Subscribe
Share
Share
Embed
The Responsibility of Investing (formerly The Principles for Responsible Investment) is a podcast by the Principles for Responsible Investment (PRI), the world’s largest global body on responsible investment, representing over $128 trillion in assets under management. Each episode features conversations with thought leaders and experts from around the world, exploring how sustainable factors are transforming the investment landscape. Listen for unique insight into how climate, nature and human rights issues are affecting asset classes and responsible investment policies.
The series helps PRI signatories - and the wider investment community - navigate responsible investment with greater precision and confidence, for the benefit of both investors and society. No matter your size, market, nor stage of the responsible investment journey, The Responsibility of Investing will bring you a new perspective every fortnight.
00:00:07:21 - 00:00:33:02
Cambria
Hello everyone, and welcome to Perry's podcast, The Responsibility of Investing. My name is Cambria Allan Retzlaff, CEO of the principles for Responsible Investment, and it is my absolute pleasure to lead today's discussion on why AI is a sustainability issue. So artificial intelligence is moving fast, faster in many ways than the frameworks investors use to understand and manage risk.
00:00:33:04 - 00:00:58:15
Cambria
And while most of the conversation in financial markets has focus on AI as a sector to invest in is a quieter but arguably more important question. What does a rapid deployment of AI across the whole economy mean for investors who care about long term investing and sustainability? That's what this conversation is about. We've brought together three people doing some amazing work.
00:00:58:15 - 00:01:25:06
Cambria
Some of the most interesting work on exactly this from the perspective, the fact of stewardship, policy and investment, governance and ecosystem coordination. The aim is simple to make. I feel like a concrete, practical, responsible investment issue. And to give you a clearer sense of what you can do about it as an investor. So let's jump in first by bringing on our guests.
00:01:25:08 - 00:01:33:01
Cambria
And so please introduce yourself to the audience. And I will start with Michael.
00:01:33:03 - 00:01:58:06
Michael
Thank you very much, Cambria. And thank you for, hosting this, very warm all by important session that I think is, is going to be very informative for you, Mike Sullivan and the guest also on the show and our listeners as well. So I'm Michael Bennett, Yamaha. I work for iOS at furry Hammies limited. The vice president's to worship director and I engage on behalf of over 2.4 trillion assets under advisement.
00:01:58:08 - 00:02:08:19
Michael
For our clients, mostly institutional investors. So kind of my daily job is, involves, talking to board members and directors and executive of the, sort of issuer companies.
00:02:08:21 - 00:02:16:02
Cambria
Thanks, Michael. And welcome and may I ask, why should a long term investor care about AI as a sustainability issue right now?
00:02:16:04 - 00:02:35:23
Michael
Well, I think you laid it out right. If you are a universal sort of owner. This is probably a whole economy effect, and it's going to touch you one way or the other to your investment portfolio. And so looking across beyond tech itself and even into other industries where things are shaping up to be very, very impactful. So it's a whole economy thinking.
00:02:35:23 - 00:02:44:24
Michael
And I think as a universal owner, you kind of want to be aware of the risks and opportunities and how, you know, you set out your investment policies and strategy.
00:02:45:01 - 00:02:55:03
Cambria
Thanks, Michael. All right. Let's move to Chris. So tell us about yourself. And then same question. Why should a long term institutional investor care about AI.
00:02:55:05 - 00:03:15:19
Chris
This can be great to be with you. Thanks to PRI for listening to this conversation. I'm Chris Jurgens, I work at Omidyar Network, which is a philanthropy and investing platform focused on responsible technology. Our mission to is ensure that our shared humanity steers our digital future. And chief of staff to the organization and our leadership team.
00:03:15:21 - 00:03:48:03
Chris
And I run our portfolio focused on responsible technology investing with a significant emphasis on AI. So the business case for a sustainable investor to focus on, I think there's three reasons. First, cited just the sheer magnitude of the AI revolution, the level of capital expenditure going into AI, greatest of any tech revolution in history as a proportion of GDP, the pace of growth of the core AI firms, the rate of enterprise adoption, fastest technology adoption in history, and how we're already seeing AI signals shaped markets.
00:03:48:03 - 00:04:14:20
Chris
The so-called SAS apocalypse, where anthropic or open AI released a new capability and a whole category of software stock. Prices moved dramatically. But then, with the sustainability lens, I think there's a strong business case with the kind of dual materiality frame to it. So if we look just that financial materiality, we think responsible AI does drive both enterprise performance and portfolio returns.
