WEBVTT
00:00:09.199 --> 00:00:12.800
Welcome to the tech episode of Dancing with Systems.
00:00:12.960 --> 00:00:16.239
What a timely discussion to have at this moment in time.
00:00:16.640 --> 00:00:21.600
We know that tech is both the enabler of every other systems transition, more of that soon.
00:00:21.920 --> 00:00:26.079
But it's also a sector in its own right with a rapidly growing footprint.
00:00:26.399 --> 00:00:47.920
And right now, those two, a force for good, a force with its own impacts, are now really intentional and the AI boom is a sort of epicenter of that defining dynamic, with data centers becoming one of the world's fastest growing consumers of water and of energy, and often concentrated in communities where people don't want them.
00:00:48.479 --> 00:00:52.320
So the narrative is turning in a way that I don't think we've seen before.
00:00:52.399 --> 00:00:56.719
And really interesting, the narrative is turning in a way that's hopping over politics.
00:00:57.280 --> 00:01:00.320
People are getting fearful of AI.
00:01:00.719 --> 00:01:06.560
And so what a great opportunity to talk about tech and systems transitions.
00:01:06.959 --> 00:01:13.760
Yeah, well, uh this is a fantastic episode and so central to everything that we're living with.
00:01:14.079 --> 00:01:18.159
Let me put a few numbers to some of those uh opening remarks that you made.
00:01:18.400 --> 00:01:29.200
So uh the the five big tech companies are are now spending more than the oil and gas industry on capital investment.
00:01:29.599 --> 00:01:34.319
And uh uh so that's 400 billion uh in 2025.
00:01:34.719 --> 00:01:40.000
Uh that's expected to jump up to 75% this year in 2026.
00:01:40.239 --> 00:01:41.840
So wow.
00:01:42.159 --> 00:01:44.319
Yeah, it's being supercharged, right?
00:01:44.640 --> 00:01:58.159
Uh and of course the the big uh uh link there is to data centers, and and we've seen some uh stepping back from commitments by those big tech companies, we've seen some concerns around emissions.
00:01:58.400 --> 00:02:01.439
Uh we do need to keep it in perspective, right?
00:02:01.519 --> 00:02:11.360
Uh data centers account for around one and a half percent of global uh consumption of electricity, so you know, much less than some other sectors.
00:02:11.680 --> 00:02:16.639
Um but it has been growing at around 12% since 2017.
00:02:17.280 --> 00:02:19.280
So, you know, it is ramping up.
00:02:19.360 --> 00:02:23.360
Uh that's four times faster than electricity consumption in other areas.
00:02:23.599 --> 00:02:27.199
So, you know, it is something that we have to be concerned about.
00:02:27.520 --> 00:02:37.520
Um but there is an open question uh, you know, could AI actually deliver some of the emission reductions as a result of all the benefits it brings?
00:02:37.599 --> 00:02:46.000
And at the moment there's an estimate by the International Energy Agency of uh reductions of perhaps 5% by 2035.
00:02:46.319 --> 00:02:52.319
Um so you know, we're we're I I guess that's something we might explore in the in this episode as well.
00:02:52.639 --> 00:02:53.680
Um so yeah.
00:02:54.560 --> 00:03:03.439
So lots of big numbers there, and also a story of exponential uh growth as well in terms of that capital investment number.
00:03:03.680 --> 00:03:06.319
Um so we're talking about tech.
00:03:06.560 --> 00:03:09.120
What is it, Wayne, that you're interested to hear about?
00:03:09.840 --> 00:03:19.199
Yeah, well, I think it makes sense that we're gonna focus in uh really on the tech of the day, which is uh which is AI and and data and so on.
00:03:19.599 --> 00:03:23.680
And what we know is that this is now the great accelerator.
00:03:24.080 --> 00:03:31.360
But the question really is is it going to accelerate the transition that we're looking for in all these other sectors?
00:03:31.680 --> 00:03:35.439
Or is it going to accelerate the decline, the problem?
00:03:35.520 --> 00:03:39.039
Uh so that's a big question, something I'd like to hear more about.
00:03:39.439 --> 00:03:45.599
Um and you know, we do hear a lot, especially on social media, about the impact of data centers.
00:03:46.000 --> 00:03:57.360
And I'm just wondering whether we're getting caught in a little bit of hype here and whether it really is as bad as social media seems to suggest that it is.
00:03:57.520 --> 00:04:00.800
So I'd I'd love some perspective from our guests on that.
00:04:01.120 --> 00:04:03.280
Um so uh what about you?
00:04:03.520 --> 00:04:04.800
What are you hoping to hear?
00:04:05.360 --> 00:04:11.280
I'm really curious about how does accountability for AI work?
00:04:11.840 --> 00:04:14.400
Is it down to the governments to develop policies?
00:04:14.639 --> 00:04:20.720
Because at the moment the policy framework is really lagging the exponential growth of particularly AI.
00:04:21.040 --> 00:04:22.560
Is it the hyperscalers?
00:04:22.959 --> 00:04:24.720
And intriguingly, is it us?
00:04:25.199 --> 00:04:33.439
Because we're beginning to understand that AI has a positive impact, but also there is a lot of fear about job losses, etc.
00:04:34.079 --> 00:04:40.480
So, where does accountability lie in order to ensure that AI is on balance of force for good?
00:04:40.720 --> 00:04:42.000
That's what I'm interested in.
00:04:42.399 --> 00:04:47.199
Yeah, and I guess that's gonna link also to the social elements, right, of AI.
00:04:47.439 --> 00:04:50.879
So um well, let's jump in and uh who've we got first.
00:04:51.519 --> 00:04:52.639
We've got Joel Makkawa.
00:04:53.199 --> 00:04:54.879
Okay, do you want to give us an introduction?
00:04:55.040 --> 00:04:55.519
Who is Joel?
00:04:55.920 --> 00:04:56.319
I can.
00:04:56.720 --> 00:05:04.560
Joel um is a wonderful person, but he's also the chair and co-founder of Trellis Group, who were formerly Green Biz Group.
00:05:04.720 --> 00:05:07.199
I think people are still getting used to that name change.
00:05:07.519 --> 00:05:15.600
Um Green Biz, Trellis, as they are now, a leading media and events platform at the intersection of business sustainability and clean technology.
00:05:16.000 --> 00:05:20.000
He's also the strategy director of the Music Sustainability Alliance.
00:05:20.480 --> 00:05:21.120
I need to go on.
00:05:22.720 --> 00:05:30.639
Um he's a former nationally syndicated columnist, he is a journalist by training with over 35 years covering sustainable business.
00:05:30.959 --> 00:05:35.199
He was one of the first journalists I remember talking about sustainability and business.
00:05:35.439 --> 00:05:40.720
And he's the author of more of a dozen books and lead author of the annual State of the Green Business Report.
00:05:41.120 --> 00:05:50.399
The Associated Press has called him the guru of green business practices, and he is inducted into the International Society of Sustainable Professionals Hall of Fame.
00:05:50.959 --> 00:05:53.839
The perfect person to set the scene on tech for us.
00:05:55.279 --> 00:05:55.680
Great.
00:05:56.079 --> 00:05:57.120
Let's hear from him now.
00:05:57.680 --> 00:06:01.519
Welcome, Joel, to this uh episode of Dancing with Systems.
00:06:02.079 --> 00:06:15.839
It's such a pleasure to have you uh as one of our veterans who's also been uh focused on tech for a long time and taken a special interest uh throughout your journalistic career, especially.
00:06:16.160 --> 00:06:23.519
And so uh we'd like to kick off by uh asking you you know, there are two ways to look at tech.
00:06:23.680 --> 00:06:30.240
One is that tech is an enabler for uh transitions in other sectors, and so we'd like to start with that.
00:06:30.480 --> 00:06:34.399
But then of course there is the tech sector itself which has its own impact.
00:06:35.040 --> 00:06:49.199
But maybe you could start by giving us your view of of how tech can be and maybe is being used to help the other sectors uh transition to net zero, nature positive, and so on.
00:06:50.160 --> 00:06:59.120
Well, you're right in that uh tech is both an enabler of of tr uh other transitions and also a sector with its own footprint.
00:06:59.279 --> 00:07:02.959
And right now, those two faces are in tension.
00:07:03.120 --> 00:07:13.759
Um the same data center build-out that's powering AI, for example, is becoming one of the largest sources of demands on energy, water, land, and and other things.
00:07:14.000 --> 00:07:29.360
So we're at this really interesting moment, and and I know that there's a debate is is, I think putting it lightly, around, you know, what is the purpose, the value of AI to society, uh, and and for whom?
00:07:29.519 --> 00:07:40.319
And so right now we're I think we're having that discussion in in a significant way, um, and and understanding that it it's not, you know, there's there's both sides.
00:07:40.560 --> 00:08:10.720
So uh, you know, simply on the enabling side, when you look at pretty much all of the challenges that we as humanity are facing: energy, water, materials, toxicity, the carbon cycle, hydraulic cycle, uh, pretty much everything, uh the the energy, electricity grid, these are among the most incredibly complex systems that we know of and we barely understand.
00:08:11.120 --> 00:08:22.720
Um in addition, these things are are are are spewing data uh 24-7, uh, whether it's about plants or about you know energy plants or or actually green plants.
00:08:23.040 --> 00:08:43.200
And um and and so making sense of that, understanding it, and being able to to optimize and predict and monitor those things is is really one of uh technology's uh great uh advances here uh and has an incredible promise.
00:08:43.360 --> 00:08:53.600
But I think the irony here is that we can now measure emissions and planetary impact far better than we can reduce them.
00:08:54.559 --> 00:08:55.200
Yeah.
00:08:55.679 --> 00:08:57.840
Yeah, and I guess that's the question.
00:08:58.240 --> 00:09:08.480
If you think about those different sectors and transitions that we're trying to enable, where do you see that technology is already having that that impact?
00:09:08.879 --> 00:09:15.360
We're already seeing it scaling, because it's not only about the data, it's also about new technologies, literally.
00:09:15.679 --> 00:09:20.559
And and where do you see where it's kind of lagging behind and there's still a gap?
00:09:21.440 --> 00:09:25.759
Well, first of all, we should acknowledge that technology is a very, very broad term.
00:09:25.919 --> 00:09:26.080
Yeah.
00:09:26.240 --> 00:09:39.519
Uh it can it we you know, we're these days we often use it to re to refer to to digital technologies, um, from from phones and computers to AI systems and and and and on and on.
00:09:39.759 --> 00:09:42.639
Um, but technology goes much broader than that.
00:09:42.720 --> 00:10:07.919
So I I'm thinking I'm assuming we're gonna be you know sticking to the you know the more recent uh uh way we talk about technology, which is uh uh, and I think this is where we're seeing the the positive impact so far is is in digital measurement and transparency, uh, infrastructure, sensors, satellites, supply chain traceability, uh grid optimization.
