Science4Parliament - Special AI edition - Part 2

Welcome to the second of the special AI editions of the Science4Parliament podcast. These episodes are a summary of three workshops which took place online in early 2024 to inform the IPU’s Artificial Intelligence (AI) resolution, which was adopted at the IPU’s General Assembly in October 2024.

This resolution aims to encourage parliamentarians to consider the social and ethical impacts of this new technology and the issues that they, as decision-makers, should be aware of when considering legislating for AI to ensure that its development and use is fair and beneficial for all of humanity.

The workshops were designed as part of the journey to the resolution, as a learning tool and also to stimulate interest and debate. The process was steered by two rapporteurs, Michelle Rempel Garner, Member of the House of Commons of Canada, and Neema Lugangira, Member of Parliament of the United Republic of Tanzania, who moderated the first two sessions, and I moderated the third session.  

Session one covered the basics of AI technologies, how they are developed and used, and how they are impacting the world, session two sees a deeper delve into the emerging impacts of AI on society and how governments need to work to harness these potential benefits and mitigate any harms. The third session was an assessment of legislation in relation to AI, what is currently in place and how to plan for what may be needed in the future.

Speakers:

Tawfik Jelassi, Assistant Director General for communications and information with UNESCO, provided the opening remarks. 

Joel Bilt, a senior fellow at the Center of International Governance Innovation and the associate professor of economics at the University of Waterloo, Canada, talked about the steps that the growth of artificial intelligence, or any other technology, is likely to follow. 

The final speaker in the session was Philip Thigo, advisor to the Government of Kenya and a member of the United Nations High-Level advisory body on artificial intelligence, he spoke about the changes that these new technologies are having on society and humanity.

More information 

The draft AI resolution, ‘The impact of artificial intelligence on democracy, human rights and the rule of law’, was published on 25 July 2024, following extensive collaborations with parliaments and experts, and adopted at the 149th IPU General Assembly in October 2024.
It is available on the IPU’s webpage, IPU.org.

The links to the other two workshop summaries are below; please do share them with anyone you think might be interested:

Science4Parliament - Special AI edition
Part 1 - https://www.buzzsprout.com/2249694/episodes/15896531
Part 3 - https://www.buzzsprout.com/2249694/episodes/15902412

If you would like to listen to the seminars in their entirety, you can do so on the IPU’s YouTube channel @IpuOrg 

If you have any comments or questions, please text the show at the link at the top of the page, email me at dnaughten@gmail.com or follow me on social media:
LinkedIn:  https://www.linkedin.com/in/denis-naughten
X:               https://x.com/DenisNaughten
Blog:         https://substack.com/@denisnaughten
Web:         https://denisn

Transcript

Science4Parliament - Special AI edition - Part 2

SPEAKERS

Denis Naughten, former chair of the Inter-Parliamentary Union working group on science and technology and member of Parliament in Ireland.
Tawfik Jelassi, assistant Director General for communications and information with UNESCO,
Joel Blit, senior fellow at the Center of International Governance Innovation and the associate professor of economics at the University of Waterloo Canada
Philip Thigo, advisor to the Government of Kenya and a member of the United Nations High Level advisory body on artificial intelligence.

Denis  00:00

Welcome to the second in a series of special artificial intelligence editions of the Science4Parliament podcast. Episode one was a summary of the basics of artificial intelligence technologies, how they are developed and used, and how they are impacting on the world and how governments need to work to harness these potential benefits and mitigate any harms. The third session looks at legislation in relation to artificial intelligence, what is currently in place and what might be needed in the future. This episode is a summary of session two, which was a deeper dive into the emerging impacts of artificial intelligence on society, on work, on labour, on human rights and the impacts that it's having on the democratic institutions themselves.  Joel Bilt, senior fellow at the Center of International Governance Innovation and the associate professor of economics at the University of Waterloo Canada, was the first speaker and talked about the steps that the growth of artificial intelligence, or any other technology is likely to follow. The final speaker in the session was Philip Thigo, advisor to the Government of Kenya and a member of the United Nations High Level advisory body on artificial intelligence, who spoke about the changes that these new technologies are having on society and humanity. 

Tawfik Jelassi, Assistant Director General for communications and information with UNESCO, provided the opening remarks. 