00:04:14:21 - 00:04:37:05
Chris
Survey from Reframe venture of VC investors last year said that 73% of companies with strong or responsible AI practices will be more financially, successful, avoiding costly incidents. I think more importantly, building customer trust. We've seen how important trust is in AI as a market. And part of that is quality product that's safe and secure and accurate and free from bias.
00:04:37:07 - 00:04:58:22
Chris
Part of that is crisis prevention and ensuring, AI doesn't have harms in terms of cybersecurity or domains like defense. But part of trust and responsibility is ensuring companies follow through on their commitments to stakeholder and society through the disruption of the AI revolution. Right? Are they staying true to their commitments in spite of energy usage of AI?
00:04:58:24 - 00:05:29:01
Chris
Are they're staying true to their commitments to people and workers despite the labor force disruption? Potential? We think the companies that have responsibility and trust at the center are those that are winning the market. And that's a, you know, a business case for sustainability. And then I think also the outside in perspective of systemic risks, and impacts, materiality, climate, energy consumption, of data centers is one of the biggest new factors driving climate commitments and and outcomes.
00:05:29:01 - 00:05:49:05
Chris
Already data centers take up 4% of U.S. electricity consumption. That'll go to 12%, within a few years. Impact on labor force. Right. Dario Modi of anthropic saying, hey, I could disrupt up to 50% of entry level jobs in the next five years. We don't know that's going to happen. But it's an uncertainty and commitments to local communities, right.
00:05:49:05 - 00:06:08:09
Chris
Already a lot of activism in the US and other places around data centers, impacts on local communities. And so the magnitude of the impacts, the business case that responsible AI is better AI and that kind of AI is going to win in the market. And systemic risks carry out, I think, are all reasons that this should be front and center for for all investors.
00:06:08:11 - 00:06:29:02
Cambria
Thanks, Chris. There's a lot to unpack there. So we'll we'll definitely dive into it. And last but certainly not least will tell us more about yourself. Your work. Because you're really sitting at the intersection of policy security and investment governance. And, you know, in your view, why should a long term institutional investor care about AI?
00:06:29:04 - 00:06:59:11
Oumou
Well, thank you so much for having me. This is certainly a really timely conversation. By way of background, my name is Emily. I am a nonresident research fellow at the center for Long Term Cybersecurity at UC Berkeley. I'm specifically on their AI security initiative, leading work on exactly this building tools, frameworks, methods to help investors sort of better understand their role in mitigating AI risk in the aggregate, at scale and in helping them mitigate the materiality risks that I can introduce.
00:06:59:13 - 00:07:36:10
Oumou
And that's three core reasons why I think I, as a core and really top sustainability issue right now. The first is that AI is today an increasingly a driver of operating cost for targeted portfolio companies. And that means where it's deployed and influencing operating margins and the way companies are thinking about their own competitive positioning. The second reason is that regulators across several major global jurisdictions are focused on how AI systems perform in real world deployment context, and because the really essential goal of regulators is to is to reduce harm in the aggregate or at scale.
00:07:36:11 - 00:08:02:19
Oumou
They're concerned with things like safety bias, reliability, interoperability and regulator interest creates, you know, new and emerging compliance and liability roles that investors can't afford to ignore. In the United States and in the UK in particular, investors have become the subject of regulations on AI security, particularly where there is a national security or national competitiveness angle. I know we'll get to this.
00:08:02:21 - 00:08:33:11
Oumou
And the third thing is, Chris, sort of head on is that the volume, and pace of capital formation and of AI adoption is driving massive build out of physical infrastructure. That's datacentres, that's water and energy systems and their associated supply chains and the sustainability considerations associated with physical infrastructure as well noted and discussed. And those include energy costs, energy availability, the environmental impact, the impact on local communities, and more.
00:08:33:13 - 00:09:00:12
Oumou
I think that any of those dynamics on their own would raise really significant sustainability considerations, but taken together, they create a real urgency. AI is becoming systemically important, important to market performance, which means that disruptions, whether they're from from resource constraints or from geopolitical issues or operational failures, can lead to valuation shocks and just in general, make the market more sensitive to correction.
00:09:00:14 - 00:09:17:15
Oumou
So from my perspective, for a long term investor, the core issue is whether these newer AI native and AI intensive businesses can scale and continue to grow over time without running into bottlenecks that hurt, returns, or hurt the environment or undermine user trust.