00:10:08.080 --> 00:10:15.279
Um these are things that don't necessarily decarbonize directly, but but they make accountability possible.
00:10:15.519 --> 00:10:22.000
And and and that's uh a significant change from the past where we didn't really understand what was going on.
00:10:22.320 --> 00:10:31.279
Um uh and and that's and and that really is a precondition for you know pretty much all of the other changes we we we want to make and need to make.
00:10:31.519 --> 00:10:36.559
Um so again, as I said before, we know how to measure things and monitor things.
00:10:36.639 --> 00:10:40.639
We just don't know how to necessarily how to change them uh for the better.
00:10:42.159 --> 00:10:48.000
And I guess that then leads us to the transition of the tech sector itself.
00:10:48.240 --> 00:11:02.240
So clearly we know that tech is already enabling us to uh map deforestation, it's giving us access to early warning systems for farmers, it's giving us access to data on air pollution incidents.
00:11:02.399 --> 00:11:08.320
So tech as an enabler of these other systems transitions, you know, as we've been discussing, is clear.
00:11:08.480 --> 00:11:15.679
But then if we look at the impact of the tech sector itself, um where where do you think we are, Joel?
00:11:15.759 --> 00:11:21.120
And and what are the sort of the big challenges that we need to overcome?
00:11:21.360 --> 00:11:24.080
Does that take us straight back to a data center?
00:11:24.720 --> 00:11:36.480
Well, inevitably it does because so much of what we're talking about is either happening in the cloud, which originates in a data center, or in AI, which of course is a data center as well.
00:11:36.559 --> 00:11:57.919
And and and and largely as a result of data centers and and all the devices that flow into and out of them, is that the sector's own transition is going backwards on the metrics that matter most because AI, the compute demand is outpacing the efficiency and clean energy gains.
00:11:58.320 --> 00:12:12.240
Um, and so uh the biggest tech companies, the the Googles and the Metas and ChatGPT and all the others are open AI, uh, their emissions are actually going up.
00:12:12.639 --> 00:12:17.919
And at a time when they're they're they're they had committed to making significant reductions.
00:12:18.480 --> 00:12:21.519
Um and so that's a challenge.
00:12:21.600 --> 00:12:24.799
And you know, we the good news is we have a lot of those solutions.
00:12:24.960 --> 00:12:35.200
We know how to uh to drive data centers uh or create data centers that are energy and water and possibly positive.
00:12:35.360 --> 00:12:38.320
Uh, we're seeing some advancement advances there.
00:12:38.799 --> 00:12:47.759
Um and and you know, the question is, you know, how do these things not just exist uh on the planet, but also exist within communities?
00:12:47.919 --> 00:12:56.480
And that's one of the challenges because uh people are afraid uh of technology in a way that I haven't seen in in a long, long time.
00:12:56.799 --> 00:13:10.080
People weren't afraid of PCs, they weren't afraid of smartphones, they weren't afraid of the internet, although all of those things had chat brought challenges, but not the kinds of fears that we're seeing now with data centers and AI.
00:13:10.320 --> 00:13:18.799
And so uh, you know, the societal impacts uh and and the backlash that we're now seeing around the world.
00:13:18.960 --> 00:13:31.919
Um there's a stat uh here in the United States in a Gallup poll that said that uh more people would prefer to live near a nuclear power plant than in a data center.
00:13:32.159 --> 00:13:33.600
I mean, that's astounding.
00:13:34.399 --> 00:13:48.320
Even if it's not exactly true, even if the people are exaggerating, uh just the thought of that or the fact that it might even be on a par with uh with a nuclear power plant is just an amazing uh uh development.
00:13:48.480 --> 00:14:13.919
And so uh we need to be addressing uh and a lot of those things are are both environmental fears, fears of uh these data centers uh uh uh using too much energy and water, but also the fears that the technology itself brings, terms of job loss, uh changes that people aren't necessarily looking forward to, uh child pornography, so many challenges, uh cybercrime and so on.
00:14:14.159 --> 00:14:25.759
So uh, you know, if you if you include go beyond the planetary boundaries and look at the societal boundaries, if you will, uh these these are uh have some significant challenges.
00:14:26.399 --> 00:14:40.159
And so, Joel, just before I am back to Wayne, it seems to me in a way that the hyperscalers, the big tech players, they've sort of lost control of the narrative a little bit here, haven't they?
00:14:40.559 --> 00:14:53.120
And you know, as a journalist, as an award-winning journalist with lots of experience in this field, what's your sense of how might they contribute to a different narrative?
00:14:53.279 --> 00:15:01.440
Because it seems to me, unless we can shift the narrative, this kind of fear is just going to keep growing.
00:15:02.320 --> 00:15:04.399
Well, that's a really tall order, Sally.
00:15:04.559 --> 00:15:09.759
I think uh the technology in and of itself isn't going to shift the narrative.
00:15:10.000 --> 00:15:44.240
What's going to shift the narrative is are are the data center providers, the AI companies, uh the hyperscalers uh coming in and showing the benefits uh and and working closely with communities to help them understand, uh first of all, address their concerns and then understand how these can be helpful in optimizing systems, in creating new opportunities, uh in uh uh reducing the kinds of uh negative impacts on the planet that we we know all too well.
00:15:44.559 --> 00:16:01.759
Um, but those right now that that these companies have been far more enamored with their own technology than than trying to make that make sense to the people who are impacted most, and those right now are in in local communities.
00:16:02.080 --> 00:16:17.600
So uh, you know, at some point there will be, as I said before, energy and water positive data centers that will actually be welcome in the communities because they'll be providing some uh environmental and social goods.
00:16:17.919 --> 00:16:23.759
Um and the, as I said, those technologies exist, but we're not that they're a ways off.
00:16:23.840 --> 00:16:52.399
And and certainly at the scale, scope, and speed of the data center build-out, particularly in China and the United States, but around the world, we're not going to be seeing those innovations anytime soon simply because most data centers are you know only maybe incrementally improved from the previous one, and very few data center builders have been fired for making the next data center look pretty much like the last one.
00:16:52.559 --> 00:17:04.400
And so getting the uh some of the technologies that have been around for years, the uh waterless cooling technologies, for example, uh are still not as widespread as they could be.
00:17:04.559 --> 00:17:26.799
And we're just now starting to see the energy technologies develop that can uh, you know, there's this uh trend in in in, at least in the States, maybe globally, called bring your own power, where in order for a data center to build in a community, they have to provide their own electricity needs on-site generation, whatever form that is.
00:17:27.200 --> 00:17:43.359
So until those become widespread and people start to see, oh, these are these are actually uh you know great benefits to our communities, then I think they'll start to you know look at the tech what the technology actually does.
00:17:43.680 --> 00:17:50.160
Um but until until then, uh it's gonna be uh I think a really bumpy ride for the tech sector.
00:17:51.039 --> 00:18:01.599
Yeah, as you're talking, Joel, it it reminds me of uh of Kenneth Bolding's uh characterization of you know the cowboy economy versus the spaceship economy.
00:18:01.680 --> 00:18:13.200
And it does feel at the moment, especially with uh big tech and AI, you know, it's it's uh it's a wild west, you know, and uh where there's seemingly no frontiers, at least that's how they're behaving.
00:18:13.599 --> 00:18:18.240
Um but also uh as you're talking, uh two two other thoughts came to mind.
00:18:18.559 --> 00:18:21.440
One is are we suffering from technology lock-in?
00:18:22.000 --> 00:18:31.839
It feels like quite a new technology, but but the kinds of setups that we have for for data centers, uh, as you say, we're kind of replicating them, but there are better ways to do them.
00:18:31.920 --> 00:18:39.200
It's just that you know, now that we've we've already uh invested in this kind of setup, it seems hard to change.
00:18:39.839 --> 00:18:46.799
Um and and and my my sort of f follow-up and final question really is so where does that leave us?
00:18:46.960 --> 00:18:57.440
What are the levers for changing this behavior, for getting to the kinds of uh efficiencies that we believe are probably out there?
00:18:57.599 --> 00:18:58.960
It's very early days, right?
00:18:59.039 --> 00:19:04.480
So this should go down an efficiency curve as well, uh, and and an innovation curve.
00:19:04.640 --> 00:19:07.039
So um where do you see the main levers?
00:19:08.000 --> 00:19:15.440
Well, in in simply put, we need to set a high bar and a and uh and a high floor.
00:19:15.680 --> 00:19:27.440
So we need to uh in communities need to uh really start pressing uh data center builders in ways that they haven't yet.
00:19:27.599 --> 00:19:37.839
There as much protest as there has been, as much of a big of a backlash as there been has been, it really hasn't had gotten traction as a national movement.
00:19:38.000 --> 00:19:49.119
I think uh we're gonna see that start to shift uh here in the United States, for example, it's gonna become uh a big issue in the in the uh 2026 midterm elections.
00:19:49.279 --> 00:19:50.880
We're seeing that already.
00:19:51.039 --> 00:20:00.480
Um, and and as it becomes a sort of a cause celeb for um and it crosses over uh left and right politics, left, right, and center.
00:20:00.559 --> 00:20:04.640
Uh it's it's amazingly bi or tripartisan here.
00:20:04.880 --> 00:20:12.559
Um so so we need uh in order to do that, we need some kinds of uh uh regulations or legislation.
00:20:12.720 --> 00:20:14.960
The data center companies want that.
00:20:15.119 --> 00:20:16.720
Uh the builders want that.
00:20:16.880 --> 00:20:26.480
I think the entire ecosystem says we need to, you know, some kinds of of guidelines, some kinds of guardrails of what we can build.
00:20:26.640 --> 00:20:29.359
Um and communities need to really press on that.
00:20:29.440 --> 00:20:35.680
And so it's it's really a combination of of laws and social pressure that's going to change things.
00:20:35.839 --> 00:20:41.839
As I said, the technology's been around, and but it there's haven't been any incentives for using it.
00:20:42.000 --> 00:20:48.240
Um and the technology, you know, it continues to improve like all you know, technologies do.
00:20:48.400 --> 00:20:57.680
Uh, the energy technologies, the waterless cooling technologies, all of those things uh are really uh year over year beginning to change.
00:20:57.759 --> 00:21:06.960
So it's inevitably, and and as it is with so many of these challenges, it's it's a number of forces working together over time.
00:21:07.839 --> 00:21:19.440
But there is a movement that is really shaping up, and I think it's going to be a really politically powerful movement that's gonna transcend uh the political parties as we've known them.
00:21:19.519 --> 00:21:32.400
And it'll be really, really interesting to watch uh how frustrated and angry people are uh and scared uh about these technologies and what it will take to change them.
00:21:32.559 --> 00:21:40.400
I I wish I had an answer of what that magic wand would be, but um I don't, and I don't think anyone else does yet at this point.
00:21:40.880 --> 00:21:41.599
Thanks, Joel.
