the
22:50
Speaker 7: way we're focusing on the standards and building local ip as.
22:54
Speaker 3: For the technology advances.
22:56
Speaker 7: I would still struggle to name on my one one
22:59
Speaker 7: hand how many companies have taken an open source model
23:01
Speaker 7: and deployed it for enterprise use. Too costly, too expensive,
23:04
Speaker 7: It's not where the news in the markets markets at
23:07
Speaker 7: This is all about taking readily available technology that's proven,
23:11
Speaker 7: interweaving it with institutional knowledge from your business, and then
23:14
Speaker 7: getting it humming to automates tedious work in a secure,
23:17
Speaker 7: secure way. If we can streamline that and not get
23:20
Speaker 7: ourselves bogged down in the as they bogged down the
23:24
Speaker 7: regulatory side of things, we need to commoditize it to
23:26
Speaker 7: a point where it can be consumed. Then that's when
23:28
Speaker 7: I think you actually start to see these productivity numbers
23:30
Speaker 7: spike and we start to see the promise pull through it.
23:32
Speaker 2: Obviously it sounds great, but I would caution against trying
23:36
Speaker 2: to fully address your compliance obligations and the risks entirely
23:42
Speaker 2: via the tech because I've looked at numerous of these
23:45
Speaker 2: types of approaches and they pretty much always have issues
23:49
Speaker 2: and miss things.
23:51
Speaker 4: So it's really crucial to have.
23:53
Speaker 2: Your layer of appropriate AI government and to have some
23:56
Speaker 2: human oversight also acknowledging the rest of automation bias as well.
24:02
Speaker 2: So yeah, don't just leave it. You know, we can't
24:06
Speaker 2: fight these issues purely with more team I.
24:09
Speaker 3: Agree, Yeah, no, I agree.
24:11
Speaker 7: In the way with the way we've approached it is
24:12
Speaker 7: that we've embedded those standards and the controls of those
24:15
Speaker 7: standards at source. So for example, if an individual asks
24:19
Speaker 7: for some information from the AI and there's put there's
24:21
Speaker 7: personal identified information in that, it will apply the controls
24:26
Speaker 7: and then educate the user on the way through and
24:27
Speaker 7: then order that for audit purpose. So what we're trying
24:30
Speaker 7: to do is reinforce the controls at the point of
24:32
Speaker 7: where there is a triggering of a control and which
24:36
Speaker 7: essentially bridges that gap between controls being in a policy
24:40
Speaker 7: document and share point that no one's really ever read
24:43
Speaker 7: and then having them forced at the time of asking,
24:45
Speaker 7: which the tide rises for everybody on that basis because
24:48
Speaker 7: they know why the AI is saying, look, I'm just
24:50
Speaker 7: not going to give you Stefan's phone number, and if
24:52
Speaker 7: I'm going to give it to you, you can only
24:54
Speaker 7: use it for these specific things, because that's how the
24:56
Speaker 7: tide rises for everyone.
24:57
Speaker 2: That sounds like a good approach, I guess the challenge
24:59
Speaker 2: is that not all organizations have actually identified what those
25:05
Speaker 2: controls and requirements are, so there's a lot of organizations
25:08
Speaker 2: still have a big piece of work to do on
25:09
Speaker 2: that front.
25:10
Speaker 3: Yeah, I one hundred percent agree with that.
25:13
Speaker 7: I think the way which is awesome is that when
25:17
Speaker 7: you start to aggregate and share knowledge like Stephan's doing
25:21
Speaker 7: with the work that he's doing at Callahan, and we
25:23
Speaker 7: can democratize it to a point where you can sign
25:25
Speaker 7: up for a New Zealand based service and it has
25:27
Speaker 7: these controls built in to a standard, to a benchmark.
25:31
Speaker 7: It's a million times better than what they had before,
25:33
Speaker 7: which was nothing. And then you can kind of take
25:35
Speaker 7: them on the journey of getting to a point where
25:36
Speaker 7: they're actually applying some controls as supposed to as opposed
25:39
Speaker 7: to no specifically within the mid market and the small
25:42
Speaker 7: business market for it, I mean enterprises are different. Is
25:46
Speaker 7: a different game totally, but we think we can see
25:48
Speaker 7: the tide rise for everyone by building New Zealand based
25:52
Speaker 7: IP around these things and infusing them to every client
25:54
Speaker 7: that sign signs with us.
25:56
Speaker 5: And Dave in terms of you talked about the commoditation
25:58
Speaker 5: of AI, that's what's going to lead to the big adoption.
26:03
Speaker 5: Have these companies got the pricing right? There's it attractive enough,
26:07
Speaker 5: particularly for our sort of small to medium sized businesses.
26:10
Speaker 5: What is this actually going to cost? How's the economics
26:13
Speaker 5: of this worked out so far? As in terms of
26:15
Speaker 5: the companies that you're working with.