00:09:17:17 - 00:09:42:09
Cambria
Great. Thanks, sumo. So so you hear we have a quite a full panel. So we have Chris who brings the ecosystem and coordination perspective. We have Michael who brings a practitioner perspective, and certainly Oumou who brings a policy security and investment governance perspective. So let's dive in a little bit more. So each of you has been engaging with different parts of the investor ecosystem on AI.
00:09:42:11 - 00:10:06:15
Cambria
So where do you see the biggest disconnect between what investors should be doing and what they're actually doing, and where are the signals, whether from regulation, from a market or from your own research that point to genuine opportunity? So with this question, I'm going to start with Chris. So tell us more about Omar's research and work. What what's surface there.
00:10:06:17 - 00:10:46:20
Chris
So at a media network, we're focused on supporting investors to make responsible AI a core corporate governance issue. And so we will work directly with investors across public and private markets and a wide range of investor networks and field building organizations. I think the core thing we're seeing on the positive side is in really fast growing appetite for investors to engage on responsible AI and some flagship investors leading the way on both company and manager engagements, but more widely in the market, seeing a really big gap between investors appetite to engage and their capacity and infrastructure to do so.
00:10:46:22 - 00:11:17:14
Chris
So, a survey of venture VC investors just for March of this year said that only 14% of, VCs rated their AI risk capabilities as good, 27% felt they had sufficient internal expertise on the topic, and that compared to 75% who thought they were pretty strong on ESG issues overall. So like a capacity skills gap, unsurprising given the newness of AI and the complexity of its impacts.
00:11:17:16 - 00:11:44:05
Chris
In terms of opportunities, I point to four things where we need to move forward. First, developers versus deploying. A lot of the engagement today is focused on the AI developers and hyperscalers, open AI, anthropic, Microsoft, etc. that's a logical place to start. Hugely influential across the whole AI value chain, but we think a lot of the risks, impacts, liabilities and opportunities are going to be at the deployment and application level of the chain.
00:11:44:07 - 00:12:12:06
Chris
And so the shift from starting with the players to, developers to the companies deploying AI across the economy is one shift we need to see and accelerate. Second, engagement is typically started with overall good governance principles. We think that's right. We think that the core governance issues, it should start with how companies are ensuring board and C-suite, oversight of AI risks and opportunities.
00:12:12:08 - 00:12:36:07
Chris
But a lot of the risks and impacts are very specter industry and business model specific. And so we think the next step is going from overall good AI governance practices to a really sharp diagnosis of what are the most material risks and issues, at the level of banking or insurance or healthcare, kind of moving to a sasb materiality assessment that's very industry specific.
00:12:36:12 - 00:12:58:05
Chris
Like we have, other issues. As the second key shift, we get to see. Third, we see a lot of frameworks out there for AI risk, management. But few of them have been built for institutional investors in mind. So you have nest, you have the UAE act, and you have individual investors that have come out with their own risk assessment frameworks.
00:12:58:08 - 00:13:17:21
Chris
We don't yet have shared, standards or frameworks across the market like we have on climate or other issues. We think the time is right to start building towards that. There was something as fast moving as AI that's going to be a more complex endeavor. And then the final shift, I think, is from individual action to more coordinated field, action.
00:13:18:00 - 00:13:55:09
Chris
So again, a lot of organizations and investors individually doing good things, but they've been limited capacity for for field orchestration. But we're excited for the emergence of collective action initiatives like the world benchmarking like Collective Impact for ethical AI, which brings together public markets asset managers to engage companies on on AI governance. We recently convened a field building group of 30 investor networks, and funders on how do we collectively move this field forward so greater opportunities to, work together to share best practice, to learn, and to make this more century great.
00:13:55:15 - 00:14:28:24
Cambria
Thanks, Chris. So, so, Chris, I've been talking, quite a bit about efforts that sort of have have grown and based on me kind of as things progress. But I'm hearing there's a need for coordination on one end and also perhaps more specificity when you're talking about industry on the other. Let me go to you. Can you, pick up the regulatory signals you think investors might be under reading here and then perhaps along with that, is there a role for government?
00:14:28:24 - 00:14:34:12
Cambria
Is there the, you know, kind of on the government side, what should investors be focusing on now.
00:14:34:14 - 00:14:52:00
Oumou
I'd start to answer this by saying that I really became, a topic of global conversation and especially of policy conversation, while as a while I was in my last role in the white House, I was in an office called the office of the National Cyber Director, which was responsible for being the lead advisor to the president on cybersecurity policy and strategy.