00:21:41.680 --> 00:21:53.440
Well, I I'm glad also that you you ended on on social movements because this is such a critical uh lever in in systems change and and maybe one that we uh we don't always pay enough attention to.
00:21:53.599 --> 00:21:59.680
So thank you once again for uh spending a bit of time with us today and uh sharing some of your insights.
00:22:00.319 --> 00:22:02.240
And uh we wish you all the best.
00:22:02.880 --> 00:22:03.359
Thank you, James.
00:22:03.680 --> 00:22:04.319
It's a great pleasure.
00:22:04.480 --> 00:22:04.720
Thank you.
00:22:04.960 --> 00:22:05.839
The best to you two.
00:22:09.599 --> 00:22:15.119
Well that was really uh interesting and I'm looking forward to dissecting some of what we've just heard there.
00:22:15.200 --> 00:22:15.920
But who's next?
00:22:16.559 --> 00:22:27.440
Well, next we have uh giving us a little bit of a deep dive uh and a company focus uh more uh in uh in the weeds uh is uh Anna Lernanesbit.
00:22:27.599 --> 00:22:37.920
She's CEO of the Climate Collective, uh where she leads an ecosystem of technology innovators and sustainability leaders creating system level solutions.
00:22:38.240 --> 00:22:42.960
Uh she's got almost 20 years of working at the intersection of impact and tech.
00:22:43.359 --> 00:23:00.799
She's held senior leadership positions at Meta, formerly Facebook, uh the World Bank Group, and the Global Environment Facility, always focused on the frontier of tech adoption and market-based based solutions, and also the social impact initiatives globally.
00:23:01.599 --> 00:23:04.160
So um, yeah, what a perfect guest.
00:23:04.640 --> 00:23:06.160
Great, let's hear from her.
00:23:09.440 --> 00:23:10.559
Welcome, Anna.
00:23:10.720 --> 00:23:14.799
Thank you for joining us um at in the Dancing for Systems podcast.
00:23:14.880 --> 00:23:16.960
Um, lovely to have you here.
00:23:17.440 --> 00:23:25.759
We have got a few questions for you, really reflecting your experience at the intersection of tech and AI and sustainability.
00:23:26.079 --> 00:23:33.519
And I'd like to start with a question that's on everyone's lips right now, which is Is AI good for sustainability?
00:23:33.680 --> 00:23:34.559
What are the challenges?
00:23:34.720 --> 00:23:36.400
What are the opportunities?
00:23:37.440 --> 00:23:39.359
Ah, thank you so much for having me.
00:23:39.519 --> 00:23:43.039
I'm really excited to be here and unpack this with you.
00:23:43.200 --> 00:23:57.599
Uh, this is a question I'm getting all the time from organizations that we train at Climate Collective, uh, from friends, from peers who've also worked in sustainability for, I mean, I've been in this space for 20 years.
00:23:57.839 --> 00:24:02.960
It's people are concerned and worried, and there's a lot of uncertainty.
00:24:03.119 --> 00:24:05.839
So very happy to be here and unpack it with you.
00:24:06.079 --> 00:24:11.200
I would say to ground ourselves first, this is gray.
00:24:11.359 --> 00:24:12.079
I'm sorry.
00:24:12.319 --> 00:24:15.440
It's neither or on its own, right?
00:24:16.079 --> 00:24:17.759
AI will be good.
00:24:17.920 --> 00:24:22.240
AI could be really positive for whatever we point it towards.
00:24:22.480 --> 00:24:41.359
If we ask AI to supercharge sustainability teams, decarbonization strategies, the clean energy transition, it will supercharge that, make it faster, better, allow it to scale, simplify financing, all the things that we really desperately need.
00:24:41.680 --> 00:24:50.720
But if we just let AI sort of exist on its own, if we embed AI in the current economic system, it will just reinforce what's already here.
00:24:50.960 --> 00:24:54.559
And as we all know, that's not exactly sustainable.
00:24:54.880 --> 00:25:00.559
Um so, two things I'd like to just add when this comes up.
00:25:01.279 --> 00:25:10.400
AI will only move the needle on sustainability in a meaningful way if a large amount of people use it for emission reductions or protection biodiversity.
00:25:11.039 --> 00:25:22.240
If we're just doing like cat pictures or birthday invitations on Chat GDP, it's not gonna solve staying within planetary boundaries, right?
00:25:22.720 --> 00:25:28.960
And second, and this is probably the more important point, not AI, not all AI is the same.
00:25:29.200 --> 00:25:32.400
Not all AI will have the same positive and negative impact.
00:25:32.720 --> 00:25:41.200
The most valuable contributions to sustainability won't come from consultants or conservation nonprofits using chatbots.
00:25:41.440 --> 00:26:03.680
That will help them and their organizational efficiency, but the big, like really big improvements that we are desperately needing will come from applied AI embedded in industrial systems, optimizing logistics or improving smart grids, finding new protein sources, accelerating carbon removal in assets.
00:26:04.000 --> 00:26:06.799
That's really where the big bang for the back will come.
00:26:07.440 --> 00:26:27.680
That's so interesting and incredibly helpful because effectively what you're saying is we need to really understand the role of AI in accelerating systems transitions that are underway and not view it as something separate or a threat or an opportunity in it in its own right.
00:26:27.920 --> 00:26:37.039
But the exam question, if you will, is you know, how might AI accelerate the energy transition, accelerate the food systems transition, health systems transition?
00:26:37.599 --> 00:26:53.440
On that basis, then how do we get that message out more concretely and cut through all the sort of noise about AI is going to destroy jobs or AI is going to clean my data for me?
00:26:53.599 --> 00:26:57.920
How do we sort of elevate that sort of strategic application of AI?
00:26:58.480 --> 00:27:00.079
That's a great question.
00:27:00.400 --> 00:27:08.160
Um I think currently the big narrative out there is kind of anchored around fear.
00:27:08.400 --> 00:27:15.039
Unless you adopt AI really fast, you're going to be missing out, you're going to lose your job, do you have enough money on your savings account?
00:27:15.440 --> 00:27:24.160
None of that is an inviting narrative, a narrative that opens up for exploration or joy or curiosity.
00:27:24.799 --> 00:27:27.279
Nor is it particularly like regenerative.
00:27:27.759 --> 00:27:38.079
So I think first we need to, well, my theory of change as an organization is kind of a mini-universe of that question you have.
00:27:38.319 --> 00:27:57.680
I think one is uh educating people, putting, uh helping people understand what AI is, what it isn't, and when it could augment or complement humans and human expertise, and when it really shouldn't, because the risks are too high.
00:27:58.319 --> 00:28:00.000
I call that AI literacy.
00:28:00.079 --> 00:28:07.680
So, like raising AI literacy amongst the general public, amongst civil society, amongst governments.
00:28:07.839 --> 00:28:13.039
I mean, how are how could governments regulate this technology if they don't quite understand it?
00:28:13.200 --> 00:28:14.319
I mean, that's right.
00:28:14.559 --> 00:28:22.319
So I think lifting understanding and talking about it is one really critical component.
00:28:22.559 --> 00:28:26.160
The other piece is to invite for transparency.
00:28:26.240 --> 00:28:34.480
I mean, we do trainings for conservation nonprofits, uh, net zero alliances, corporate sustainability teams.
00:28:34.720 --> 00:28:40.160
And I can't even tell them with real numbers exactly what their AI footprint is.
00:28:40.559 --> 00:28:49.519
So that creates this kind of uncertainty where people are like a bit worried but pushed to hurry up and quickly about this so they're not missing out.
00:28:49.680 --> 00:29:09.680
And as humans, we always always find a way to game the system where we always start optimizing for like who's using the most tokens, who's burning the biggest AI, who has the biggest AI bills, and and like none of those indicators are KPIs that we should really measure ourselves as a society.
00:29:09.920 --> 00:29:17.039
So yeah, I think agency, like building agency amongst people, civil society is really the key here.
00:29:17.680 --> 00:29:37.039
And Anna, when you talk about the narrative and the fears and the uncertainties, certainly in our space where we're concerned about sustainability, the first thing that comes up is AI's energy footprint, and then AI's water footprint, and then of course all of the ethical issues around how AI might be used.
00:29:37.200 --> 00:29:48.480
And I wonder how you see that uh also in contrast with the potential upside, even on energy or or on water or on many other solutions.
00:29:48.720 --> 00:29:55.839
Um is there a kind of net positive uh narrative as well that we're we're failing to get across?
00:29:57.279 --> 00:29:57.920
Yes.
00:29:58.160 --> 00:30:03.119
Um there are very little research with actual data on this uh yet.
00:30:03.279 --> 00:30:17.279
Um it's a new topic, but you know, that there's just not enough information out there, so we can't quite measure the exact energy footprint or the exact water usage of certain um facilities.
00:30:17.519 --> 00:30:27.839
I think that's silly because think about when you go to a grocery store, you can often find out which fishing boat a specific fish was caught on.
00:30:28.079 --> 00:30:30.319
You can even find out what were their fishing practices.
00:30:30.559 --> 00:30:35.119
Like we've figured out a way to trace a lot of information through value chains.
00:30:35.359 --> 00:30:38.079
Why can't I know where my AI was processed?
00:30:38.240 --> 00:30:43.920
If I'm using Claude, why can't I know what data center those prompts were routed via?
00:30:44.160 --> 00:30:46.079
I mean, technically it's totally possible.
00:30:46.240 --> 00:30:50.240
So it's not just a matter of like, I can't access that information.
00:30:50.480 --> 00:30:50.799
Yeah.
00:30:51.519 --> 00:30:58.400
It makes a big difference, I think, if your AI was processed in a data center operating on a dirty grid, like I'm from northern Sweden.
00:30:58.559 --> 00:31:06.799
I'd feel really good if my prompts were routed through that data center in northern Sweden that has almost entirely renewable energy.
00:31:06.960 --> 00:31:12.400
But I would feel really uncomfortable if it's going through a data center that's not powered by natural gas.
00:31:12.559 --> 00:31:20.079
Because then kind of my impact as a sustainability organization just feels kind of it just doesn't feel it feels misaligned.
00:31:20.559 --> 00:31:20.880
Yeah.
00:31:21.359 --> 00:31:25.839
And maybe also to remind people that we're at such an early stage with AI.
00:31:25.920 --> 00:31:30.720
And if we looked at the efficiency of the internal combustion engine when it just came out, it would be terrible.
00:31:30.960 --> 00:31:32.319
Same with solar panels.
00:31:32.559 --> 00:31:37.920
So, you know, I expect a technology innovation uh curve that will happen there as well.
00:31:38.240 --> 00:31:45.359
But what do you say to companies now who are grappling to internalize, to adopt, to apply AI?
00:31:45.599 --> 00:31:54.480
What you hear so much about is is maybe speed over wisdom, I would say, just trying to do it faster and as quickly as possible.
00:31:54.640 --> 00:32:13.599
Um and then I've also heard talk among businesses about the difference between using AI as a tool and becoming AI native, which I think speaks more to what you were saying is oh, we're actually using it to change our products and our services and our processes, not just to become more efficient.