Tawfik Jelassi  01:41

This session will explore a key question, what are the implications of AI for our societies and our economies? Directly linked to this question is the imperative, how can parliamentarians collaborate to forging and enabling environment with a two-pronged approach, crafting and enacting laws that foster green innovation and development while safeguarding a human rights standard, privacy, data protection and security. We know that the rapid evolution of AI requires a forward-looking approach to understand the challenges and opportunities. As parliamentarians, you hold the crucial role in shaping the legislative and policy framework which should and must guide AI and digital transformation towards sustainable and inclusive growth for all we believe that to be effective in AI policy making, we must address four key questions. First, how can we implement agile legislative approaches? And we know that time is of the essence here. We know that the laws and the legal framework has always been trailing technological advances, so we need to be agile. We need to be quite fast in putting in place the legal legislative framework needed. Second key question, why are intersectoral collaboration and stakeholder engagement? Why are they so vital for effective policies on artificial intelligence? Thirdly, how does ensuring gender equality and inclusivity in AI policies, how does it mitigate the bias and foster equity in our societies? Fourth and last, why is it essential to address knowledge gaps through capacity building and capacity development initiatives?

Denis  03:59

Joel Bildt, senior fellow at the Center of international governance innovation, and associate professor of economics at the University of Waterloo Canada, gave us a technology history lesson looking at how changes and development happens.

Joel Blit  04:18

What I hope to give you is a better understanding of how AI is going to be developing over time. So what to expect, what kinds of changes to expect over the next couple of decades, both to the economy and to the world and society more broadly, but maybe even more significantly, I hope to give you a sense of what opportunities AI is going to be presenting to your country, to your economy and to government as well. So how government can reinvent itself by using these technologies, you know, around AI, so that's those are really my main goals. So let me start with this. I often give a version of this talk to executives for major international companies and their child. Challenge is really to reimagine their products and services, in other words, their business model, to reimagine their business model around artificial intelligence. And they're also facing some time pressure, of course, because their competitors are doing the same. 

Joel Blit  05:12

But the reality here is that industry executives have it very easy, because this is the only thing really that they need to worry about. Government, on the other hand, has three separate things to worry about. So for the first is, is very much the same, right? You need to be thinking about how you can reorganize the way that you're offering your services, your programs, around this new technology. So you need to be reinventing yourself so that that isn't common. But in addition to that, as government, you need to be leaders in promoting the adoption of AI so that you can actually have so you can have an economy that prospers, a country that prospers. And you also need to be thinking about regulating AI to minimize its downsides and to make sure that the benefits are spread widely and not just to the few. And so, you know, relatively speaking, the challenge of government is absolutely monstrous compared to what industry executives are facing. There's really these three different things that you need to be thinking about at the same time. Now I hope that my remarks there are going to help you think at least at a high level about what is coming down the pipeline, and therefore what you can do in terms of these three things. That's my goal today. Obviously, we're only going to have so much time in sort of another 15 - 20 minutes. 

 Joel Blit  06:24

Okay, so let's, let's jump right in, and let's go back into a trip into history, right? So let me talk about general purpose technologies and their adoption. And so when I talk about general purpose technologies, I'm talking about things like steam engine, I'm talking about electricity, I'm talking about computers. And of course, the most recent general-purpose technology is artificial intelligence. And so general-purpose technologies, this is a technical term that we use in economics, but it really first technologies that have these three different characteristics. First, they're pervasive. And what that means is that it's not a technology that's going to impact one sector or one industry, it's a technology that's going to impact many different sectors. Second, it's a technology that is subject to ongoing improvement. So it's not a technology that appears and is set once and for all. It's a technology that's constantly changing and improving. And thirdly, it's subject to innovation complementarities. And what that means is that there's other things that are evolving alongside it, and are and are feeding into the technology. So the best historical example I can give you is, if we think about computers, it would be software, right? So as software gets better, computers get more and more valuable, and as computers get more valuable, there's a bigger incentive to develop better software. And so there's a strategic interplay between the two. Now, without getting into the economic details of these characteristics and what they imply, these characteristics imply, first of all, that general purpose technologies are transformational, and secondly, that it truly takes decades to feel their impact. I was on a panel quite recently, and I was asked, is AI hype? That was the question, and my answer was, well, relative to what it's going to deliver in the next two years, yes, AI is hype. We're not going to see the kinds of changes that a lot of people are expecting in the short term, but if you give it a decade, it's going to make changes that we can't even imagine today. And so this is the first lesson coming out of previous disruptive technologies. We actually have a little bit more time than we think, even though this one in particular is moving really quickly. But I think the more significant lesson is that these the adoption path of all of these technologies follow a they always follow a fairly similar adoption path. And so this can shed light into how we can expect AI to be adopted over time.