00:14:52:00 - 00:15:11:02
Oumou
And so, as you can imagine, once GPT three was released, really our team spent a lot of time thinking about the role of government, the role of the public sector in general, really the role of actors across society and shoring up some of the risks that we all observe, and trying to drive towards the outcomes that we all seek.
00:15:11:04 - 00:15:55:08
Oumou
One of the things I'd say is that one of the most important things for investors to understand is that AI, from a security perspective, is a force multiplier for attackers. It accelerates many of the tech and security risks that we already face, and introduces a new layer of, operational and technical risk for organizations on the attack side, AI has made cyber attacks cheaper, faster, more convincing, and eliminated some of the like previously requisite skill and funding needed to launch convincing convincing offensive phishing campaigns deepfakes, malware and other forms of social engineering, which generally makes these attacks harder to track, harder to take down.
00:15:55:10 - 00:16:17:16
Oumou
On the defense side, companies are plugging AI into workflows and security tools, which then expand the attack surface and can expose sensitive data if those systems are tricked or misused. So in that way, AI systems themselves can become targets for manipulation. Which means they themselves are can become modes of failure or critical failure points inside companies.
00:16:17:18 - 00:16:41:01
Oumou
For investors, I think the implication, is twofold because, one, there is sort of this new, space of direct risk to enterprise resilience and to business continuity, as well as considerable liability exposure, because, again, regulators are concerned about harm reduction in the aggregate. And then second, this is something that that affects lives and affects public trust.
00:16:41:03 - 00:17:10:13
Oumou
These dynamics really affect how much confidence people have. And the systems that they rely on every day. And so from my perspective, the security dimension of AI risk is as extremely salient, because that's what regulators are going to be really, really focused on. And the United States, for example, outbound investment rules can sometimes restrict or require disclosure of, cross-border investment, particularly where there are dual use military surveillance kind of applications.
00:17:10:15 - 00:17:30:24
Oumou
And then in the UK, the same thing is also the case under the National Security and the Investment Act, the formation of regulations around, around AI security and investors specifically becoming the targets and the subjects of those regulations, I think, is something that we will likely see more of in the coming years as the geopolitical environment becomes more dynamic.
00:17:31:01 - 00:17:59:23
Cambria
Okay. So we heard from Chris. We're talking collective action, the growth of collective action. And also the importance of grounding in industry. We were just talked about, the, the government side and what regulators will be watching, but also some areas for investors to, to watch out for, which certainly include security, especially as I practice has become more intertwined with, regular business.
00:18:00:00 - 00:18:21:07
Cambria
So, Michael, tell us about what you're seeing on engagement front. Where are the signals, from regulation, from market or from your own research that point to genuine opportunity. And, you know, where is the disconnect between what investors are doing and maybe where they should be going, where they should be going?
00:18:21:09 - 00:18:48:12
Michael
I think sometimes we do have play the the newness of AI quite a bit. And obviously it is part of the ontology now and the building taxonomy, many of us remember before I as I said, it was all the fears about big data, which was the big thing every company wanted to get into big data. Then we jumped on to this idea of predictive analytics, which is a form of sort of machine learning based AI sort of principle and an approach.
00:18:48:14 - 00:19:15:03
Michael
And we got into metaverse and in fact, I brought a lot of excitement. We saw companies that are, setting out metaverse strategy, you know, changing names and a whole apparatus of different sort of new lingua that came into being. And then we started seeing some of the Alban academic research that sort of brought a lot of AI, sort of now what we call the agent ecosystem or environment into play, that people were now able to personalize your usage of it.
00:19:15:08 - 00:19:36:12
Michael
And I think the inference point is what kind of created this sort of big brouhaha, and pressure for investment, but also for investors as well, and for governments around the world. So from an investor point of view, I do sometimes we do the words like responsible there's responsible that it becomes part of the nomenclature very, very quickly.
00:19:36:16 - 00:20:11:16
Michael
Some of us remember, you know, clean coal, clean this. We are very, very quick to kind of gravitate towards this sort of principle, this principle that and investors have to kind of deal with the the reality of the fact that they have a fiduciary, whatever. Nice. They don't want to or your, an asset manager or within other spectrum of the investments that are vehicles, you have the core fiduciary responsibility that are called fiduciary responsibility in the US, obviously, under the Delaware sort of doctrine, including business judgment rule for a lot of boards, requires you to think about risk and opportunity, right.