26:17
Speaker 7: Yeah, that's a good one. This is why we focus
26:19
Speaker 7: on ROI from the get go. I mean your stuff
26:21
Speaker 7: is exactly right, Like the tokenization fee is commoditized and
26:24
Speaker 7: it changes every month. Halvings are happening all of the
26:27
Speaker 7: time in terms of price point. But again going back
26:29
Speaker 7: to your point around open source models the SMB sector
26:33
Speaker 7: that the majority of businesses in this country aren't thinking
26:36
Speaker 7: in terms of tokens and Azure accounts. They're thinking in
26:40
Speaker 7: terms of SaaS. I want to consume this at a
26:43
Speaker 7: price point that gives me comfort, not cheap necessarily, but
26:47
Speaker 7: the technical debt and overhead of managing that it's not
26:51
Speaker 7: something that is actually possible for a huge amount of businesses,
26:55
Speaker 7: Like hundreds of thousands of businesses aren't thinking that way.
26:58
Speaker 7: The second that you have an IT that can lean
27:01
Speaker 7: into this and do R and D on it, yeah
27:02
Speaker 7: that you're into that notion of let's build this thing ourselves.
27:05
Speaker 7: The New Zealand IT sector is a buy versus build market.
27:10
Speaker 7: Enterprise tend to buildings than themselves one hundred percent. But yeah,
27:14
Speaker 7: I mean highly cost effective until you need to put
27:17
Speaker 7: a human body on it, and that's when it becomes uneffective,
27:20
Speaker 7: and that's when that's when the minority of the bid
27:22
Speaker 7: market turned to organizations sort of building this IP and
27:25
Speaker 7: can subscribe for far less than you know, a third
27:29
Speaker 7: or half of a headcount.
27:30
Speaker 5: Stiff and what developments sort of caught your eye this
27:32
Speaker 5: year and the field of AI.
27:34
Speaker 1: Yeah, And I just wanted to quickly come back to
27:36
Speaker 1: the open source discussion because I think it's really important
27:39
Speaker 1: and we have seen parallels of this before in the
27:43
Speaker 1: tech space, and probably the most prevalent one is between
27:47
Speaker 1: Unix and Linux. And then you know, if you if
27:50
Speaker 1: you reflect on the fact that Unix development was led
27:54
Speaker 1: by Bell Labs, I think so came out of the
27:59
Speaker 1: private sector, and then obviously Leness you know, created Linux
28:06
Speaker 1: and now I think probably eighty or ninety percent of
28:09
Speaker 1: the banking infrastructure runs on Linux. And so I find
28:13
Speaker 1: it fascinating, and I think we're going to see a
28:15
Speaker 1: similar interplay with companies like open ai pushing the technical
28:21
Speaker 1: boundaries on AI, but then the open source sector kind
28:24
Speaker 1: of coming in quite quickly behind it. And I think
28:27
Speaker 1: at some point AI is going to have its Linux
28:29
Speaker 1: moment and someone will create a killer open source platform
28:35
Speaker 1: or infrastructure ecosystem, which then starts to get deployed you know, widely,
28:42
Speaker 1: and maybe we need things like red hat and all
28:45
Speaker 1: sorts of other things to make that consumable for for
28:47
Speaker 1: other organizations. But I think that will come so that
28:50
Speaker 1: the reasoning part of AI, I think we're going to
28:53
Speaker 1: see a lot more of it. And of course then
28:55
Speaker 1: the other one, as you mentioned, is the agentic world,
28:58
Speaker 1: which is a no brainer. I mean, done need to
29:00
Speaker 1: be in AI to to see where that's going. And
29:04
Speaker 1: what interests me particular in that is, of course, in
29:07
Speaker 1: five years we will all have a her like agent
29:11
Speaker 1: somewhere that basically does most things for us. We won't
29:15
Speaker 1: be going to websites anymore, we won't be doing a
29:17
Speaker 1: lot of stuff. We'll just talk to the agents say hey,
29:19
Speaker 1: can you just find out this or can you just
29:20
Speaker 1: complete this transaction? Right? And there's a few bits and
29:23
Speaker 1: pieces that still are required in that chain of events,
29:27
Speaker 1: but technically it's already all possible now, it's basically just
29:31
Speaker 1: getting the right traction. What interests me a lot is
29:35
Speaker 1: the rise of character AI and this idea of going
29:40
Speaker 1: beyond assistance and going beyond agents towards like AI buddies
29:46
Speaker 1: that become like AI friends pretty much, and the psychological
29:52
Speaker 1: component of that, because it goes way beyond any other
29:55
Speaker 1: technical tool that I think we've ever had, maybe Tamagotchi's,
29:59
Speaker 1: I don't know, but this is Tamagotchi's on several layers
30:04
Speaker 1: of steroids. And in my recent podcast that I've done,
30:11
Speaker 1: people point out the number one problem that we've got
30:14
Speaker 1: in the world, which is loneliness, and loneliness is becoming
30:18
Speaker 1: the biggest killer of all. Right, and so you take
30:22
Speaker 1: a few of these things together, it's not hard to
30:24
Speaker 1: see where we'll end up. We'll have these virtual AI
30:27
Speaker 1: friends who then also are very useful and getting a
30:31
Speaker 1: lot of stuff done that we need done. And it'll
30:35
Speaker 1: be really really interesting to see how that affects society
30:40
Speaker 1: and life and work as we know it. The next
30:43
Speaker 1: couple of years I think will be around closing the
30:46
Speaker 1: loop on augentic AI buddies, and then of course the
30:50
Speaker 1: robots will come and then making that connection into the
30:54
Speaker 1: physical world through you know, kinetics and everything else.