Joel Blit  08:47

And so they largely follow these three phases of adoption, these technologies. And so this is what I call the three R's, the three phases of adoption. And so in the first phase, which I call the replace phase, this is where the new technology displaces the old way of doing things. And so often it's old technology, but some, but often as well, it's just people doing things. And so the new technology is going to displace the sort of the new technology displaces the old technology, and we see increases in efficiency and exist and, you know, more efficient ways of doing things. Now, that said, the actual processes themselves, the business models, don't actually change, right? So in the replace phase the business models, the processes are fairly fixed, and what happens is that the old technology gets pulled out, and then the new technology gets put in to do that particular task. Now the processes themselves, the business models, they don't change until the next phase, which I call the reimagined phase. And so in this reimagined phase, the new technology allows the complete reimagining of processes and business models around the new technology. And lastly, we get into the recombined phase, and this is where the technology starts combining with other technologies to create entirely new things, because this is really important for policy makers, especially. There's going to be winners and losers with some of these technologies. Winners are going to be the kinds of people that have entrepreneurial skills, critical thinking skills, problem solving skills, decision making skills, and the losers are going to be the people that have the kinds of skills that AI can do.  

Joel Blit  10:17

The changes for policy makers. Because not only do governments have to use this technology to reinvent themselves in the way that they offer services, right, they need to move through these phases, but they also need to be promoting the adoption of AI in the broader economy. If you want prosperity for your country, prosperity for all, you can't be stifling this, and you need to be driving this technology forward. And that means that you need to be driving literacy. You need to be driving trust. You need to promoting adoption and driving innovation around AI, and, of course, entrepreneurship around AI. And lastly, of course, incredibly important regulation, because there's going to be many challenges around this. I haven't gone into details here, but we really need to strike the right balance between promoting it and at the same time regulating it.

Denis  11:05

Philip Thigo, advisor to the Government of Kenya, and a member of the UN High Level advisory body on artificial intelligence, talks about the remarkable journey that technology has taken, from the iPad to one which is now changing how society operates. 

 
Philip Thigo  11:22

We hardly understand this, but the  iPad was born in 2010, and it is more powerful 200 or 300 times more powerful than the computer that sent man to the moon in the Apollo mission. It just tells you how quickly and how rapidly things have changed, but also how this era basically talks about where we are. Things are moving very quickly, and so how technology is reshaping how we live, learn, work and communicate. We can no longer distinguish the world of work and our personal lives, and I think all of us experiencing so we are definitely a globalised world where we are living without borders, and so we have to think about technology or governance within a borderless and connected world. And we are highly mobile. You couldn't have done this before. We're doing this on Zoom. People have very new behaviors in terms of we live a very public life. There's feedback, there's collaboration, and so this sharing, AI would not have been possible without these new behaviors, where we just have vast amounts of data that we are putting out there so that these technologies can train themselves.

Philip Thigo  12:36

 I think where we are right now, and that is why I think there's a lot of concern is this era of generative artificial intelligence, right, where now the traditional systems are now able to design, recognize patterns, make predictions, create new content and in many forms, right? Not just text, can do audio, videos and more. And I think that's where we are in land. I think I don't know, a lot of us have seen ChatGPT ,Midjourney, and when I was having a chat with some colleagues, just to get people's thoughts they said we are still very far from AI being able, for example, to remember things, to be able to do kind of functions of what human beings can do. Basically, AI is not even a four-year-old human being right now. So for now, maybe that's the comfort, but it keeps on running as they develop the fifth iteration. But I think the reality is that generative AI is neither artificial nor intelligence intelligent, and the idea is that artificial intelligence still is trained on human data and depending on what types of data sets we put out there, it can be damaging and garbage out and so I think it's very important that we begin to recognize that. 

Philip Thigo  13:49

And of course, with this, we are seeing disruption of jobs, business disruptions, issues of ethics and identity because of the biases we've seen. We're seeing potential issues of security, security and conflict when AI is being deployed in the field of war based on autonomous weapons, the whole issue of of inequality, but also the need for agile governance, because you're governing in an era of uncertainty. But then who are we governing? We know there are many personas. Uh, human citizens are not homogenous, and so, how do you ensure that we still take care of a farmer who has a simple smartphone, does not have good intellectual activity? How do you ensure that an elderly person is not displaced, even dislocated in an era of technological acceleration? How do we still ensure that a tech entrepreneur in a small village in Tanzania or Nairobi can still compete with a Fortune 500 of Silicon Valley company like IBM, Amazon Web Service. But also, how do we ensure that we still protect these spaces for youth activism and civic participation and engagement of democratic processes? And so we've seen the challenges around this, of course, the issues. Privacy and security as technology evolves, so are the potential nefarious use of technology like hackings and cybersecurity and cybercrimes increasing. The issue of digital divide has even gone beyond digital to an opportunity divide because of where we see technology going. Of course, we've seen resistance to change, and I think the first speaker mentioned this. So how do we overcome bureaucracy and tradition? Again, leveraging on technology, but then all of these things are being done within a resource constraint or within physical constraints. But also our procurement systems do not understand how to procure technology, because it's not about servers anymore. It's really about every aspects of our lives and government. 