00:20:11:16 - 00:20:27:07
Michael
Downside, you know, an upside risk all together. And I think that is a core aspect of every investor sort of work. Every day you wake up and you are thinking about your clients, and you're thinking about what your clients need to know and need to have to ensure that you are seeing the returns that you are supposed to see.
00:20:27:09 - 00:21:04:12
Michael
And I would like to go down that said a fiduciary part into the fact that we within the ears sort of ecosystem on behalf of our clients, our clients as an institutional asset owners, but they also have members many cases. You're pensioners now. I'm bringing this down to that level because I think sometimes when we talk about AI, we tend to put it in the spectrum of framework, this sort of concept and this sort of product that exists in vacuum, these are monies and dollars and decisions that are moving from one part of our economy to the other, and in classic print to sort of creative destruction, sort of concept.
00:21:04:14 - 00:21:24:09
Michael
We are seeing the evolution of our sort of a global economy to a larger aspect. Having better seen this before. Yes, we have a telecom, a boom, the rumor mania in the UK and that solid destruction of significant GDP at the end of the era, even the telecom sort of boom, we saw how different value chains got impacted.
00:21:24:15 - 00:21:48:03
Michael
So when you're talking about AI, you're not just talking about just the hyperscalers or the frontier AI model creators, but you also talking about the utility players. You're talking about the community water. So the municipal water entities, you also talking about pension is you also talking about our own sort of livelihoods, right. Utility costs, you know, as we've seen in some sort of market, have gone up by few percent, right?
00:21:48:09 - 00:22:12:09
Michael
We've seen demand for new sort of investment into infrastructure along the transmission line for energy. We're seeing sort of decisions about how we're going to address climate change in the face of the fact that we are going to see, on prem, especially behind the meter, solutions that are not necessarily about renewables, that are products, but we are seeing, in a sense, some sort of fossil based sort of, aspect to it.
00:22:12:12 - 00:22:36:22
Michael
So you have sort of new technologies somehow being built on the back of old technology. And that raises what I call a grid of uncertainty. So the uncertainty is what we are dealing with. I think we can come up with all the frameworks and principles which in the iOS environment, we do with our engagement is really driven by by good corporate governance is driven up by ensuring that the board and management are thinking about long term decision.
00:22:36:24 - 00:22:54:20
Michael
But I nisso has a short term, medium term and long term risk. I think, we'll talk about some of the geopolitical consequences that we are seeing. You. Right. One of the things that we are seeing the market right now is you have to have an AI strategy on AI submission, or you don't get the full capital inflow that you required.
00:22:54:22 - 00:23:15:03
Michael
So even if you don't have any thought through your strategy, you have to say something, right. And so many people coming to the market, sometimes without a very robust strategy. And our job as engages is to really ask boards and companies about how do you thinking through that long term investment horizon? Yes, it is okay to talk about the fact that you have an internal AI strategy.
00:23:15:06 - 00:23:46:18
Michael
If you is a company that is trying to adopt AI is okay to talk about AI investment. If you are a hyperscaler that is trying to lay the pipe in terms of the infrastructure for the future, it is okay to think about investment or reasons that are required. But in the same breath, we want to be sure that the business judgment rule that is required of board members are brought to effect, which is that you are thinking about those core risks and opportunities, because at the end of the day, remember what I talk about, who actually the beneficiary of a lot of the work that we do, it is those pension years is those future
00:23:46:20 - 00:24:07:15
Michael
that is relying on incomes and benefits, that comes from the monies that are being invested, that we are engaged in on behalf. So there's a wider societal implication that what is design implication is that it affects all of us, and our engagement is driven by the idea that as an investors have dealt with risk. Part of what we do is always risk adjustment return.
00:24:07:17 - 00:24:32:05
Michael
And the goal is to try to minimize downside risk and increase upside risk. And if we can do that, that requires not just having better models and better policies. But you know, I'll be selfish here to say that it requires better dialog. And that's why, you know, my job, I at least I think is very important because I wake up going into some of those meetings asking questions that help us understand more of what is going on.
00:24:32:10 - 00:24:57:09
Michael
There's so much noise about the impact of AI on jobs, about the impact of AI, on regulations, on climate, on other sort of sustainable development of gender. There's impact on utilities, on water backlash, misunderstanding energy use. All of those and many different sectors are trying to figure out what the solution are. I think the runway is quite long in terms of how far we go, if history is any sort of lesson to us.
