Responsible AI: governing market failure

If society seeks or needs responsible development and use of AI technologies, how is this best achieved?

This month the Australian government published its analysis of public submissions on its April 2019 proposed 'Ethical AI Principles', and published a revised set of principles: here. 

In my April submission (in this repository) among other things I put three points, which I summarise here as I believe they remain 'live':

1. A national conversation

The first point was about processes, such as the public enquiry, of arriving at and promoting such lists of principles (whatever their content). This process or that of the Australian Human Rights Commission are no substitute for a genuine, scaled national conversation, indeed a global one. As I submitted, that conversation is not about 'what should our ethical AI principles look like' but (if AI is truly as transformative as we think) about the more fundamental question 'how should we live [and what role do we want and not want for technology in that attempt at flourishing]'.

2. The missing governance piece

The second point was to ask how the listed principles are intended to take or be given effect, which is a question not of ‘principles for ethical AI’ but of ‘the governance of principles for ethical AI’. Every major government and tech company has or is producing such lists. What are the mechanisms by which, in various contexts, we think they are best given effect? Since they are 'ethical' principles, I hesitate to say 'how are they complied with' and 'what are the consequences of non-compliance'. Which leads to my third point.

3. Ethics vs law / regulation

The third point was to suggest that the real question (in seeking submissions) ought not to be whether the 8 listed principles in the Australian framework are the ‘right’ or best or most complete ethical principles. Some ethical AI frameworks have more (e.g. Future of Life's 23), some have less (e.g. the OECD's 5, or Google's 7). The prior question ought to be whether responsible AI development and use is best approached as a question of ethics rather than as a question of law and regulation.

I reflected on this third issue in a previous post (here): there is a very live law and regulation aspect here (as useful as ethics-based approaches are, and complementary to law).

This month's revised approach notes:

• "The framework may need to be supplemented with regulations, depending on the risks for different AI applications. New regulations should only be implemented if there are clear regulatory gaps and a failure of the market to address those gaps."

This is, on one view, a remarkable proposition, if not an outright abdication of governmental responsibility for promoting responsible AI. 

It is a proposition, unless I am mistaken, that in relation to AI -- which the Australian framework process explicitly states is so fast-evolving, so profoundly transformative, so pervasive -- posits that:

(a) law and regulation is only a 'supplement' to ethics-based approaches; and

(b) the market [whatever that means!] should be left to address 'compliance' with ethical principles, and the people's elected law-making bodies should only have a role where gaps [whatever that means!] are 'clear' .

For one thing, by the time we diagnose that there has been a market failure to encourage or enforce responsible AI development and use, it will be rather too late to start asking law-makers to get out their legislative drafting pens and address 'gaps'.

Lawyers and law-makers can stand down: we are not needed here, or now. Australia, that sophisticated regulatory state, has decided that the market -- which of course has proven soooo socially responsible hitherto -- can regulate this issue just fine.

Jo 

'Flourishing Cities': partnering for resilience

How can effective, appropriate cross-sector partnering help in 'future proofing' the developing world cities of the future?

'Flourishing Cities' is the theme this week of our annual Challenges of Government conference at the Blavatnik School of Government here in Oxford.

From an Africa perspective (as this blog is), the conference programme speaks to the scale of urban development and human security challenges in fast-growing cities. Far from Oxford's pleasant setting, these challenges come across very vividly in the heaving expanse of cities from the Cape Flats in South Africa to Cairo.

(The Nile city was the setting of a related post a year ago now on the 'turnaround challenge' for business in mega-cities (here)).

Yet by focusing among other things on the progress of one troubled Colombian city, the conference agenda also speaks to the largely untapped potential for scaling-up dialogue and partnership between policymakers, business, civic groups (and often donors).

The evident and largely latent scope for greater cooperation and collaboration holds considerable promise for unlocking developmental bottlenecks in ways that make commercial sense for business and investors, too.

At some level, there are strong shared interests across sectors in moving beyond guarantees of minimal security so as to enable human flourishing and the attainment of basic aspirations. It goes almost without saying that business and governments share an interest and indeed an imperative in promoting more inclusive and sustainable growth and poverty-reduction; more and better education, job-creation, and productivity; greater shared prosperity and reduced inequality (of income, healthcare, security and so on); greater social cohesion and reduced radicalisation; and so on.