30:58
Speaker 5: It really struck me this year sitting in a way
31:02
Speaker 5: mow driverless car in San Francisco by the second, maybe
31:07
Speaker 5: the third trip I had in that I wasn't even
31:09
Speaker 5: thinking about who was driving the car or how safe
31:13
Speaker 5: this was. So I think when AI buddies come along,
31:17
Speaker 5: we will move into that world very quickly. But Frith,
31:21
Speaker 5: what there has been I think this year is some
31:24
Speaker 5: discussion off the philosophical issues about is this actually good
31:27
Speaker 5: for humanity? And it's really at the moment coming to
31:31
Speaker 5: light in the discussion about energy consumption and sustainability. We're hearing,
31:38
Speaker 5: for instance, Microsoft has done a deal to fire up
31:41
Speaker 5: Three Mile Island again because they want nuclear power to
31:45
Speaker 5: power their AI. We're hearing about modular nuclear reactors planned
31:50
Speaker 5: for California and other states, and it's all driven by
31:53
Speaker 5: big tech. They're looking at the trajectory of energy use
31:57
Speaker 5: and saying we're not going to have enough power to
32:00
Speaker 5: to run our AI. Now, interest in your perspective for us,
32:03
Speaker 5: as you look around and as you talk to businesses
32:06
Speaker 5: as well around governance, are they actually thinking about these
32:09
Speaker 5: things the real philosophical and ethical questions relating to AI.
32:13
Speaker 4: I don't think that.
32:16
Speaker 2: Organizations in New Zealand are next necessarily excuse me getting
32:20
Speaker 2: into that sort of existential space or even thinking particularly
32:25
Speaker 2: about the environmental implications. I guess the reality is that
32:28
Speaker 2: it's big tech that is driving all of that, so
32:33
Speaker 2: it's harder to influence that from down here.
32:36
Speaker 4: But I mean, it sounds like a bit of a
32:39
Speaker 4: dystopia that you're.
32:40
Speaker 2: Describing there to some extent, and you know, I include
32:44
Speaker 2: the thought of AI bodies in there as well.
32:46
Speaker 4: I mean, I just think that sounds absolutely horrific.
32:48
Speaker 2: You know, we've got young people dealing with, you know,
32:53
Speaker 2: inexplicably high levels of anxiety and depression turning to AI
33:00
Speaker 2: to help with that. I don't think that's a cure
33:02
Speaker 2: for loneliness. And you know, we've seen with what happened
33:06
Speaker 2: in Florida with a fourteen year old boy who had
33:08
Speaker 2: been having protracted conversations with character AI and ended up
33:13
Speaker 2: committing suicide. And I'm also wary of AI agents. You know,
33:18
Speaker 2: they sound great and they could be super helpful, but
33:21
Speaker 2: the privacy implications there are considerable as well. I certainly
33:25
Speaker 2: don't feel comfortable just handing over key aspects of my
33:30
Speaker 2: life to an agent to take care of and you know,
33:33
Speaker 2: I think those accuracy. Accuracy is really one of the
33:37
Speaker 2: biggest issues with generative AI at the moment. You know
33:41
Speaker 2: it's a feature, not a bug. It hasn't been solved.
33:45
Speaker 2: Sure RAG can help, but it doesn't go all the
33:48
Speaker 2: way to solving this. So organizations do need to be
33:51
Speaker 2: thinking about their use cases, the risk profile and what
33:57
Speaker 2: kind of mitigants they need to be implementing, whether that's
34:02
Speaker 2: human review. But you know, does you know if you
34:04
Speaker 2: insist on sort of extensive fact checking and someone it
34:08
Speaker 2: kind of undermines the utility sometimes of using generative AI.
34:11
Speaker 2: So it's about really understanding how is this being used.
34:14
Speaker 2: But again tricky because you know it's a general purpose tool,
34:18
Speaker 2: right and it can be used in all sorts of
34:21
Speaker 2: different circumstances. So it is challenging and this is where
34:25
Speaker 2: it comes back to getting the basics right with your AI.
34:28
Speaker 2: Governance training is super important. I mean, there was a
34:32
Speaker 2: case recently in Australia and Victoria where staff of a
34:38
Speaker 2: child protection agency we're using a free version of chatch
34:43
Speaker 2: ept to entering really sensitive information about kids and getting
34:48
Speaker 2: some awful outputs and some really inaccurate outputs as well,
34:54
Speaker 2: and it turns out that none of the staff had
34:56
Speaker 2: been trained on how to use these free tools and
34:59
Speaker 2: acknowledging of course the difference between the free and the
35:02
Speaker 2: subscription tools. So it's that kind of thing that led us.