00:24:57:15 - 00:25:20:08
Michael
But I think from an engagement perspective, our core duty is to understand how those who are responsible for the decision making are thinking about all of that different risk and opportunity, and applying that business judgment rule. Because you have to remember, in the US, the core mandate of the board is to shareholders. That is something that is embedded within the doctrine of the Delaware sort of system.
00:25:20:10 - 00:25:52:16
Michael
Right. And the business judgment rule is how you are held accountable. In Europe. It's quite different if you when you look at the framing of the EA we are at, there's a lot of responsibility on individual directors. You could be sued as an individual director based on the output of frontier AI models. Right? And I think that is where we have to draw attention for a lot of our investment, investment communities that our ability to engage in the dialog, whether it is benign or actually very extensive and comprehensive, it is the most important thing because trust me, the companies themselves are having the same conversation.
00:25:52:18 - 00:26:00:10
Michael
The roadmap is not there. It is not unusual, but it's an unusual technology that transform the coal fired power supply.
00:26:00:12 - 00:26:31:01
Cambria
This is really a governance question as much as anything else. Thanks. So let's let's, continue on our journey and drill down a little bit into the sustainability lens. So we've established why this matters and where the opportunity sits. Let's get into a little bit more of the substance, because, you know, when you're actually looking on at what AI deployment means across different sustainability themes, the picture gets more complex and also more concrete.
00:26:31:07 - 00:26:47:20
Cambria
So I want to dig in a little bit more, into what each of you is actually finding on the ground. So let's, let's keep pulling on that thread. So, Michael, let's go back to you. You've been in the room with energy companies, utilities, major tech firms. You you alluded to a bit of that in your earlier response.
00:26:47:24 - 00:26:57:21
Cambria
Can you tell us a little bit more about, you know, what you're finding and is the market getting it right? Are there some areas where there might be some recalibration?
00:26:57:23 - 00:27:18:08
Michael
No. You raise a fundamental challenge that is in the current currently in the market space, I would like to sort of bring one thing that I said earlier on, which is the idea of uncertainty. I mean, at the beginning of this sort of big boomerang, around AI and data center in particular, we had a lot of numbers about what is the potential energy demand required.
00:27:18:08 - 00:27:34:13
Michael
And some of the numbers were really out there. We saw some projections from Goldman Sachs. We saw some from McKinsey. From the US. I well, not from the just a US energy agency, but also the IEA as well. But I think the crux of it is that we know we're going to see significant energy demand right now.
00:27:34:13 - 00:27:55:08
Michael
What does that mean? Especially in the US context? I used to live you obviously in the Washington DC area. My last other place I live was Reston. Around me was a lot of data centers. Right. And anybody who lives in Reston is very, very, familiar with this constant dialog with City council, members. I happen to live in Dallas, Texas now, where it is booming again.
00:27:55:08 - 00:28:22:04
Michael
I was telling my friends that I can't seem to get away from data centers because around me, I'm back into one of city council meetings about data centers. So it pulls me everywhere I go. And I think the real thing here is that we because energy demand is a big part of it. It brings it back into a lot of our sustainability discussions around climate change, around water, you, around, you know, sound and noise sort of pollution.
00:28:22:06 - 00:28:57:05
Michael
We hear things about community backlash, land use and land is also another big part of it. And so if you've been around a lot of the ESG, sustainability environmental discussion, the most feels like deja vu. You know, this is a 2.0 of another era of discussion. And companies across utility tech companies and other value chains and Uma and I have spoken about this before, including sourcing critical minerals that are important part of the value chain around certain materials, including rare, that are required for so many different technological innovation.
00:28:57:07 - 00:29:16:05
Michael
Or at the heart of it, when you look at critical minerals and you talk to companies that actually make racks and you know them, the HP's Adele's of the world, I mean, they require a lot of this minerals. And if this minerals are going to be in the discussion, well, guess what you're talking about another resource dependency with its own environmental consequences.
00:29:16:11 - 00:29:36:22
Michael
We know the challenges about with mining, the particular the missions involved and the process and the human rights implications of the value chain. And now you add on top of that, as Bruno will kind of allude to, as well, is this geopolitical security around these minerals where EU has its own critical minerals policy, Australia has come up with one, Canada, all of those.
00:29:36:22 - 00:30:03:10
Michael
And then you have the natural competition for those same resources along auto and that sort of tech companies. Another traditional use of individual consumer, appliances. So we are inflection point where if you are on a board of a company and you are thinking about all the risk within the value chain, it goes away from inference into, wow, the kind of infrastructure required is never seen.