At this level, the case for partnering is not hard to make. Much harder is to give effect to such ideas, and to ensure that there are principles to guide the pragmatism involved in greater cooperation across sectors in meeting the development agenda.

Moreover, there is a growing shift to focus on city-level issues, from investors to development agencies. Schemes and initiatives proliferate: Rockefeller's '100 Resilient Cities', Columbia University's 'Millennium Cities' initiative; IBM's 'Smart Cities' initiative, and the list goes on. Again, the case for focusing on urban development and resilience challenges is easy to make, even if partnering is hard: finding the right incentives for sustained cooperation, the right relationship parameters, deciding who counts as 'business' or 'the private sector' in selecting partners, negotiating relationships with national- and provincial-level governments.

As the previous post on 'innovation' noted (here), there can be a tendency in using the term 'partnership' to gloss over not just how hard and entrenched development challenges can be, but also how hard, piecemeal, or political partnering can be (even if 'development' was easy!).

What we are working on at Blavatnik, among other things, is getting to the heart of the fundamental concepts around why partnering works or does not work, is appropriate or inappropriate, is embraced or resisted.

These challenges apply generically: one myth in the recent turn to city-level programmes, conferences, investment strategies (etc.) is that by descending to the supposedly more agile and adaptive city level of analysis or administration, one can by-pass some of the enduring problems of cross-sector collaboration and get more done.

This is an appealing idea -- but not necessarily a sound one.

Jo 

Africa: the use and abuse of 'innovation'

No-one wants to be heard to oppose 'innovation'.

It is not fashionable to suggest that the search for new approaches, whether in business or policy-making, can perhaps distract us from honest analysis of why conventional approaches are not working.

While the 'innovation!' cry seems everywhere at present, it is not often clear what it means to foster innovation in any field.

This is so, too, in relation to the issues often covered in this blog: the regulation of responsible business in developing societies, and the governance of public-private development cooperation.

Implicit in that call, of course, is a recognition that conventional approaches are not working. 'Old' approaches (for example, to promoting the rule of law in other societies) are often conceptually sound but face enduring barriers in practice. No amount of privileging innovation over implementation might be able to surmount these barriers. Yet the prevailing 'innovation' rhetoric can hold out the false promise that these barriers might be side-stepped altogether.

We hear a lot about the need and/or scope for innovative approaches to unlocking Africa's potential. I'll call this the 'innovation trend'.

In parallel is current rhetoric around Africa as a continent whose poorer and unemployed millions are best understood as would-be innovators and entrepreneurs. I'll call this the 'innovator trend'.

In development policy circles, the rhetoric of the innovation trend is undeniably positive. It supports a perspective that seeks potentially transformative break-throughs and short-cuts. The innovator trend is equally positive: it has an empowering intent that casts the continent's poorer people as having economic and development agency and potential, rather than as mere passive recipients or observers.

First, the 'innovator trend'. The goodwill that accompanies current portraits of the continent as full of latent 'innovators and entrepreneurs' has an unintended dark side. Many millions of Africa's rural and urban poor are not 'innovating' but simply trying to secure less precarious livelihoods; to cast them as incipient entrepreneurs is less condescending and opens the way for (self-)empowering approaches. But it can also represent a denial of the reality of poverty, and of the political context for addressing it.

I wrote on this recently in something published by the Bertha Centre for Social Innovation at the Graduate School of Business at the University of Cape Town, South Africa: here.

Turning to the 'innovation trend', it is becoming rather skewed. By 'innovation' in African development is typically these days meant technological innovation, and increasingly digital / IT tech-based innovation. The field is attended by what I think is unfounded hype about the transformative potential of IT tech-based solutions.

In particular, much of the excitement around at present is based on unsound analogies that take mobile phone sector trends in Africa and, from these, project that the continent's new dawn is not only at hand and handheld, but is less than a finger-length away. This conversation suggests that if only the right technologies and ap's could be designed and scaled-up, in development and growth terms Africa would soon join Asia, and then overtake Scandinavia... This is not helpful even if it is uplifting.