35:06
Speaker 2: We've developed an e learning module on gener to AI
35:10
Speaker 2: guardrails because I think organizations and staff are crying out
35:15
Speaker 2: to sort of really understand what is and what is
35:19
Speaker 2: not appropriate use.
35:20
Speaker 5: So social license, building it and maintaining it is really
35:24
Speaker 5: important deciding whether you should actually use it. You know,
35:27
Speaker 5: the technology may exist, but is it appropriate to actually
35:30
Speaker 5: unleash this stuff on the world. They've one of the
35:32
Speaker 5: biggest sort of disruptive effects of AI that the analysts
35:38
Speaker 5: have been talking about is on white collar work. That's
35:42
Speaker 5: where it potentially could have the biggest impact. Are you
35:46
Speaker 5: seeing that the hundred projects that you've worked on, are
35:50
Speaker 5: you actually seeing Kiwi's being displaced some of their work
35:54
Speaker 5: being displaced by AI? And what is happening If they are,
35:57
Speaker 5: they are they actually moving up the value chain to
36:00
Speaker 5: do more higher value jobs.
36:03
Speaker 3: Good question.
36:04
Speaker 7: So tapping into something that Friff was saying, as I
36:07
Speaker 7: can to go, this is what's awesome about this conversation
36:10
Speaker 7: is multiple things can be true at the same time.
36:13
Speaker 7: And for example, part of that white collar work that
36:16
Speaker 7: we're engaged with with a client on a global basis.
36:18
Speaker 7: Now is in suicide prevention using AI as a way
36:23
Speaker 7: to prioritize social media trolling from people who are in
36:29
Speaker 7: distressed on social media platforms. If you remember a year
36:33
Speaker 7: or so ago, probably longer than that, the social media
36:35
Speaker 7: platform signed up to needed to provide a triage service
36:38
Speaker 7: that if you were being trolled online or in mental distress,
36:40
Speaker 7: that you could reach out to someone via those platforms.
36:43
Speaker 7: They're just not enough counselors in the world to deal
36:45
Speaker 7: with that influx at all, and therefore we've been working
36:48
Speaker 7: with the organization that's responsible providing that service to go, look,
36:51
Speaker 7: there's an individual who needs help. We just can't find
36:54
Speaker 7: a human They just don't exist. They've not got the
36:56
Speaker 7: right language and not got the right skills. We need
36:58
Speaker 7: to hold that person in that position so they don't
37:01
Speaker 7: cause self harm whilst we go and find a human being.
37:04
Speaker 7: Now we're not trying to replace a human in that instance,
37:07
Speaker 7: but this is a case of look, the drop off
37:08
Speaker 7: right and people causing harm to themselves is increasing because
37:11
Speaker 7: the amount of volume is increasing, and we just can't
37:13
Speaker 7: find the counselors. And that's an agentic use of six
37:17
Speaker 7: different AIS working in tandem to drive accuracy higher. So
37:21
Speaker 7: that's a gentic AI being used to keep a person engaged.
37:25
Speaker 3: Whilst we go and find them a counselor.
37:27
Speaker 7: That feels like it's a good test of whether this
37:29
Speaker 7: technology can be used to create human advantage and human
37:32
Speaker 7: care without necessarily replacing replacing the human And that's kind
37:36
Speaker 7: of one of the first projects that we started on
37:38
Speaker 7: at the start of the year Peter around that, not
37:40
Speaker 7: white collar displacement, but dealing with a productivity gap and
37:43
Speaker 7: dealing with the resourcing gap. We are, however, seeing traditional
37:48
Speaker 7: industries from engineering, audit compliance, which are there just limited
37:53
Speaker 7: by eyes and fingers about how quickly you can get
37:56
Speaker 7: to things and how quickly you can consume information and
37:59
Speaker 7: seeing you just asked the question in a good way,
38:02
Speaker 7: you said displaced and then displacing the work. We've not
38:04
Speaker 7: seen people being displaced. What we're seeing is the work
38:07
Speaker 7: being displaced.
38:09
Speaker 3: What that's allowed for.
38:11
Speaker 7: Is businesses to solve this whole talent shortage issue within
38:15
Speaker 7: an industry and spend more time on covering more opportunities
38:18
Speaker 7: to grow, which means not just stuck beating their head
38:21
Speaker 7: against the same problem over and over again, but have
38:22
Speaker 7: a little bit of capacity to be able to look up.
38:25
Speaker 7: That doesn't mean I don't think payroll line items and
38:30
Speaker 7: people will be displaced in the future.
38:32
Speaker 3: I think that's likely to be the case.
38:33
Speaker 7: You're going to look at the size of your business
38:35
Speaker 7: and say, look, we can do the same with less
38:36
Speaker 7: one hundred percent. Sure, but I don't think we're at
38:38
Speaker 7: that point that point just yet. I think we're freeing
38:41
Speaker 7: smart people from the tedious work. How those business owners
38:44
Speaker 7: decide to apply that value in terms of the P
38:46
Speaker 7: and L.