00:30:03:10 - 00:30:24:06
Michael
Business may seem like this before, right? And that raises a lot of question. Do tech companies have the right kind of board and the right kind of executives who understand the resource and infrastructure and the physical requirements of this space? I mean, when you look across the employment landscape right now, you see a lot of tech companies hiring advisors around energy sourcing and negotiations.
00:30:24:10 - 00:30:42:16
Michael
I mean, it's a big deal. Frontier AI operators are doing the same thing. Hyperscalers are doing the same thing. But we also have to be very careful because in the utility space, there was also a lot of speculation. One investor or one particular developer could bit the same utility demand to about three different sort of utilities and that have an escalation effect.
00:30:42:16 - 00:31:05:02
Michael
And this has this whole thing in the bubble discussion of real. Is the demand very real? Absolutely. Before the high demand, we had transmission delays that were up to seven years. Now we had a new administration. We are seeing some policy rollbacks or policy changes that are effectively trying to create opportunity for all of this. But for investors, I think the bigger questions are actually quite simple.
00:31:05:04 - 00:31:27:23
Michael
Thus, those companies involved in evolution had the right kind of board of directors and oversight that really at the right questions to get to some semblance of the right answers. Right. We're not saying the perfect answers, but the right answers. And he did do what actions are they taking? Are they changing policies? Are they changing strategies? Are they now looking at your risk mitigation sort of issues?
00:31:28:04 - 00:31:51:04
Michael
How are they approaching a lot of these things around environmental concerns? I could give numbers here about climate implications and all of those different kinds of stuff. But I think where we are right now is not a matter of do A or B or C or D, I'm an economist. I always say there's going to be a trade off if a country or a company is going to take one particular path on energy, they're going to have to forfeit something out.
00:31:51:06 - 00:32:11:02
Michael
We're in a game where the speed to scale is actually the new nomenclature. And it's very, very important that we as investors understand why that is an important because to win the race, we have to get the right infrastructure in place. And the right infrastructure requires companies and governments to make certain trade decisions. And I think that's what the tension is.
00:32:11:07 - 00:32:17:00
Michael
So I'll go back to governance again, because that is where everything rests, at least from my point of view.
00:32:17:02 - 00:32:45:23
Cambria
Now that's, that's that's great. And as, as someone who, who lives in Michigan where a lot of these issues are very much live, especially the energy affordability issue, the a natural resources issue, the how much agency does the community have? I mean, all of these issues that we often talk about in sustainability are just crashing into each other, and investors are, you know, kind of have to navigate all of that because all of that stuff matters in whether the investment is successful and all of that.
00:32:45:23 - 00:33:09:05
Cambria
So it's it's just a fascinating area. Let's pivot a little bit more on this. I'm going to go to you. You've been thinking about AI's a security and stability issue, and you talked about it a little bit in your opening. What should investors understand about this, and how might it affect people's livelihoods as well as the investment risk?
00:33:09:07 - 00:33:32:02
Oumou
Yeah. Well, I I'll reiterate what I said earlier, which is one that I really stands to accelerate the range of really adverse tech harms, security harms that we're already experiencing at scale. And, the traditional software ecosystem, for cyber security, for example, you might hear about things like ransomware incidents, taking healthcare systems offline, hospitals offline, schools offline.
00:33:32:04 - 00:33:56:10
Oumou
I really stands to augment, accelerate and really enlarge the success of of those kinds of operations. So that's one risk. I'll go a level deeper though, because I think there are really two key specific security issues that investors have to pay attention to. One is on, data security. AI is, of course, built on data. It runs on data.
00:33:56:10 - 00:34:18:01
Oumou
It lives off of data. And the security of the data that underlies the moments that the models, the components of the models and, and the systems that users actually interface with really is going to be a source, some of the most significant reputational and compliance risk that investors might face. So that's one thing I'll flag.
00:34:18:03 - 00:34:41:22
Oumou
The other thing I'd say, is that it's important for investors to, to resist the signal that, that, that, the market appears to give that says the first mover advantage is the really the one that counts the speed of adoption and the pace, of adoption isn't necessarily the same. As innovating. It's not necessarily the same as, enabling the technology to really scale and provide the benefits that we all expect.