I noted (or rather, ranted) in a previous post that one cannot necessarily 'leapfrog' all development, governance and growth problems in Africa simply by waving the wand of 'innovation'. Nor are there tech-based solutions to the political, policy and governance issues that are inseparable from refashioning whole economies and societies.

Innovation vocabularies in the African development context have a negative dimension. By placing emphasis on hopes for tech-based quick-fixes to enduring developmental challenges that require conventional reform efforts, the turn to 'innovation' rhetoric might in fact represent a form of fatigue.

If so, championing innovation also smacks of a form of desperation in the face of enduring conventional bottlenecks, barriers, deficiencies and dysfunction.

I am not saying that innovation is easy, only that development is hard...

Jo

ps -- here is a previous post, reflecting on how corruption is a form of 'innovation' -- a new way around systems that are not 'working' in the eyes of some who have their own motives. How does a society capture the evident social capital that fosters innovative corruption and harness it for pro-social outcomes?

Business and Africa's development: an agenda

Pre-Autumn Oxford, and this week a new grad student moved in next door. So this post gets nerdy on the nexus of responsible business and responsive government in African societies.

The topic is big, but if this blog's themes were translated into a research agenda, what might be some principal questions?

I try below to list 10 hypothetical thesis research topics. They are not the 10 biggest questions around 'Africa Rising' generally. Partly this reflects an implicit call on what issues relate to a public role for the private sector, and which are firmly matters for government only: this blog is not about public policy in general. Many of the issues affecting Africa's trajectory are global ones even if they have important localised impacts, from climate change to negotiations on trade barriers.

Instead the list is an exercise in indulgence were I to be one of these new post-grads choosing a topic.

You will notice that some of them are essentially diagnostic: where are we now? There's a reason for that. Working on medium- and longer-term upside scenarios for Africa's unlocked potential -- generally or by sector -- is very interesting work. There is quite a bit of it, and every few months more glossy reports. Yet the trick to such projections is basing them in accurate stock of where things are now. The paucity or unreliability of data make this no easy task -- as Morten Jerven has continued to show.

Taking stock, deciding baselines, and building scenarios requires, of course, asking the right questions. So does the task of imagining the 'upside' -- what does it comprise, what does it mean to conceptualise steady growth that is inclusive and sustainable?

Well, here are 10 topics. They are not necessarily in order of priority. They are framed brief and broad as research questions, albeit ones with a degree of abstraction (macro-level) and with a heavy policy rather than academic or conceptual dimension.

1. 'Inclusive growth': What is in fact happening to income inequality in the region's major economies -- is there any role for business on this issue in fast-changing markets, or is its social responsibility only to grow? More generally, how can tax system design in African conditions best balance private sector incentives with public goods imperatives, and how do we institutionalise appropriate public-private dialogue on tax issues?

2. 'Africapitalism': Is there any evidence of an emerging identifiable 'African' model of private enterprise with smoother edges in terms of sustainability and social + environmental impact, a model consistent with prevailing political ideas of the developmental state ... or is 'Africapitalism' just a neat phrase with no real content, in economies whose structural patterns are well entrenched?

3. 'Business and development': Policymakers valorise small and medium enterprises, but what do we really know about their impact on job-creation and poverty-reduction in Africa? Assuming we know this, what can realistically be done about financing, regulatory and other obstacles to local business creation and continuation on the continent -- how can donors, lenders and big business help?

4. 'Business for development': Related to 3, in what ways can systematically engaging bigger business in the sustainable development and inclusive growth agenda help, including by linking informal or smaller-scale actors into bigger value-chains? Why is this proving so hard? Where has rhetoric on private sector engagement yielded significant results capable of sustaining replicable models?

5. 'Innovation: nothing new?': Mobiles (and related platforms) have had a significant impact in Africa, including in ways that address or leapfrog altogether some stubborn development bottlenecks. This continues to spurn a lot of hype about Africa's 'digital lions' and the transformative potential of the internet in African economies. Yet what evidence is there about links between private consumption or public investment in / of ICTs and significant change in core areas of the economy such as agriculture? How might the internet/digital/knowledge economy prove truly transformative? Or is the current donor buzzing around innovation and ICT only going to prove a distraction from education and skills issues and from addressing some basic infrastructural, policy and regulatory barriers to growth in traditional sectors?