00:32:14.640 --> 00:32:15.279
Yeah.
00:32:17.359 --> 00:32:22.960
All of these points are quite poignant right now, and I think you're real you're right.
00:32:23.119 --> 00:32:39.759
We're so early that even someone like me, who I spend all my waking hours thinking and reading and doing sustainability and AI, and still there's so much that happens every day that I'm struggling to keep up with.
00:32:40.319 --> 00:33:03.599
Um I think for corporations, when we work with sustainability teams, corporate sustainability teams, they are struggling, they're struggling from kind of a large responsibility and burden of reporting requirements often and kind of communicating why a green option makes sense for the for the for the corporate narrative.
00:33:03.759 --> 00:33:05.680
And I think AI can be really helpful here.
00:33:05.839 --> 00:33:12.000
AI really can act as a kind of bridge that ties together different investments and different strategies across a company.
00:33:12.240 --> 00:33:42.400
AI can also help them produce certain things, let's say like sustainability reporting, using AI to enhance that can make a lot of the kind of burdensome, heavy data processing slim down and instead allow them to focus on kind of the more where their human judgment is more important and kind of our where connections and and and and that humanity is more valued.
00:33:42.559 --> 00:33:52.319
So I think there's there are really great examples already emerging of how corporates can embed AI in their own sustainability strategies and programs.
00:33:53.039 --> 00:33:59.759
I think a responsible AI approach would be to also support regulation.
00:34:00.160 --> 00:34:07.519
I think corporates have a really important role these days in like underlining the need for regulation.
00:34:08.239 --> 00:34:13.280
We missed the boat on regulating social media and see where that got us.
00:34:13.440 --> 00:34:18.880
Um I think self-regulation has also not worked well for climate or nature contributions.
00:34:18.960 --> 00:34:22.079
So I think we have an opportunity here to do this better.
00:34:22.480 --> 00:34:28.639
Um and last one, I would I would tell corporates too, you know, look at your suite of partners.
00:34:28.800 --> 00:34:29.840
Who do you partner with?
00:34:30.000 --> 00:34:31.920
Do they know how to use AI well?
00:34:32.079 --> 00:34:37.199
Could you support them as you're learning how to do this responsibly and sustainably?
00:34:37.280 --> 00:34:39.760
Could you support your partners in doing the same?
00:34:40.079 --> 00:34:45.119
For example, we partner with Clarnow, which is a large fintech company in Sweden.
00:34:45.440 --> 00:34:49.280
They're a long-standing backer of climate and nature entrepreneurs.
00:34:49.519 --> 00:34:55.039
The first thing they did was to ask their portfolio companies if they have the resources to use AI well.
00:34:55.360 --> 00:35:00.960
When they realized they didn't, they then came to us and asked if we could help build AI literacy in their companies.
00:35:01.440 --> 00:35:13.760
And just from some trainings and interventions, we're already seeing this massive boost in their in these organizations' output, saving costs, reaching new fundraising targets, and having an impact.
00:35:14.800 --> 00:35:16.480
And that wasn't possible before.
00:35:16.559 --> 00:35:27.840
So I think corporations have a really new level of kind of responsibility looking inwards, but also responsibility looking outwards for their partners and and value chain.
00:35:30.079 --> 00:35:34.960
Anna, just um our last question to you, if we can.
00:35:35.280 --> 00:35:48.239
Um that was a really lovely and different insight into AI, and particularly how AI can be framed as potentially accelerating these systems transitions.
00:35:48.719 --> 00:35:57.519
Would love I guess your advice for organizations that are committed to their role in systems transitions.
00:35:57.679 --> 00:36:07.840
What's your kind of top one, two pieces of advice for change makers in businesses that want to wrap their arms around AI and use it for systems transformation?
00:36:08.079 --> 00:36:11.519
And what's your advice for us as individuals?
00:36:15.039 --> 00:36:20.320
I think we have a lot to learn from how corporations approached the SDGs.
00:36:20.639 --> 00:36:32.239
One of my jobs at when I worked at Meta was to build out Meta's response to the sustainable development goals with the goal of sort of accelerating progress on the SDGs.
00:36:33.440 --> 00:36:35.119
We then looked inwards first.
00:36:35.199 --> 00:36:38.719
We thought, like, okay, what are we calling superpowers because it's a tech company.
00:36:38.800 --> 00:36:40.880
So what are Meta's superpowers?
00:36:41.039 --> 00:36:45.440
Uh what does what could Meta bring to the table that's unique for them?
00:36:45.679 --> 00:36:59.599
And in our case, when we looked at then trying to match that with the different SDGs that were priorities, we found, well, we have unique data sets, with unique uh AI capabilities, engineers and compute.
00:36:59.840 --> 00:37:17.039
So then we kind of anchored our interventions and our programs and our partnership outreach around those things that we knew we were already doing really well and could do more of in sort of a for good social impact contribution.
00:37:17.280 --> 00:37:41.199
So I would say for at least for corporations, think about what you're good at and try to anchor your sustainable AI strategies, interventions, principles around that, because that will just lead to a much more long-term sustainable contribution versus you know doing something kind of like that's not core to your business.
00:37:41.599 --> 00:37:45.519
I think that also holds like the world is a tough place today.
00:37:45.760 --> 00:37:55.280
So staying close to kind of your lane, what you're really strong at, I think will ensure that these interventions are more longer-term uh sustainable.
00:37:55.760 --> 00:37:56.000
Great.
00:37:56.320 --> 00:38:12.639
And then for individuals, we did this crazy thing in February when I called a bunch of women I worked with, 40 women actually, that I worked with throughout my career, in all with leadership roles in climate.
00:38:12.800 --> 00:38:15.760
And I all asked them, like, Do you do you feel good about AI?
00:38:15.840 --> 00:38:17.440
Do you know what it is?
00:38:17.920 --> 00:38:19.360
Do you understand it?
00:38:19.679 --> 00:38:24.320
Are you using it for your own sort of efficiency at work?
00:38:24.639 --> 00:38:32.639
And they're all, you know, generally all of them just said, I have no idea what this is, it scares me, I'm really concerned about the environmental impact.
00:38:32.719 --> 00:38:35.920
I'm sitting on the sidelines feeling really stressed.
00:38:36.159 --> 00:38:40.559
So we just came together as a, we call it the women in climate community.
00:38:40.800 --> 00:38:44.239
So over the last 16 weeks, we've just met once a week.
00:38:44.559 --> 00:38:55.840
And my team brings some skills, some kind of basic tutorials, and we experiment with different technologies and we talk about what they're good at, what they're not so good at.
00:38:56.079 --> 00:39:06.400
Um, and then we have a WhatsApp group where we share articles that we read that upsets us or that gets us happy and excited, and we discuss them in community.
00:39:06.639 --> 00:39:12.800
And I think the I didn't know that I needed a women's group, but I really did.
00:39:12.960 --> 00:39:31.599
Um, and I think joining together, even if it's a group of colleagues, maybe it's a handful of neighbors or it's friends, just coming together and kind of unpack this topic that's moving so fast that is has the potential to really like rewire society, economies, communities.
00:39:32.559 --> 00:39:36.000
I think this is a time where we should kind of do that together.
00:39:36.400 --> 00:39:49.679
I love that because what you're essentially saying is let's harness the energy in the system and dance with it to flow that energy into the direction we want it to flow and let's realise we all have agency.
00:39:49.840 --> 00:39:51.760
There's something we can all do.
00:39:52.239 --> 00:39:54.480
Thank you so much, Anna, for joining us.
00:39:54.639 --> 00:39:56.559
It's been lovely to chat with you.
00:39:56.719 --> 00:39:58.079
Um thank you.
00:39:58.480 --> 00:39:59.280
Thanks, Anna.
00:39:59.599 --> 00:40:00.239
Thank you.
00:40:04.559 --> 00:40:07.599
That was also really, really interesting.
00:40:07.840 --> 00:40:09.440
What struck you, Wayne?
00:40:10.159 --> 00:40:19.519
Well, I think what what came across for me is that uh you know, whatever you point AI at, it will accelerate.
00:40:19.760 --> 00:40:20.079
Right?
00:40:20.159 --> 00:40:21.599
That that's really what it's doing.
00:40:21.679 --> 00:40:23.280
It's like a supercharger.
00:40:23.760 --> 00:40:34.480
Um and the question that emerged from these uh discussions was um, you know, it doesn't help to accelerate in a broken system.
00:40:35.199 --> 00:40:40.719
So it doesn't uh alleviate the necessity to actually change the system itself.
00:40:41.119 --> 00:40:46.639
And if we're able to do that, then AI could probably uh accelerate the solutions.
00:40:47.039 --> 00:40:52.559
So the question we always have to ask when we talking about AI is AI for what?
00:40:52.800 --> 00:40:54.480
What's it being pointed at?
00:40:54.880 --> 00:40:55.519
How about you?
00:40:55.679 --> 00:40:56.880
What what intrigued you?
00:40:57.440 --> 00:40:59.280
I think that's really interesting actually.
00:40:59.360 --> 00:41:01.760
Uh AI needs the right architecture.
00:41:02.000 --> 00:41:16.880
Um But I think what struck for me was uh social movements uh when it comes to driving that accountability for AI and particularly data centres, might do what policy isn't yet doing.
00:41:17.039 --> 00:41:25.360
I was really struck by the fact that people can see the data centres, they are lowering water tables in some parts of the US.
00:41:25.760 --> 00:41:38.159
Joel was talking about the fact that uh this issue has stepped over, you know, traditional political lines, and communities are finding their voice, and I think that is really interesting.
00:41:38.480 --> 00:41:45.760
So could communities, could society become the accountability mechanism for AI and tech?
00:41:46.239 --> 00:41:50.800
Yeah, maybe a lever we we hadn't necessarily anticipated.
00:41:51.360 --> 00:41:53.440
So um could be very powerful.
00:41:53.519 --> 00:41:57.039
Well, what's the dance vibe you get for for for the sector, Sally?
00:41:57.440 --> 00:42:05.920
So I've thought about this a little bit, and I'm gonna go for the pase dobla because it is a dance of incredible power.
00:42:06.079 --> 00:42:10.639
You've got the bull charging at the matador, but who's gonna win?
00:42:10.800 --> 00:42:25.039
Because they're both really, really powerful, and it's dynamic, it's fast-paced, and so when I think about tech, there's a big pass dobla going on right now between the hyperscalers, community, society, policy, and where will that cape go?
00:42:25.360 --> 00:42:27.360
Yeah, and there are red flags there, right?
00:42:27.679 --> 00:42:29.760
Absolutely, as there are in tech.
00:42:30.159 --> 00:42:38.400
Well, uh that that's it for this episode, and uh we look forward to you joining us next time uh when we dive into another sector.
00:42:38.960 --> 00:42:39.840
Thank you.