00:34:41:24 - 00:35:08:19
Cambria
All right. So we've been talking about security. We've been talking about communities. We've been talking about infrastructure. But certainly the one of the biggest topics that we often hear of is labor and labor displacement. And it's probably the risk we hear most about. And perhaps, you know, our listeners might feel the least equipped to assess.
00:35:08:21 - 00:35:29:11
Cambria
So, Chris, let's go over to you. Because Omidyar has, done a lot of work in this space in general, what should investors actually be asking about? How AI is affecting the workforce and workforce of companies, and what would a credible answer look like?
00:35:29:13 - 00:35:57:02
Chris
Such an important question as there's so much uncertainty about, labor market impacts for investors, but for workers themselves, I think a number of areas investors should probe. First, the central debate about AI labor is the augmentation versus automation question, automating jobs fully away or leveraging AI to make the existing employees, more productive and innovative. The path of least resistance is to go to automation, where you can cut costs.
00:35:57:04 - 00:36:29:04
Chris
But we think a lot of the sustainable, long term value creation is going to be rooted in augmenting workers. So I think investors should probe specifically with companies. Where are you looking to automate? Where are you looking to augment and how and link that not just to the cost cutting, but to innovation and value creation. Second, we know from decades of technology implementations that technology implementation drives more value when employees are engaged in how that technology is designed utilized in the workplace.
00:36:29:06 - 00:37:03:12
Chris
So investors should be asking, do workers have a seat at the table in how AI is being designed and deployed? And AI is not just happening to them, but it's it's with them. So that leads to better value creation. Third, companies are investing massive amounts in the technology and infrastructure. Investors should ask, are your investments in workforce training, and preparedness commensurate with your investments in AI infrastructure for spending hundreds of millions of dollars on the tech and workforce enablement, then you're missing the productivity gains.
00:37:03:14 - 00:37:33:22
Chris
Fourth, we know from AI that a lot of the innovation, experience to date experience the date that a lot of the innovation comes from experimentation, from individuals playing with AI and seeing what works. But you can't have that in a corporate IT department with everything running amok, setting a fourth line of inquiry could be, how are you enabling experimentation through controlled sandboxes and places where people can experience while having good governance in place, from an overall technology governance perspective?
00:37:33:24 - 00:38:01:24
Chris
And then I think a final ask that we have, always from a governance perspective, as digital was fosters is disclosure on AI and, the workforce. How much are you spending on training? More data, more disclosure, more forthcoming information because there's so much uncertainty about the workforce. We need information to to try to get some sense of clarity of where this is taking things at an enterprise level and at the overall level of, the labor market.
00:38:02:01 - 00:38:20:21
Cambria
All right. So we've covered a lot of ground. Before we get into the practical side, if you had to identify the single biggest opportunity for investors in this space, where would you focus, each of you, one sentence. Let's go. Michael.
00:38:20:23 - 00:38:32:05
Michael
I'm just going to be selfish and say, governance, governance, governance, governance is, it is the old trick in the book, and it still remains the most strongest and most effective part of it.
00:38:32:07 - 00:38:42:02
Oumou
I go governance, governance, governance, governance, period. And just don't be afraid to lean in. You're helping shape the eventual outcomes, which will be good a plus.
00:38:42:02 - 00:39:15:24
Chris
One of my colleagues on the governance front, but an opportunity we think investors should be focused on that we haven't talked about yet is the proactive opportunity to invest in building a better AI stack, safer AI? We invested in a company that makes AI chips that use a fraction of the energy of the existing alternatives. There's a huge market around AI safety and assurance infrastructure, things like deepfake detection, algorithmic auditing, red teaming, the kind of, seatbelts and catalytic converters and airbag analogy for cars.
00:39:15:24 - 00:39:25:08
Chris
We need that same sort of, infrastructure for AI. And that is a huge proactive investment opportunity in venture private markets and in the corporate open markets.
00:39:25:10 - 00:39:47:19
Cambria
Great. All right. Thanks everyone. So we've covered a huge amount of ground today. And I hope what's come through is that this isn't a distant or technical issue. It's already showing up in portfolios and company disclosures and regulatory frameworks. In the conversations investors are having with the companies they own. But knowing why it matters is only half the story.
00:39:47:21 - 00:40:07:12
Cambria
In our next conversation, we're going to get practical. What can investors actually do? What tools and frameworks are emerging to help them? And where does collective action make the biggest difference? I hope you'll join us for that next week. Until then, I'm Canberra Ali Retzlaff, and this is the responsibility of investing.