6. 'New investors': What evidence is there that Chinese and other investors have an inferior social or environmental footprint in Africa relative to other (Western) firms? On the basis of this, what scope still exists to shape 'new' investors' approaches in ways that promote ideals around sustainability, good governance and human rights?

7. 'Future-proofing cities': In what ways are business and governments (including sub-national governments) working together to address service-provision and other shared issues in Africa's more significant fast-growing urban areas? What can be done to scale-up some of these initiatives, and how do they relate to broader national and donor development strategies, including in terms of being coherent with rural development issues?

8. 'Public-private partnerships': What pro-development role do PPPs really have to play, what is their record of success, why is there reluctance on either side, what could be done to ensure they meet the potential often attributed to them? In particular, recent high-level summits have called for innovative public-private financing mechanisms to 'share risks while maximising financial returns alongside development impact' (a tall order ...): what models work / might work, what can be done to ensure they're taken up especially for public infrastructure funding?

9. 'Farming fundamentals': Perhaps I am biased, but it seems to me the focus on Africa's urban consumer classes, youth demographic, urban labour surplus, manufacturing potential (etc) is still wide of the main mark. That mark is agriculture, and related value-adding services and industries. What does the last decade really tell us about the scope for private investment in these (very diverse) sectors to have significant developmental impact, in particular through bringing in smallholders?

10. 'Fragility and prosperity': in what specific ways does donor and government policy towards private sector development or engagement require adaptation for countries and areas affected by fragility, conflict and violence? How do we attract reputable firms to risky places? Does major investment necessarily increase human security in fragile regions, where might it have had the opposite effect?

There are any number of other questions and re-framing of the listed ones. There it is. The potential and problems relating to Africa's women and girls mean that the listed things could benefit from a gender dimension.

Responsible policymakers and investors would be asking essentially the same questions about the nature of the continent's growth path: one question underlying all those on the list is how to foster responsibility and attention to longer-term horizons within government and business. That challenge is hardly unique to Africa.

Jo

WEF Africa: leapfrogs and left-behinds

One can seek short-cuts on long-term problems. But for all the excitement about innovation in Africa, one simply cannot 'leapfrog' all problems.

There is much enthusiasm about the scope for technological innovations and related information-sharing platforms to unlock Africa's growth, development and even democratic potential.

Much of this enthusiasm is justified and good. A hot-off-the-press example of this is a recent post by Michael Hastings on why technology-based 'solutions' for health, education, financial services and other issues provide major reasons for optimism about Africa.

That post relates to this week's World Economic Forum (Africa) in Abuja, Nigeria (7-9 May). 

This event also reveals, and champions, considerable business and social innovation in contemporary Africa in order to resolve / avoid infrastructural and developmental bottlenecks, scale-up markets, reach new consumers, provide new services, and so on. Some of the innovation is in terms of new forms of relationships (principally between business and government) for achieving inclusive, sustained and sustainable growth in Africa. Some of these are innovative relationships around innovative technologies, such as increasing government transparency through making more public documents available online.

As said, this is very well and good: may a thousand million centers of energy and daring send ripples of hope and waves of green, inclusive growth through the continent. I do not say this sarcastically. For example, one solution to Africa's energy poverty (and one with considerable other benefits, including in carbon footprint terms) is the growth in off-grid localised generation and distribution networks.

Yet the mini-grid trend is itself indicative of an issue that all the WEF-style discussion of innovation, entrepreneurship, and leapfrogging cannot and should not obscure.

This is that businesses and enterprises of all sizes in Africa are, like its individuals and families, typically compelled to be innovative in many respects because of poor state capacity to provide basic public goods and services. My former Oxford Analytica colleague Hannah Waddilove remarked this week that what can be seen positively as 'entrepreneurship' for example in providing bottled water for retail is also indicative of the failure of state service-provision.

Previous posts have noted that the scope for public-private cooperation in meeting the development agenda is unrealised, as is the untapped potential for private provision of public goods in Africa. However, a theme of these posts has also been that the state still very much matters for long-term sustainable development in Africa, perhaps more than ever.