00:00:09.199 --> 00:00:12.800
Welcome to the tech episode of Dancing with Systems.
00:00:12.960 --> 00:00:16.239
What a timely discussion to have at this moment in time.
00:00:16.640 --> 00:00:21.600
We know that tech is both the enabler of every other systems transition, more of that soon.
00:00:21.920 --> 00:00:26.079
But it's also a sector in its own right with a rapidly growing footprint.
00:00:26.399 --> 00:00:47.920
And right now, those two, a force for good, a force with its own impacts, are now really intentional and the AI boom is a sort of epicenter of that defining dynamic, with data centers becoming one of the world's fastest growing consumers of water and of energy, and often concentrated in communities where people don't want them.
00:00:48.479 --> 00:00:52.320
So the narrative is turning in a way that I don't think we've seen before.
00:00:52.399 --> 00:00:56.719
And really interesting, the narrative is turning in a way that's hopping over politics.
00:00:57.280 --> 00:01:00.320
People are getting fearful of AI.
00:01:00.719 --> 00:01:06.560
And so what a great opportunity to talk about tech and systems transitions.
00:01:06.959 --> 00:01:13.760
Yeah, well, uh this is a fantastic episode and so central to everything that we're living with.
00:01:14.079 --> 00:01:18.159
Let me put a few numbers to some of those uh opening remarks that you made.
00:01:18.400 --> 00:01:29.200
So uh the the five big tech companies are are now spending more than the oil and gas industry on capital investment.
00:01:29.599 --> 00:01:34.319
And uh uh so that's 400 billion uh in 2025.
00:01:34.719 --> 00:01:40.000
Uh that's expected to jump up to 75% this year in 2026.
00:01:40.239 --> 00:01:41.840
So wow.
00:01:42.159 --> 00:01:44.319
Yeah, it's being supercharged, right?
00:01:44.640 --> 00:01:58.159
Uh and of course the the big uh uh link there is to data centers, and and we've seen some uh stepping back from commitments by those big tech companies, we've seen some concerns around emissions.
00:01:58.400 --> 00:02:01.439
Uh we do need to keep it in perspective, right?
00:02:01.519 --> 00:02:11.360
Uh data centers account for around one and a half percent of global uh consumption of electricity, so you know, much less than some other sectors.
00:02:11.680 --> 00:02:16.639
Um but it has been growing at around 12% since 2017.
00:02:17.280 --> 00:02:19.280
So, you know, it is ramping up.
00:02:19.360 --> 00:02:23.360
Uh that's four times faster than electricity consumption in other areas.
00:02:23.599 --> 00:02:27.199
So, you know, it is something that we have to be concerned about.
00:02:27.520 --> 00:02:37.520
Um but there is an open question uh, you know, could AI actually deliver some of the emission reductions as a result of all the benefits it brings?
00:02:37.599 --> 00:02:46.000
And at the moment there's an estimate by the International Energy Agency of uh reductions of perhaps 5% by 2035.
00:02:46.319 --> 00:02:52.319
Um so you know, we're we're I I guess that's something we might explore in the in this episode as well.
00:02:52.639 --> 00:02:53.680
Um so yeah.
00:02:54.560 --> 00:03:03.439
So lots of big numbers there, and also a story of exponential uh growth as well in terms of that capital investment number.
00:03:03.680 --> 00:03:06.319
Um so we're talking about tech.
00:03:06.560 --> 00:03:09.120
What is it, Wayne, that you're interested to hear about?
00:03:09.840 --> 00:03:19.199
Yeah, well, I think it makes sense that we're gonna focus in uh really on the tech of the day, which is uh which is AI and and data and so on.
00:03:19.599 --> 00:03:23.680
And what we know is that this is now the great accelerator.
00:03:24.080 --> 00:03:31.360
But the question really is is it going to accelerate the transition that we're looking for in all these other sectors?
00:03:31.680 --> 00:03:35.439
Or is it going to accelerate the decline, the problem?
00:03:35.520 --> 00:03:39.039
Uh so that's a big question, something I'd like to hear more about.
00:03:39.439 --> 00:03:45.599
Um and you know, we do hear a lot, especially on social media, about the impact of data centers.
00:03:46.000 --> 00:03:57.360
And I'm just wondering whether we're getting caught in a little bit of hype here and whether it really is as bad as social media seems to suggest that it is.
00:03:57.520 --> 00:04:00.800
So I'd I'd love some perspective from our guests on that.
00:04:01.120 --> 00:04:03.280
Um so uh what about you?
00:04:03.520 --> 00:04:04.800
What are you hoping to hear?
00:04:05.360 --> 00:04:11.280
I'm really curious about how does accountability for AI work?
00:04:11.840 --> 00:04:14.400
Is it down to the governments to develop policies?
00:04:14.639 --> 00:04:20.720
Because at the moment the policy framework is really lagging the exponential growth of particularly AI.
00:04:21.040 --> 00:04:22.560
Is it the hyperscalers?
00:04:22.959 --> 00:04:24.720
And intriguingly, is it us?
00:04:25.199 --> 00:04:33.439
Because we're beginning to understand that AI has a positive impact, but also there is a lot of fear about job losses, etc.
00:04:34.079 --> 00:04:40.480
So, where does accountability lie in order to ensure that AI is on balance of force for good?
00:04:40.720 --> 00:04:42.000
That's what I'm interested in.
00:04:42.399 --> 00:04:47.199
Yeah, and I guess that's gonna link also to the social elements, right, of AI.
00:04:47.439 --> 00:04:50.879
So um well, let's jump in and uh who've we got first.
00:04:51.519 --> 00:04:52.639
We've got Joel Makkawa.
00:04:53.199 --> 00:04:54.879
Okay, do you want to give us an introduction?
00:04:55.040 --> 00:04:55.519
Who is Joel?
00:04:55.920 --> 00:04:56.319
I can.
00:04:56.720 --> 00:05:04.560
Joel um is a wonderful person, but he's also the chair and co-founder of Trellis Group, who were formerly Green Biz Group.
00:05:04.720 --> 00:05:07.199
I think people are still getting used to that name change.
00:05:07.519 --> 00:05:15.600
Um Green Biz, Trellis, as they are now, a leading media and events platform at the intersection of business sustainability and clean technology.
00:05:16.000 --> 00:05:20.000
He's also the strategy director of the Music Sustainability Alliance.
00:05:20.480 --> 00:05:21.120
I need to go on.
00:05:22.720 --> 00:05:30.639
Um he's a former nationally syndicated columnist, he is a journalist by training with over 35 years covering sustainable business.
00:05:30.959 --> 00:05:35.199
He was one of the first journalists I remember talking about sustainability and business.
00:05:35.439 --> 00:05:40.720
And he's the author of more of a dozen books and lead author of the annual State of the Green Business Report.
00:05:41.120 --> 00:05:50.399
The Associated Press has called him the guru of green business practices, and he is inducted into the International Society of Sustainable Professionals Hall of Fame.
00:05:50.959 --> 00:05:53.839
The perfect person to set the scene on tech for us.
00:05:55.279 --> 00:05:55.680
Great.
00:05:56.079 --> 00:05:57.120
Let's hear from him now.
00:05:57.680 --> 00:06:01.519
Welcome, Joel, to this uh episode of Dancing with Systems.
00:06:02.079 --> 00:06:15.839
It's such a pleasure to have you uh as one of our veterans who's also been uh focused on tech for a long time and taken a special interest uh throughout your journalistic career, especially.
00:06:16.160 --> 00:06:23.519
And so uh we'd like to kick off by uh asking you you know, there are two ways to look at tech.
00:06:23.680 --> 00:06:30.240
One is that tech is an enabler for uh transitions in other sectors, and so we'd like to start with that.
00:06:30.480 --> 00:06:34.399
But then of course there is the tech sector itself which has its own impact.
00:06:35.040 --> 00:06:49.199
But maybe you could start by giving us your view of of how tech can be and maybe is being used to help the other sectors uh transition to net zero, nature positive, and so on.
00:06:50.160 --> 00:06:59.120
Well, you're right in that uh tech is both an enabler of of tr uh other transitions and also a sector with its own footprint.
00:06:59.279 --> 00:07:02.959
And right now, those two faces are in tension.
00:07:03.120 --> 00:07:13.759
Um the same data center build-out that's powering AI, for example, is becoming one of the largest sources of demands on energy, water, land, and and other things.
00:07:14.000 --> 00:07:29.360
So we're at this really interesting moment, and and I know that there's a debate is is, I think putting it lightly, around, you know, what is the purpose, the value of AI to society, uh, and and for whom?
00:07:29.519 --> 00:07:40.319
And so right now we're I think we're having that discussion in in a significant way, um, and and understanding that it it's not, you know, there's there's both sides.
00:07:40.560 --> 00:08:10.720
So uh, you know, simply on the enabling side, when you look at pretty much all of the challenges that we as humanity are facing: energy, water, materials, toxicity, the carbon cycle, hydraulic cycle, uh, pretty much everything, uh the the energy, electricity grid, these are among the most incredibly complex systems that we know of and we barely understand.
00:08:11.120 --> 00:08:22.720
Um in addition, these things are are are are spewing data uh 24-7, uh, whether it's about plants or about you know energy plants or or actually green plants.
00:08:23.040 --> 00:08:43.200
And um and and so making sense of that, understanding it, and being able to to optimize and predict and monitor those things is is really one of uh technology's uh great uh advances here uh and has an incredible promise.
00:08:43.360 --> 00:08:53.600
But I think the irony here is that we can now measure emissions and planetary impact far better than we can reduce them.
00:08:54.559 --> 00:08:55.200
Yeah.
00:08:55.679 --> 00:08:57.840
Yeah, and I guess that's the question.
00:08:58.240 --> 00:09:08.480
If you think about those different sectors and transitions that we're trying to enable, where do you see that technology is already having that that impact?
00:09:08.879 --> 00:09:15.360
We're already seeing it scaling, because it's not only about the data, it's also about new technologies, literally.
00:09:15.679 --> 00:09:20.559
And and where do you see where it's kind of lagging behind and there's still a gap?
00:09:21.440 --> 00:09:25.759
Well, first of all, we should acknowledge that technology is a very, very broad term.
00:09:25.919 --> 00:09:26.080
Yeah.
00:09:26.240 --> 00:09:39.519
Uh it can it we you know, we're these days we often use it to re to refer to to digital technologies, um, from from phones and computers to AI systems and and and and on and on.
00:09:39.759 --> 00:09:42.639
Um, but technology goes much broader than that.
00:09:42.720 --> 00:10:07.919
So I I'm thinking I'm assuming we're gonna be you know sticking to the you know the more recent uh uh way we talk about technology, which is uh uh, and I think this is where we're seeing the the positive impact so far is is in digital measurement and transparency, uh, infrastructure, sensors, satellites, supply chain traceability, uh grid optimization.