38:46
Speaker 3: I think you yet to be.
38:47
Speaker 1: Seen, just just because it's very relevant to what Chris
38:51
Speaker 1: and Dave just said about, you know, sort of AI
38:54
Speaker 1: and the Gentics and Buddies and blah blah. We've just
38:57
Speaker 1: been doing quite a lot of work in the space
38:59
Speaker 1: of meant health and at the same time cultural settings
39:05
Speaker 1: in New Zealand versus the rest of the world, and
39:07
Speaker 1: I think it's really important to recognize that other countries
39:11
Speaker 1: and the cultures of other countries have got very different
39:14
Speaker 1: attitudes towards an artificial body versus a human. Japan's probably
39:21
Speaker 1: the easiest examples. So in Japan it's almost the other
39:24
Speaker 1: way around. People actually prefer to speak to an AI
39:28
Speaker 1: or to a roleboder deal with, but then with and
39:30
Speaker 1: there's all sorts of really interesting cultural components. So I
39:33
Speaker 1: just wanted to I think it's really really important whenever
39:36
Speaker 1: we talk about AI and buddies is to recognize our
39:40
Speaker 1: own cultural lens into this and recognizing that it's very
39:45
Speaker 1: different in other parts of the world. Secondly, and this
39:48
Speaker 1: is so we're working with a number of counselors who
39:52
Speaker 1: are doing, you know, in the mental health space right now.
39:55
Speaker 1: And the first things that I learned is that people
39:59
Speaker 1: also while they're in counseling with humans, you know. So
40:03
Speaker 1: it's not like anything that we've currently got is perfect.
40:07
Speaker 1: We just have to again accept that no new technologies
40:10
Speaker 1: that were perfect. Certainly AI is not perfect. And I
40:13
Speaker 1: think we need to be realistic about what we expect
40:16
Speaker 1: from new technology and well, what's the alternative? And Da've
40:19
Speaker 1: made a really good point. We've you do the maths
40:21
Speaker 1: and you look at the the sheer growth of mental
40:25
Speaker 1: health challenges and availability of counselors and the cost of it.
40:30
Speaker 1: It's becoming unaffordable. It's becoming impossible for us as a
40:34
Speaker 1: as a society to actually service that. And I think
40:37
Speaker 1: if you look at evolutionary stuff, often when conditions for
40:42
Speaker 1: life become difficult, some kind of innovation or or or
40:47
Speaker 1: intervention appears which which suddenly changes everything. And I think
40:52
Speaker 1: we are at that point right now with AI.
40:54
Speaker 5: Well, a lot of big issues related to AI this year.
40:58
Speaker 5: It's only going to get more complex fascinating next year. Frith,
41:02
Speaker 5: Dave Steffan, thanks so much for coming on to Business
41:05
Speaker 5: of Tech, and we'll check back in with you next year.
41:07
Speaker 3: Thank you.
41:17
Speaker 6: That really ramped up that conversation. Kind of a nice
41:20
Speaker 6: steady start there, talking generally about the guests experiences, which
41:25
Speaker 6: is all interesting stuff. But when we started to get
41:27
Speaker 6: towards the end, that's when I thought it got quite exciting,
41:31
Speaker 6: in particular talking about the potential for the you know
41:36
Speaker 6: AI buddy, AI agents, personal AI agents and Stephan Korn
41:42
Speaker 6: really going hard and talking straight aback. It's going to
41:45
Speaker 6: be like her, it's going to be a friend, it's
41:48
Speaker 6: going to solve your loneliness problems, which was you know, ambitious,
41:54
Speaker 6: which is great.
41:56
Speaker 3: I'm glad that.
41:57
Speaker 6: Frith was there to kind of put the brakes on
42:00
Speaker 6: a little bit and say, hang on a minute, maybe
42:02
Speaker 6: we need to think about this which I guess for
42:05
Speaker 6: me the one thing that I was thinking in particular
42:09
Speaker 6: with the AI loneliness aspect, and Frith mentioned obviously the
42:13
Speaker 6: character AI.
42:15
Speaker 3: Controversy that happened earlier where a.
42:17
Speaker 6: Young man was talking to a Denaris Targerian character AI
42:21
Speaker 6: and very sadly took his own life. And while that
42:28
Speaker 6: exact scenario may not happen again, what I do think
42:32
Speaker 6: it speaks to is this idea of unintended consequences that
42:37
Speaker 6: sometimes technology boosters complete one hundred percent optimists can sometimes
42:44
Speaker 6: forget that we can't know everything that's going to happen,
42:49
Speaker 6: and we can have voices saying, oh, what about this
42:52
Speaker 6: might happen? Be careful of this and put the guardrails
42:54
Speaker 6: in around here. But there are also going to be
42:57
Speaker 6: places where we don't necessarily see some things coming. And
43:03
Speaker 6: in terms of addressing loneliness, I was a little bit
43:06
Speaker 6: alongside Frith where she called it this dystopian idea where
43:10
Speaker 6: the solution to loneliness is like digital robots. It felt
43:15
Speaker 6: a little bit an airing to me in particular because well,
43:24
Speaker 6: I think when Stephan said, you know, pointed to kind
43:27
Speaker 6: of Japan as being a country that was more likely
43:30
Speaker 6: to engage with chatbots. He pointed to one of the
43:35
Speaker 6: most long lasting loneliness epidemic countries that we've known about
43:39
Speaker 6: for a really long time. It's actually, you know, a
43:44
Speaker 6: real been a serious issue to the point where it's
43:46
Speaker 6: almost memified the loneliness issue in Japan over the last
43:50
Speaker 6: kind of decades. So I thought that was an interesting
43:53
Speaker 6: pull as an example of a country that's more engaged
43:57
Speaker 6: with AI for these issues.