Philip Thigo  15:46

And finally, how do you measure impact? So I think a lot of these challenges are what we see, but potentially, I think what policy makers also we are seeing the stress that it's creating simply because of how technology works, whether these borderless companies really blurring the issue of identity, issues of legal jurisdictions, but also taxes. We were seeing companies are more exponential, that workers either work in your country, but also others work remotely, so as a governed and also creating stress on the tripartite. We're seeing new risks around us, as I mentioned, data, but also around subversion of elections. There are, of course, issues of ethics and inclusion, and the largest disruption, of course, is out of what I share these presentations. But I think the new challenges on governance, and we see it, is that again, there's a continued erosion of trust, simply because of how technology is out there, the governance models that could manage and mitigate this are not Just Kelly, Fast enough you and so how do we ensure that we grow technology as a public good, but also that the dispensation, or the governance dispensation, has an imagination that then can bring the positive visions of technology for the benefits society? There's a lot of scaremongering right now, and some of us are saying the similarly when the cars were being built. It's also important to figure out, how can you use the same technologies in their duality for development outcomes?

Philip Thigo  18:15

We need, honestly, as much as we are, as much as we think around governance, is to think about resource distribution and access. There's already an imbalance on how AI the fundamentals of AI, data, compute and talent is distributed is yes, within wealthy nations, but also large technology funds. How do you how do we mitigate against that? The second piece of course, is there's an AI divide beyond a technology divide, especially in the Global South, how do we, even as parliaments, who have power over budgets, still figure out that the basics don't change? We still need to invest on data systems. We need to invest in Internet and broadband access. We need to invest in how we reform our education systems so that you can skill, reskill and upskill our generations, especially the global south. To be able to compete in this market, we need to invest in electricity. A third of the African continent, for example, hddas power. But globally, even in this early age of AI, already, 10% of global energy resources goes towards compute, and it will only increase. So how do we ensure that that is done? But also, how do we ensure that a lot of this infrastructure developments are done in a way that a lot of countries that are AI poor do not get into more debt and can do this within constrained fiscal space? How do we build local AI ecosystems. Think about, I mean, we have translation here, in this, in this call itself. It means some communities need their data and languages digitized so that AI models can be trained on those languages, so that they can have a future. Are beyond being invisible within this new AI award. So when I talk about bias, it's about data gaps and data sets that are not that not exist. So how do we potentially invest and ensure that that happens? But also, how do we double down on issues of access, on issues of benefits and issues of capacity and two things for me, for parliament, I think for me, is, one is, how do we ensure that because your people's representatives, how do we enhance citizen participation and AI fluency? How do we as much as possible design this mixed regulation that, yes, put in guardrails, understand the risks, but then do not stifle innovation? And finally, how do we ensure that we ourselves adopt an agile growth and digital mindset that also couples by international collaboration and then local action in terms of use cases, so that we can better understand where the failures are, where the risks are, but there were the opportunities to learn. 

Denis  20:59

Thank you for listening to this special artificial intelligence edition of the Science4Parliament podcast. I hope that this brief overview of the interParliamentary Union's artificial intelligence workshops for parliamentarians will give you food for thought and will help you in your work as you strive to contextualize and regulate these new technologies as they are evolving. The draft artificial intelligence resolution entitled ‘The impact of artificial intelligence on democracy, human rights and the rule of law’ was published on the 25th of July, 2024 following extensive collaborations with parliaments and experts, and was adopted at the 149th InterParliamentary Union General Assembly in October of 2024. It is available on the InterParliamentary Union (IPU) web page, ipu.org, the links to the other two workshop summaries are in the episode notes. Please do share them with whoever you think might be interested. If you'd like to listen to the seminars in their entirety, you can do so on the IPU's YouTube channel, which is at IPU org under the artificial intelligence heading, and all episodes of the Science4Parliament podcast, are on Spotify, Apple podcasts, or wherever you get your podcasts. If you have any comments or questions, please text the show at the link at the top of the page, or email me at denis.naughten@oireachtas.ie