00:10:08.080 --> 00:10:15.279
Um these are things that don't necessarily decarbonize directly, but but they make accountability possible.
00:10:15.519 --> 00:10:22.000
And and and that's uh a significant change from the past where we didn't really understand what was going on.
00:10:22.320 --> 00:10:31.279
Um uh and and that's and and that really is a precondition for you know pretty much all of the other changes we we we want to make and need to make.
00:10:31.519 --> 00:10:36.559
Um so again, as I said before, we know how to measure things and monitor things.
00:10:36.639 --> 00:10:40.639
We just don't know how to necessarily how to change them uh for the better.
00:10:42.159 --> 00:10:48.000
And I guess that then leads us to the transition of the tech sector itself.
00:10:48.240 --> 00:11:02.240
So clearly we know that tech is already enabling us to uh map deforestation, it's giving us access to early warning systems for farmers, it's giving us access to data on air pollution incidents.
00:11:02.399 --> 00:11:08.320
So tech as an enabler of these other systems transitions, you know, as we've been discussing, is clear.
00:11:08.480 --> 00:11:15.679
But then if we look at the impact of the tech sector itself, um where where do you think we are, Joel?
00:11:15.759 --> 00:11:21.120
And and what are the sort of the big challenges that we need to overcome?
00:11:21.360 --> 00:11:24.080
Does that take us straight back to a data center?
00:11:24.720 --> 00:11:36.480
Well, inevitably it does because so much of what we're talking about is either happening in the cloud, which originates in a data center, or in AI, which of course is a data center as well.
00:11:36.559 --> 00:11:57.919
And and and and largely as a result of data centers and and all the devices that flow into and out of them, is that the sector's own transition is going backwards on the metrics that matter most because AI, the compute demand is outpacing the efficiency and clean energy gains.
00:11:58.320 --> 00:12:12.240
Um, and so uh the biggest tech companies, the the Googles and the Metas and ChatGPT and all the others are open AI, uh, their emissions are actually going up.
00:12:12.639 --> 00:12:17.919
And at a time when they're they're they're they had committed to making significant reductions.
00:12:18.480 --> 00:12:21.519
Um and so that's a challenge.
00:12:21.600 --> 00:12:24.799
And you know, we the good news is we have a lot of those solutions.
00:12:24.960 --> 00:12:35.200
We know how to uh to drive data centers uh or create data centers that are energy and water and possibly positive.
00:12:35.360 --> 00:12:38.320
Uh, we're seeing some advancement advances there.
00:12:38.799 --> 00:12:47.759
Um and and you know, the question is, you know, how do these things not just exist uh on the planet, but also exist within communities?
00:12:47.919 --> 00:12:56.480
And that's one of the challenges because uh people are afraid uh of technology in a way that I haven't seen in in a long, long time.
00:12:56.799 --> 00:13:10.080
People weren't afraid of PCs, they weren't afraid of smartphones, they weren't afraid of the internet, although all of those things had chat brought challenges, but not the kinds of fears that we're seeing now with data centers and AI.
00:13:10.320 --> 00:13:18.799
And so uh, you know, the societal impacts uh and and the backlash that we're now seeing around the world.
00:13:18.960 --> 00:13:31.919
Um there's a stat uh here in the United States in a Gallup poll that said that uh more people would prefer to live near a nuclear power plant than in a data center.
00:13:32.159 --> 00:13:33.600
I mean, that's astounding.
00:13:34.399 --> 00:13:48.320
Even if it's not exactly true, even if the people are exaggerating, uh just the thought of that or the fact that it might even be on a par with uh with a nuclear power plant is just an amazing uh uh development.
00:13:48.480 --> 00:14:13.919
And so uh we need to be addressing uh and a lot of those things are are both environmental fears, fears of uh these data centers uh uh uh using too much energy and water, but also the fears that the technology itself brings, terms of job loss, uh changes that people aren't necessarily looking forward to, uh child pornography, so many challenges, uh cybercrime and so on.
00:14:14.159 --> 00:14:25.759
So uh, you know, if you if you include go beyond the planetary boundaries and look at the societal boundaries, if you will, uh these these are uh have some significant challenges.
00:14:26.399 --> 00:14:40.159
And so, Joel, just before I am back to Wayne, it seems to me in a way that the hyperscalers, the big tech players, they've sort of lost control of the narrative a little bit here, haven't they?
00:14:40.559 --> 00:14:53.120
And you know, as a journalist, as an award-winning journalist with lots of experience in this field, what's your sense of how might they contribute to a different narrative?
00:14:53.279 --> 00:15:01.440
Because it seems to me, unless we can shift the narrative, this kind of fear is just going to keep growing.
00:15:02.320 --> 00:15:04.399
Well, that's a really tall order, Sally.
00:15:04.559 --> 00:15:09.759
I think uh the technology in and of itself isn't going to shift the narrative.
00:15:10.000 --> 00:15:44.240
What's going to shift the narrative is are are the data center providers, the AI companies, uh the hyperscalers uh coming in and showing the benefits uh and and working closely with communities to help them understand, uh first of all, address their concerns and then understand how these can be helpful in optimizing systems, in creating new opportunities, uh in uh uh reducing the kinds of uh negative impacts on the planet that we we know all too well.
00:15:44.559 --> 00:16:01.759
Um, but those right now that that these companies have been far more enamored with their own technology than than trying to make that make sense to the people who are impacted most, and those right now are in in local communities.
00:16:02.080 --> 00:16:17.600
So uh, you know, at some point there will be, as I said before, energy and water positive data centers that will actually be welcome in the communities because they'll be providing some uh environmental and social goods.
00:16:17.919 --> 00:16:23.759
Um and the, as I said, those technologies exist, but we're not that they're a ways off.
00:16:23.840 --> 00:16:52.399
And and certainly at the scale, scope, and speed of the data center build-out, particularly in China and the United States, but around the world, we're not going to be seeing those innovations anytime soon simply because most data centers are you know only maybe incrementally improved from the previous one, and very few data center builders have been fired for making the next data center look pretty much like the last one.
00:16:52.559 --> 00:17:04.400
And so getting the uh some of the technologies that have been around for years, the uh waterless cooling technologies, for example, uh are still not as widespread as they could be.
00:17:04.559 --> 00:17:26.799
And we're just now starting to see the energy technologies develop that can uh, you know, there's this uh trend in in in, at least in the States, maybe globally, called bring your own power, where in order for a data center to build in a community, they have to provide their own electricity needs on-site generation, whatever form that is.
00:17:27.200 --> 00:17:43.359
So until those become widespread and people start to see, oh, these are these are actually uh you know great benefits to our communities, then I think they'll start to you know look at the tech what the technology actually does.
00:17:43.680 --> 00:17:50.160
Um but until until then, uh it's gonna be uh I think a really bumpy ride for the tech sector.
00:17:51.039 --> 00:18:01.599
Yeah, as you're talking, Joel, it it reminds me of uh of Kenneth Bolding's uh characterization of you know the cowboy economy versus the spaceship economy.
00:18:01.680 --> 00:18:13.200
And it does feel at the moment, especially with uh big tech and AI, you know, it's it's uh it's a wild west, you know, and uh where there's seemingly no frontiers, at least that's how they're behaving.
00:18:13.599 --> 00:18:18.240
Um but also uh as you're talking, uh two two other thoughts came to mind.
00:18:18.559 --> 00:18:21.440
One is are we suffering from technology lock-in?
00:18:22.000 --> 00:18:31.839
It feels like quite a new technology, but but the kinds of setups that we have for for data centers, uh, as you say, we're kind of replicating them, but there are better ways to do them.
00:18:31.920 --> 00:18:39.200
It's just that you know, now that we've we've already uh invested in this kind of setup, it seems hard to change.
00:18:39.839 --> 00:18:46.799
Um and and and my my sort of f follow-up and final question really is so where does that leave us?
00:18:46.960 --> 00:18:57.440
What are the levers for changing this behavior, for getting to the kinds of uh efficiencies that we believe are probably out there?
00:18:57.599 --> 00:18:58.960
It's very early days, right?
00:18:59.039 --> 00:19:04.480
So this should go down an efficiency curve as well, uh, and and an innovation curve.
00:19:04.640 --> 00:19:07.039
So um where do you see the main levers?
00:19:08.000 --> 00:19:15.440
Well, in in simply put, we need to set a high bar and a and uh and a high floor.
00:19:15.680 --> 00:19:27.440
So we need to uh in communities need to uh really start pressing uh data center builders in ways that they haven't yet.
00:19:27.599 --> 00:19:37.839
There as much protest as there has been, as much of a big of a backlash as there been has been, it really hasn't had gotten traction as a national movement.
00:19:38.000 --> 00:19:49.119
I think uh we're gonna see that start to shift uh here in the United States, for example, it's gonna become uh a big issue in the in the uh 2026 midterm elections.
00:19:49.279 --> 00:19:50.880
We're seeing that already.
00:19:51.039 --> 00:20:00.480
Um, and and as it becomes a sort of a cause celeb for um and it crosses over uh left and right politics, left, right, and center.
00:20:00.559 --> 00:20:04.640
Uh it's it's amazingly bi or tripartisan here.
00:20:04.880 --> 00:20:12.559
Um so so we need uh in order to do that, we need some kinds of uh uh regulations or legislation.
00:20:12.720 --> 00:20:14.960
The data center companies want that.
00:20:15.119 --> 00:20:16.720
Uh the builders want that.
00:20:16.880 --> 00:20:26.480
I think the entire ecosystem says we need to, you know, some kinds of of guidelines, some kinds of guardrails of what we can build.
00:20:26.640 --> 00:20:29.359
Um and communities need to really press on that.
00:20:29.440 --> 00:20:35.680
And so it's it's really a combination of of laws and social pressure that's going to change things.
00:20:35.839 --> 00:20:41.839
As I said, the technology's been around, and but it there's haven't been any incentives for using it.
00:20:42.000 --> 00:20:48.240
Um and the technology, you know, it continues to improve like all you know, technologies do.
00:20:48.400 --> 00:20:57.680
Uh, the energy technologies, the waterless cooling technologies, all of those things uh are really uh year over year beginning to change.
00:20:57.759 --> 00:21:06.960
So it's inevitably, and and as it is with so many of these challenges, it's it's a number of forces working together over time.
00:21:07.839 --> 00:21:19.440
But there is a movement that is really shaping up, and I think it's going to be a really politically powerful movement that's gonna transcend uh the political parties as we've known them.
00:21:19.519 --> 00:21:32.400
And it'll be really, really interesting to watch uh how frustrated and angry people are uh and scared uh about these technologies and what it will take to change them.
00:21:32.559 --> 00:21:40.400
I I wish I had an answer of what that magic wand would be, but um I don't, and I don't think anyone else does yet at this point.
00:21:40.880 --> 00:21:41.599
Thanks, Joel.
00:21:41.680 --> 00:21:53.440
Well, I I'm glad also that you you ended on on social movements because this is such a critical uh lever in in systems change and and maybe one that we uh we don't always pay enough attention to.