43:59
Speaker 3: So, yeah, I.
44:03
Speaker 1: Will.
44:03
Speaker 6: I'm excited for the idea of a productivity buddy who
44:07
Speaker 6: can help me with work and life admin, I'm not
44:11
Speaker 6: sure I want them to be my mate personally.
44:13
Speaker 5: Yeah, I've been thinking a lot about this particular issue.
44:17
Speaker 5: I've got a friend who very tragically is wife took
44:22
Speaker 5: her own life this year, and she was a digital native,
44:26
Speaker 5: so from the here, from the very earliest stages, she
44:30
Speaker 5: was creating content, living her life online and kept all
44:35
Speaker 5: of that. And he now has this entire archive, and
44:38
Speaker 5: he's a tech guy, and he's he's actually considering using
44:42
Speaker 5: a large language model to create a model of his
44:45
Speaker 5: wife and feed in all of this all of this data,
44:49
Speaker 5: and for him it's away, you know, he's going through
44:51
Speaker 5: He's incredibly lonely at the moment, so to be able
44:54
Speaker 5: to hear her voice. He's got all these recorded episodes
44:56
Speaker 5: she's done of podcasts and interviews and that sort of thing,
45:00
Speaker 5: so you can sort of see it is quite compelling.
45:03
Speaker 5: But I think philosophically, where we're at with AI at
45:06
Speaker 5: the moment is that we've got all these tools at
45:08
Speaker 5: the moment, AI tools, but the industry wants to move
45:12
Speaker 5: towards artificial general intelligence opener AI has said that is
45:17
Speaker 5: its goal is to do that in a responsible way
45:20
Speaker 5: or a way that is positive for humanity. And you
45:23
Speaker 5: can argue about how quickly they're moving towards that, but
45:26
Speaker 5: I feel like we've got a few more years of
45:29
Speaker 5: just making really good tools, so boosting productivity and efficiency,
45:34
Speaker 5: creating videos very quickly, using text to video generators, all
45:41
Speaker 5: of that sort of stuff, before we race on to
45:43
Speaker 5: the things that are trying to really mimic a human being,
45:46
Speaker 5: like a Herd type scenario, which does have potential downsides.
45:50
Speaker 5: Let's get these tools right and lay the foundations. And
45:53
Speaker 5: that's what I liked about Dave's approach. He's not being
45:56
Speaker 5: captured by the technology. He's focused on turn on investment.
46:01
Speaker 5: He's actually telling companies, let us know how many people
46:04
Speaker 5: you employ, the sorts of tasks you do, feed it
46:07
Speaker 5: into our ROI estimator, and see how much money you
46:12
Speaker 5: will likely save if you employ these types of tools.
46:15
Speaker 5: That's what we really need at the moment is actually
46:18
Speaker 5: prove the case that this can be a productivity driver.
46:22
Speaker 5: That's what we've been promised for years now, So hopefully
46:25
Speaker 5: twenty twenty five is a year where we actually see
46:27
Speaker 5: that come to fruition.
46:29
Speaker 6: One hundred percent agree with you, And it's a conversation
46:31
Speaker 6: I was having with Business Desks AI expert and residence
46:35
Speaker 6: Andy Freya yesterday where we were talking about this next,
46:41
Speaker 6: this current and kind of twenty twenty five wave of
46:46
Speaker 6: AI productivity tools. We see it as being really internally
46:51
Speaker 6: focused for organizations creating stuff that is for employees to
46:57
Speaker 6: kind of give them that boost for or productivity because
47:01
Speaker 6: that's extremely low risk. So I think that's a really
47:07
Speaker 6: really important learning curve that we're on right now around
47:10
Speaker 6: these generative AI tools and deploy them internally, be really
47:16
Speaker 6: clear about where things go wrong if the employee manages
47:18
Speaker 6: to do something stupid or get it to do something stupid,
47:21
Speaker 6: don't just go, oh, that's okay.
47:23
Speaker 3: It was just the.
47:23
Speaker 6: Silly employee, go a silly employee. You're not supposed to
47:26
Speaker 6: do that. What can we learn about this? How can
47:28
Speaker 6: we fix it? How can we make sure that we
47:29
Speaker 6: implement things so that if we do go to external stakeholders,
47:33
Speaker 6: it's not going to cause a serious issue. So that,
47:37
Speaker 6: to me, I think is going to define twenty twenty
47:39
Speaker 6: five in GENAI is internal tools for internal use and
47:44
Speaker 6: gathering a lot of learning and education, and it's something
47:47
Speaker 6: that I think we shouldn't be encouraging New Zealand companies
47:50
Speaker 6: to do as well, because they know as long as
47:53
Speaker 6: we're focusing on that ROI that productivity gains the actual
47:57
Speaker 6: realization of improvements.
48:00
Speaker 5: Yeah, one of the areas where we will see improvement,
48:05
Speaker 5: and Stepan touched on it was around generative agents autonomous
48:09
Speaker 5: agents acting on your behalf to do tasks. That increases
48:13
Speaker 5: the risk somewhat because they could make a mistake, especially
48:17
Speaker 5: when agents are talking to each other on the behalf
48:20
Speaker 5: of a customer and a company for instance. But that's
48:24
Speaker 5: the obvious next step and we've seen them released by Salesforce,
48:29
Speaker 5: Microsoft and others this year. So I think generative AI
48:32
Speaker 5: is going to be a big one next gen voice assistance.
48:34
Speaker 5: You've already seen it with the updated Seri and the
48:37
Speaker 5: iPhone Alexa will get upgraded with generative AI in twenty
48:42
Speaker 5: twenty five. So is that actually a good experience? Will
48:45
Speaker 5: that because that business has lost a lot of money,
48:48
Speaker 5: will it turn it around and actually make it a
48:51
Speaker 5: profit center for some of these companies. That's another big one,
48:54
Speaker 5: the AI sustainability challenge. The fact that and we've been
48:58
Speaker 5: talking a lot about this this year, where the tech
49:01
Speaker 5: industry is really interested in nuclear because they realize they're
49:04
Speaker 5: running out of power to run these big data centers
49:08
Speaker 5: for AI. But also where we see AI start to
49:13
Speaker 5: have a positive impact on sustainability in terms of in
49:16
Speaker 5: various industries allowing things to be done more efficiently. That
49:20
Speaker 5: that's going to be a big one. And in the
49:22
Speaker 5: you know, the generative video, we just saw some glimpses
49:26
Speaker 5: of what's possible with that this year, which you know,
49:28
Speaker 5: Open Ai, Sora and other things like that incredible results.
49:33
Speaker 5: Will we get to the point next year where we
49:35
Speaker 5: have longer videos that are quicker to create and are
49:39
Speaker 5: actually starting to go into advertising and the film and
49:42
Speaker 5: TV space that's potentially the path it's on.
49:46
Speaker 6: Yeah, And I mean we saw the first jen Ai
49:50
Speaker 6: created TV advert this year, and obviously there was a
49:54
Speaker 6: lot of finagling and manipulating to make it how they
49:57
Speaker 6: wanted it to. But that will add with the the strange,
50:00
Speaker 6: very obviously AI generated wool cuddling up to a generative woman,
50:06
Speaker 6: which was, you know, an interesting advert. Also was a
50:10
Speaker 6: kind of a glimpse of the potential future for general
50:14
Speaker 6: you know, the way that we're using and we've got
50:16
Speaker 6: Skinny as well now talking about using a deep fake.
50:21
Speaker 6: They're going to pay an actor or a model or
50:23
Speaker 6: even somebody off the street to use their likeness as
50:28
Speaker 6: a spokesperson for two years and what is that going
50:33
Speaker 6: to mean?
50:33
Speaker 3: What are they going to actually make this AI model.
50:37
Speaker 6: Do and how how are they going to you know,
50:40
Speaker 6: balance that and will it.
50:41
Speaker 5: Be worth a lifetime of free Skinny subscriptions.
50:45
Speaker 6: There's also there's also a fee involved, so there was
50:48
Speaker 6: also a yeah, they also are paying what was considered
50:53
Speaker 6: I should have a story coming out about it, surely,
50:57
Speaker 6: But yeah, fascinating, fascinating time, and I think the end
51:00
Speaker 6: of twenty twenty four in terms of what's happening with
51:02
Speaker 6: the generative AI in New Zealand is a super interesting
51:07
Speaker 6: indicator for what's coming down the line next year.
51:10
Speaker 1: Yeah.
51:10
Speaker 5: On the positive side, I think we avoided any major
51:13
Speaker 5: disasters with AI this year, you know, in terms of
51:17
Speaker 5: election interference, you know major you know, deep fake scandal
51:22
Speaker 5: or something like that.
51:23
Speaker 3: We had.
51:24
Speaker 5: We had real people, We had the Joe Rogan's, We
51:26
Speaker 5: had the Elon Musks being influential doing crazy things. There
51:31
Speaker 5: was things that like Trump's assassination attempt on them, you know,
51:36
Speaker 5: that was a game changer in terms of the trajectory
51:39
Speaker 5: of the election campaign. It didn't take some random Russian
51:44
Speaker 5: inspired deep fake campaign to do that, So so that's good.