00:21:53.599 --> 00:21:59.680
So thank you once again for uh spending a bit of time with us today and uh sharing some of your insights.
00:22:00.319 --> 00:22:02.240
And uh we wish you all the best.
00:22:02.880 --> 00:22:03.359
Thank you, James.
00:22:03.680 --> 00:22:04.319
It's a great pleasure.
00:22:04.480 --> 00:22:04.720
Thank you.
00:22:04.960 --> 00:22:05.839
The best to you two.
00:22:09.599 --> 00:22:15.119
Well that was really uh interesting and I'm looking forward to dissecting some of what we've just heard there.
00:22:15.200 --> 00:22:15.920
But who's next?
00:22:16.559 --> 00:22:27.440
Well, next we have uh giving us a little bit of a deep dive uh and a company focus uh more uh in uh in the weeds uh is uh Anna Lernanesbit.
00:22:27.599 --> 00:22:37.920
She's CEO of the Climate Collective, uh where she leads an ecosystem of technology innovators and sustainability leaders creating system level solutions.
00:22:38.240 --> 00:22:42.960
Uh she's got almost 20 years of working at the intersection of impact and tech.
00:22:43.359 --> 00:23:00.799
She's held senior leadership positions at Meta, formerly Facebook, uh the World Bank Group, and the Global Environment Facility, always focused on the frontier of tech adoption and market-based based solutions, and also the social impact initiatives globally.
00:23:01.599 --> 00:23:04.160
So um, yeah, what a perfect guest.
00:23:04.640 --> 00:23:06.160
Great, let's hear from her.
00:23:09.440 --> 00:23:10.559
Welcome, Anna.
00:23:10.720 --> 00:23:14.799
Thank you for joining us um at in the Dancing for Systems podcast.
00:23:14.880 --> 00:23:16.960
Um, lovely to have you here.
00:23:17.440 --> 00:23:25.759
We have got a few questions for you, really reflecting your experience at the intersection of tech and AI and sustainability.
00:23:26.079 --> 00:23:33.519
And I'd like to start with a question that's on everyone's lips right now, which is Is AI good for sustainability?
00:23:33.680 --> 00:23:34.559
What are the challenges?
00:23:34.720 --> 00:23:36.400
What are the opportunities?
00:23:37.440 --> 00:23:39.359
Ah, thank you so much for having me.
00:23:39.519 --> 00:23:43.039
I'm really excited to be here and unpack this with you.
00:23:43.200 --> 00:23:57.599
Uh, this is a question I'm getting all the time from organizations that we train at Climate Collective, uh, from friends, from peers who've also worked in sustainability for, I mean, I've been in this space for 20 years.
00:23:57.839 --> 00:24:02.960
It's people are concerned and worried, and there's a lot of uncertainty.
00:24:03.119 --> 00:24:05.839
So very happy to be here and unpack it with you.
00:24:06.079 --> 00:24:11.200
I would say to ground ourselves first, this is gray.
00:24:11.359 --> 00:24:12.079
I'm sorry.
00:24:12.319 --> 00:24:15.440
It's neither or on its own, right?
00:24:16.079 --> 00:24:17.759
AI will be good.
00:24:17.920 --> 00:24:22.240
AI could be really positive for whatever we point it towards.
00:24:22.480 --> 00:24:41.359
If we ask AI to supercharge sustainability teams, decarbonization strategies, the clean energy transition, it will supercharge that, make it faster, better, allow it to scale, simplify financing, all the things that we really desperately need.
00:24:41.680 --> 00:24:50.720
But if we just let AI sort of exist on its own, if we embed AI in the current economic system, it will just reinforce what's already here.
00:24:50.960 --> 00:24:54.559
And as we all know, that's not exactly sustainable.
00:24:54.880 --> 00:25:00.559
Um so, two things I'd like to just add when this comes up.
00:25:01.279 --> 00:25:10.400
AI will only move the needle on sustainability in a meaningful way if a large amount of people use it for emission reductions or protection biodiversity.
00:25:11.039 --> 00:25:22.240
If we're just doing like cat pictures or birthday invitations on Chat GDP, it's not gonna solve staying within planetary boundaries, right?
00:25:22.720 --> 00:25:28.960
And second, and this is probably the more important point, not AI, not all AI is the same.
00:25:29.200 --> 00:25:32.400
Not all AI will have the same positive and negative impact.
00:25:32.720 --> 00:25:41.200
The most valuable contributions to sustainability won't come from consultants or conservation nonprofits using chatbots.
00:25:41.440 --> 00:26:03.680
That will help them and their organizational efficiency, but the big, like really big improvements that we are desperately needing will come from applied AI embedded in industrial systems, optimizing logistics or improving smart grids, finding new protein sources, accelerating carbon removal in assets.
00:26:04.000 --> 00:26:06.799
That's really where the big bang for the back will come.
00:26:07.440 --> 00:26:27.680
That's so interesting and incredibly helpful because effectively what you're saying is we need to really understand the role of AI in accelerating systems transitions that are underway and not view it as something separate or a threat or an opportunity in it in its own right.
00:26:27.920 --> 00:26:37.039
But the exam question, if you will, is you know, how might AI accelerate the energy transition, accelerate the food systems transition, health systems transition?
00:26:37.599 --> 00:26:53.440
On that basis, then how do we get that message out more concretely and cut through all the sort of noise about AI is going to destroy jobs or AI is going to clean my data for me?
00:26:53.599 --> 00:26:57.920
How do we sort of elevate that sort of strategic application of AI?
00:26:58.480 --> 00:27:00.079
That's a great question.
00:27:00.400 --> 00:27:08.160
Um I think currently the big narrative out there is kind of anchored around fear.
00:27:08.400 --> 00:27:15.039
Unless you adopt AI really fast, you're going to be missing out, you're going to lose your job, do you have enough money on your savings account?
00:27:15.440 --> 00:27:24.160
None of that is an inviting narrative, a narrative that opens up for exploration or joy or curiosity.
00:27:24.799 --> 00:27:27.279
Nor is it particularly like regenerative.
00:27:27.759 --> 00:27:38.079
So I think first we need to, well, my theory of change as an organization is kind of a mini-universe of that question you have.
00:27:38.319 --> 00:27:57.680
I think one is uh educating people, putting, uh helping people understand what AI is, what it isn't, and when it could augment or complement humans and human expertise, and when it really shouldn't, because the risks are too high.
00:27:58.319 --> 00:28:00.000
I call that AI literacy.
00:28:00.079 --> 00:28:07.680
So, like raising AI literacy amongst the general public, amongst civil society, amongst governments.
00:28:07.839 --> 00:28:13.039
I mean, how are how could governments regulate this technology if they don't quite understand it?
00:28:13.200 --> 00:28:14.319
I mean, that's right.
00:28:14.559 --> 00:28:22.319
So I think lifting understanding and talking about it is one really critical component.
00:28:22.559 --> 00:28:26.160
The other piece is to invite for transparency.
00:28:26.240 --> 00:28:34.480
I mean, we do trainings for conservation nonprofits, uh, net zero alliances, corporate sustainability teams.
00:28:34.720 --> 00:28:40.160
And I can't even tell them with real numbers exactly what their AI footprint is.
00:28:40.559 --> 00:28:49.519
So that creates this kind of uncertainty where people are like a bit worried but pushed to hurry up and quickly about this so they're not missing out.
00:28:49.680 --> 00:29:09.680
And as humans, we always always find a way to game the system where we always start optimizing for like who's using the most tokens, who's burning the biggest AI, who has the biggest AI bills, and and like none of those indicators are KPIs that we should really measure ourselves as a society.
00:29:09.920 --> 00:29:17.039
So yeah, I think agency, like building agency amongst people, civil society is really the key here.
00:29:17.680 --> 00:29:37.039
And Anna, when you talk about the narrative and the fears and the uncertainties, certainly in our space where we're concerned about sustainability, the first thing that comes up is AI's energy footprint, and then AI's water footprint, and then of course all of the ethical issues around how AI might be used.
00:29:37.200 --> 00:29:48.480
And I wonder how you see that uh also in contrast with the potential upside, even on energy or or on water or on many other solutions.
00:29:48.720 --> 00:29:55.839
Um is there a kind of net positive uh narrative as well that we're we're failing to get across?
00:29:57.279 --> 00:29:57.920
Yes.
00:29:58.160 --> 00:30:03.119
Um there are very little research with actual data on this uh yet.
00:30:03.279 --> 00:30:17.279
Um it's a new topic, but you know, that there's just not enough information out there, so we can't quite measure the exact energy footprint or the exact water usage of certain um facilities.
00:30:17.519 --> 00:30:27.839
I think that's silly because think about when you go to a grocery store, you can often find out which fishing boat a specific fish was caught on.
00:30:28.079 --> 00:30:30.319
You can even find out what were their fishing practices.
00:30:30.559 --> 00:30:35.119
Like we've figured out a way to trace a lot of information through value chains.
00:30:35.359 --> 00:30:38.079
Why can't I know where my AI was processed?
00:30:38.240 --> 00:30:43.920
If I'm using Claude, why can't I know what data center those prompts were routed via?
00:30:44.160 --> 00:30:46.079
I mean, technically it's totally possible.
00:30:46.240 --> 00:30:50.240
So it's not just a matter of like, I can't access that information.
00:30:50.480 --> 00:30:50.799
Yeah.
00:30:51.519 --> 00:30:58.400
It makes a big difference, I think, if your AI was processed in a data center operating on a dirty grid, like I'm from northern Sweden.
00:30:58.559 --> 00:31:06.799
I'd feel really good if my prompts were routed through that data center in northern Sweden that has almost entirely renewable energy.
00:31:06.960 --> 00:31:12.400
But I would feel really uncomfortable if it's going through a data center that's not powered by natural gas.
00:31:12.559 --> 00:31:20.079
Because then kind of my impact as a sustainability organization just feels kind of it just doesn't feel it feels misaligned.
00:31:20.559 --> 00:31:20.880
Yeah.
00:31:21.359 --> 00:31:25.839
And maybe also to remind people that we're at such an early stage with AI.
00:31:25.920 --> 00:31:30.720
And if we looked at the efficiency of the internal combustion engine when it just came out, it would be terrible.
00:31:30.960 --> 00:31:32.319
Same with solar panels.
00:31:32.559 --> 00:31:37.920
So, you know, I expect a technology innovation uh curve that will happen there as well.
00:31:38.240 --> 00:31:45.359
But what do you say to companies now who are grappling to internalize, to adopt, to apply AI?
00:31:45.599 --> 00:31:54.480
What you hear so much about is is maybe speed over wisdom, I would say, just trying to do it faster and as quickly as possible.
00:31:54.640 --> 00:32:13.599
Um and then I've also heard talk among businesses about the difference between using AI as a tool and becoming AI native, which I think speaks more to what you were saying is oh, we're actually using it to change our products and our services and our processes, not just to become more efficient.