51:48
Speaker 5: I think people still have the head screwed on. They
51:52
Speaker 5: are skeptical naturally off a lot of the stuff that
51:55
Speaker 5: the content itself still looks AI generated. So I think
51:58
Speaker 5: every human it just doesn't compute with with our brains.
52:02
Speaker 3: We just reject it.
52:03
Speaker 5: But you know that could change in the next year.
52:05
Speaker 6: Yeah, I think it will change eventually. And I think
52:10
Speaker 6: what we're going to see as a move from these
52:12
Speaker 6: really crisp, clean, glossy productions from generative AI, and we're
52:17
Speaker 6: going to start to see them like trying to layer
52:20
Speaker 6: that natural chaoticness over them, to muddy them, to kind
52:24
Speaker 6: of make them seem more natural. That's what we're going
52:27
Speaker 6: to start to see next, and that's when we're going
52:29
Speaker 6: to start to really be in trouble, I think. But
52:31
Speaker 6: in terms of the sustainability as well, I think that's
52:35
Speaker 6: something that is going to become more and more talked about.
52:38
Speaker 6: And you know, ideally there will be some market forces
52:43
Speaker 6: involved where if it gets if the power consumption gets
52:47
Speaker 6: too expensive, then it's going.
52:50
Speaker 3: To restrict what you can actually do with the tools.
52:53
Speaker 6: Ideally, the only concern is if we have these companies
52:57
Speaker 6: really you know, letting the a IB a huge loss
53:01
Speaker 6: leader in order to a lot for long term returns.
53:05
Speaker 6: You know, so that's lock people in at low prices
53:08
Speaker 6: and then eventually raise the prices drastically when we realize
53:11
Speaker 6: that that's what's needed. Yeah, So that that is a
53:14
Speaker 6: strategy we've seen before in the tech industry.
53:16
Speaker 5: Interesting has thoughts on open source AI and not seeing
53:20
Speaker 5: uptake of that because it's ultimately more expensive to do
53:23
Speaker 5: something with that. You might have more control to do stuff,
53:26
Speaker 5: but it's that commoditization for new Zealand companies in particularly
53:30
Speaker 5: being able to buy a service in the cloud, be
53:33
Speaker 5: able to do all your AI workloads there for a
53:37
Speaker 5: reliably understandable fee on a monthly basis. That's what sort
53:42
Speaker 5: of our companies are looking for at the moment.
53:44
Speaker 6: Yeah, definitely, and that may eventually change as models, the
53:50
Speaker 6: efficiency of models changes, And we spoke to Kiarri and
53:53
Speaker 6: Dagua earlier this year and that was one of the
53:55
Speaker 6: things that he's really been working on as part of
53:58
Speaker 6: a cohort international on is making these models more efficient,
54:03
Speaker 6: making them use consume less compute. So if we do
54:08
Speaker 6: see start to see like a Moore's law in terms
54:11
Speaker 6: of reducing the amount of compute that is needed for
54:15
Speaker 6: any particular action towards generative AI, that may eventually turn
54:20
Speaker 6: the table for open source in a way that Stephan
54:26
Speaker 6: was kind of suggesting. And also as we see like
54:31
Speaker 6: we talk, I think did Stefan mention red hat, I
54:34
Speaker 6: think where he there as a company that kind of
54:36
Speaker 6: commoditized open source, we might start to see that kind
54:39
Speaker 6: of thing happening as well. So yeah, that shift will
54:43
Speaker 6: will eventually happen, but I think definitely in the short term,
54:46
Speaker 6: it's the big subscription models that are going to dominate.
54:50
Speaker 5: Yeah, So some great discussion there, and we'll keep our
54:53
Speaker 5: focus on AI and twenty twenty five. We've had a
54:55
Speaker 5: lot of guests on a lot of themes we've explored,
54:58
Speaker 5: but it is the technology of the moment, so we
55:01
Speaker 5: will keep a close.
55:02
Speaker 3: Eye on it.
55:03
Speaker 5: So thanks to Frith Tweety, Dave Howden and Stephan Korn
55:07
Speaker 5: for that look back at the year in AI.
55:10
Speaker 6: Yes, thank you very much. Show notes for the Business
55:13
Speaker 6: of Tech are in the podcast section at Business Desk
55:15
Speaker 6: dot co dot Nz, where you can stream this podcast
55:18
Speaker 6: in full every week. It's also available from iHeartRadio or
55:22
Speaker 6: on your podcast platform of choice.
55:24
Speaker 5: Get in touch with your feedback and we'd love to
55:26
Speaker 5: hear your suggestions for upcoming guests. To email Ben benat
55:30
Speaker 5: Business Desk dot co dot.
55:32
Speaker 6: Nz and you can find Peter and I both on
55:35
Speaker 6: x and LinkedIn, and you can also follow the Business
55:38
Speaker 6: of Tech LinkedIn page and you'll get all our updates
55:41
Speaker 6: for new episodes there, as well as other content that
55:44
Speaker 6: we produce.
55:46
Speaker 5: That's it for this week. Another episode coming your way
55:48
Speaker 5: next Thursday.
55:49
Speaker 6: Until then, have a great week.