00:32:14.640 --> 00:32:15.279
Yeah.
00:32:17.359 --> 00:32:22.960
All of these points are quite poignant right now, and I think you're real you're right.
00:32:23.119 --> 00:32:39.759
We're so early that even someone like me, who I spend all my waking hours thinking and reading and doing sustainability and AI, and still there's so much that happens every day that I'm struggling to keep up with.
00:32:40.319 --> 00:33:03.599
Um I think for corporations, when we work with sustainability teams, corporate sustainability teams, they are struggling, they're struggling from kind of a large responsibility and burden of reporting requirements often and kind of communicating why a green option makes sense for the for the for the corporate narrative.
00:33:03.759 --> 00:33:05.680
And I think AI can be really helpful here.
00:33:05.839 --> 00:33:12.000
AI really can act as a kind of bridge that ties together different investments and different strategies across a company.
00:33:12.240 --> 00:33:42.400
AI can also help them produce certain things, let's say like sustainability reporting, using AI to enhance that can make a lot of the kind of burdensome, heavy data processing slim down and instead allow them to focus on kind of the more where their human judgment is more important and kind of our where connections and and and and that humanity is more valued.
00:33:42.559 --> 00:33:52.319
So I think there's there are really great examples already emerging of how corporates can embed AI in their own sustainability strategies and programs.
00:33:53.039 --> 00:33:59.759
I think a responsible AI approach would be to also support regulation.
00:34:00.160 --> 00:34:07.519
I think corporates have a really important role these days in like underlining the need for regulation.
00:34:08.239 --> 00:34:13.280
We missed the boat on regulating social media and see where that got us.
00:34:13.440 --> 00:34:18.880
Um I think self-regulation has also not worked well for climate or nature contributions.
00:34:18.960 --> 00:34:22.079
So I think we have an opportunity here to do this better.
00:34:22.480 --> 00:34:28.639
Um and last one, I would I would tell corporates too, you know, look at your suite of partners.
00:34:28.800 --> 00:34:29.840
Who do you partner with?
00:34:30.000 --> 00:34:31.920
Do they know how to use AI well?
00:34:32.079 --> 00:34:37.199
Could you support them as you're learning how to do this responsibly and sustainably?
00:34:37.280 --> 00:34:39.760
Could you support your partners in doing the same?
00:34:40.079 --> 00:34:45.119
For example, we partner with Clarnow, which is a large fintech company in Sweden.
00:34:45.440 --> 00:34:49.280
They're a long-standing backer of climate and nature entrepreneurs.
00:34:49.519 --> 00:34:55.039
The first thing they did was to ask their portfolio companies if they have the resources to use AI well.
00:34:55.360 --> 00:35:00.960
When they realized they didn't, they then came to us and asked if we could help build AI literacy in their companies.
00:35:01.440 --> 00:35:13.760
And just from some trainings and interventions, we're already seeing this massive boost in their in these organizations' output, saving costs, reaching new fundraising targets, and having an impact.
00:35:14.800 --> 00:35:16.480
And that wasn't possible before.
00:35:16.559 --> 00:35:27.840
So I think corporations have a really new level of kind of responsibility looking inwards, but also responsibility looking outwards for their partners and and value chain.
00:35:30.079 --> 00:35:34.960
Anna, just um our last question to you, if we can.
00:35:35.280 --> 00:35:48.239
Um that was a really lovely and different insight into AI, and particularly how AI can be framed as potentially accelerating these systems transitions.
00:35:48.719 --> 00:35:57.519
Would love I guess your advice for organizations that are committed to their role in systems transitions.
00:35:57.679 --> 00:36:07.840
What's your kind of top one, two pieces of advice for change makers in businesses that want to wrap their arms around AI and use it for systems transformation?
00:36:08.079 --> 00:36:11.519
And what's your advice for us as individuals?
00:36:15.039 --> 00:36:20.320
I think we have a lot to learn from how corporations approached the SDGs.
00:36:20.639 --> 00:36:32.239
One of my jobs at when I worked at Meta was to build out Meta's response to the sustainable development goals with the goal of sort of accelerating progress on the SDGs.
00:36:33.440 --> 00:36:35.119
We then looked inwards first.
00:36:35.199 --> 00:36:38.719
We thought, like, okay, what are we calling superpowers because it's a tech company.
00:36:38.800 --> 00:36:40.880
So what are Meta's superpowers?
00:36:41.039 --> 00:36:45.440
Uh what does what could Meta bring to the table that's unique for them?
00:36:45.679 --> 00:36:59.599
And in our case, when we looked at then trying to match that with the different SDGs that were priorities, we found, well, we have unique data sets, with unique uh AI capabilities, engineers and compute.
00:36:59.840 --> 00:37:17.039
So then we kind of anchored our interventions and our programs and our partnership outreach around those things that we knew we were already doing really well and could do more of in sort of a for good social impact contribution.
00:37:17.280 --> 00:37:41.199
So I would say for at least for corporations, think about what you're good at and try to anchor your sustainable AI strategies, interventions, principles around that, because that will just lead to a much more long-term sustainable contribution versus you know doing something kind of like that's not core to your business.
00:37:41.599 --> 00:37:45.519
I think that also holds like the world is a tough place today.
00:37:45.760 --> 00:37:55.280
So staying close to kind of your lane, what you're really strong at, I think will ensure that these interventions are more longer-term uh sustainable.
00:37:55.760 --> 00:37:56.000
Great.
00:37:56.320 --> 00:38:12.639
And then for individuals, we did this crazy thing in February when I called a bunch of women I worked with, 40 women actually, that I worked with throughout my career, in all with leadership roles in climate.
00:38:12.800 --> 00:38:15.760
And I all asked them, like, Do you do you feel good about AI?
00:38:15.840 --> 00:38:17.440
Do you know what it is?
00:38:17.920 --> 00:38:19.360
Do you understand it?
00:38:19.679 --> 00:38:24.320
Are you using it for your own sort of efficiency at work?
00:38:24.639 --> 00:38:32.639
And they're all, you know, generally all of them just said, I have no idea what this is, it scares me, I'm really concerned about the environmental impact.
00:38:32.719 --> 00:38:35.920
I'm sitting on the sidelines feeling really stressed.
00:38:36.159 --> 00:38:40.559
So we just came together as a, we call it the women in climate community.
00:38:40.800 --> 00:38:44.239
So over the last 16 weeks, we've just met once a week.
00:38:44.559 --> 00:38:55.840
And my team brings some skills, some kind of basic tutorials, and we experiment with different technologies and we talk about what they're good at, what they're not so good at.
00:38:56.079 --> 00:39:06.400
Um, and then we have a WhatsApp group where we share articles that we read that upsets us or that gets us happy and excited, and we discuss them in community.
00:39:06.639 --> 00:39:12.800
And I think the I didn't know that I needed a women's group, but I really did.
00:39:12.960 --> 00:39:31.599
Um, and I think joining together, even if it's a group of colleagues, maybe it's a handful of neighbors or it's friends, just coming together and kind of unpack this topic that's moving so fast that is has the potential to really like rewire society, economies, communities.
00:39:32.559 --> 00:39:36.000
I think this is a time where we should kind of do that together.
00:39:36.400 --> 00:39:49.679
I love that because what you're essentially saying is let's harness the energy in the system and dance with it to flow that energy into the direction we want it to flow and let's realise we all have agency.
00:39:49.840 --> 00:39:51.760
There's something we can all do.
00:39:52.239 --> 00:39:54.480
Thank you so much, Anna, for joining us.
00:39:54.639 --> 00:39:56.559
It's been lovely to chat with you.
00:39:56.719 --> 00:39:58.079
Um thank you.
00:39:58.480 --> 00:39:59.280
Thanks, Anna.
00:39:59.599 --> 00:40:00.239
Thank you.
00:40:04.559 --> 00:40:07.599
That was also really, really interesting.
00:40:07.840 --> 00:40:09.440
What struck you, Wayne?
00:40:10.159 --> 00:40:19.519
Well, I think what what came across for me is that uh you know, whatever you point AI at, it will accelerate.
00:40:19.760 --> 00:40:20.079
Right?
00:40:20.159 --> 00:40:21.599
That that's really what it's doing.
00:40:21.679 --> 00:40:23.280
It's like a supercharger.
00:40:23.760 --> 00:40:34.480
Um and the question that emerged from these uh discussions was um, you know, it doesn't help to accelerate in a broken system.
00:40:35.199 --> 00:40:40.719
So it doesn't uh alleviate the necessity to actually change the system itself.
00:40:41.119 --> 00:40:46.639
And if we're able to do that, then AI could probably uh accelerate the solutions.
00:40:47.039 --> 00:40:52.559
So the question we always have to ask when we talking about AI is AI for what?
00:40:52.800 --> 00:40:54.480
What's it being pointed at?
00:40:54.880 --> 00:40:55.519
How about you?
00:40:55.679 --> 00:40:56.880
What what intrigued you?
00:40:57.440 --> 00:40:59.280
I think that's really interesting actually.
00:40:59.360 --> 00:41:01.760
Uh AI needs the right architecture.
00:41:02.000 --> 00:41:16.880
Um But I think what struck for me was uh social movements uh when it comes to driving that accountability for AI and particularly data centres, might do what policy isn't yet doing.
00:41:17.039 --> 00:41:25.360
I was really struck by the fact that people can see the data centres, they are lowering water tables in some parts of the US.
00:41:25.760 --> 00:41:38.159
Joel was talking about the fact that uh this issue has stepped over, you know, traditional political lines, and communities are finding their voice, and I think that is really interesting.
00:41:38.480 --> 00:41:45.760
So could communities, could society become the accountability mechanism for AI and tech?
00:41:46.239 --> 00:41:50.800
Yeah, maybe a lever we we hadn't necessarily anticipated.
00:41:51.360 --> 00:41:53.440
So um could be very powerful.
00:41:53.519 --> 00:41:57.039
Well, what's the dance vibe you get for for for the sector, Sally?
00:41:57.440 --> 00:42:05.920
So I've thought about this a little bit, and I'm gonna go for the pase dobla because it is a dance of incredible power.
00:42:06.079 --> 00:42:10.639
You've got the bull charging at the matador, but who's gonna win?
00:42:10.800 --> 00:42:25.039
Because they're both really, really powerful, and it's dynamic, it's fast-paced, and so when I think about tech, there's a big pass dobla going on right now between the hyperscalers, community, society, policy, and where will that cape go?
00:42:25.360 --> 00:42:27.360
Yeah, and there are red flags there, right?
00:42:27.679 --> 00:42:29.760
Absolutely, as there are in tech.
00:42:30.159 --> 00:42:38.400
Well, uh that that's it for this episode, and uh we look forward to you joining us next time uh when we dive into another sector.
00:42:38.960 --> 00:42:39.840
Thank you.