In this sense, innovation that by-passes state incapacity may be imperative, welcome, or inevitable. Yet it creates a risk that short-cuts and leapfrogs -- valorised as 'innovation' -- might have a long-term negative effect. They might result in undermining the capacity or incentive for the state to respond to, and provide for, its citizens. If we tie progress to innovation that has a primarily commercial orientation but do not 'innovate' to link this in to building more capable, responsive states, many people might be left behind (even more) when the leaping begins.

Jo

See this previous post on 'corruption as innovation' in the context of state incapacity and bottlenecks.

See these pieces on inclusive growth (here), and a post-WEF Davos one on inequality and risk in Africa's growth path (here).

See too a WEF-related post I wrote this week on the 'African Arguments' platform (Royal Africa Society), here.

Private innovation, public goods: leapfrogs, short-cuts and pragmatic principles

If they both reflect a kind of innovation, how do we distinguish 'corruption' from 'entrepreneurship' -- and encourage only the latter?

In my day job covering contemporary Africa's political economy, two distinct narratives have consistently high prominence -- and I wonder about the link between them.

One is the hugely debilitating effect of various forms of corruption -- manipulating public goods and processes for private ends. The other is the much-praised capacity of individuals and groups to make ends meet (or even multiply) despite poor or problematic public services and systems; one constantly encounters anecdotes about how this is a continent filled with highly innovative and enterprising people whose irrepressible spirit of commerce and exchange holds great promise whatever the state does or fails to do.(*)

How can we see these as part of the same issue, not as unrelated parallel stories?

Is it possible to find -- in admiring the ingenuity involved in some corrupt or illicit practices -- some silver lining about the scope for more efficient, effective or legitimate relations between those in public office and private firms or individuals? (**) Can the same spirit-of-enterprise that sometimes manifests as corruption be harnessed to increase not constrict public choice? Can the particularised trust that enables corrupt relationships be seen as the same raw material from which one can envisage a richer reservoir of generalised trust in which greater and wider prosperity is possible for more people?

Most literature focusses on how corruption stifles innovation because, for instance, it undermines trust in how a partial state might treat the fruits of any enterprise (see for example Mauro 1995; Mbaku 2007; Anokhin and Schulze 2009; cf. Mironov 2005). But what if at least some of the same creative thinking that goes into corrupt practices is from the same pool or resource of spirit-of-enterprise that might be capable of finding useful good short-cuts or leapfrogs that make government more efficient and responsive and better able to serve the wider public interest? (***)

Thus how do we foster 'good' creativity and innovation by private entities and individuals not just in commerce or social services but also in designing or influencing the provision of public goods such as security, public financial integrity or the rule of law?

You will see that I have no idea! This post involves fumbling about in the hope of stumbling upon the beginnings of a theory that manifests principled pragmatism and looks for ways to see how transactions and relationships that appear illicit and damaging may have lessons in terms of regulating the interface of private interests and public authorisation, which is where corruption occurs. If I find it I will also find a neater name than 'Governance-by-outcome-not-process'. It is in procedures that opportunities for corruption lie; those who seek to short-cut such procedures (for nefarious or even just frustrated reasons) may be telling us something about designing institutions and regulations to minimise opportunities for rent-seeking by officials.

There are familiar balances involved in idealising forms of regulation and governance: for example, one wants public servants to be responsive and pragmatic (accessible and clean), but not too responsive and malleable (accessible but corrupt). Moreover, of course, not all the regulation that matters or works comes from the state. We talk about fostering 'bottom-up' initiatives, but are also typically sceptical about those that come from non-state sources.

But I don't mean this -- rather I mean being prepared to accept (a) that some corruption fosters growth or distribution, or indicates that formal licit approval is too hard for some sectors of the public (ie, corruption could represent a reaction to bad policy, a sign that governance is not working or is requiring anti-social short-cut actions which could instead be investigated and the innovative approaches directed towards pro-social outcomes); (b) some corruption nodes indicate bureaucratic bottlenecks that simply should be relaxed or removed (rather than 'strengthened' by anti-corruption measures) and (c) corruption sometimes indicates that individual actors have found a more efficient route than policymakers prescribe, which may have virtuous implications (see for example Leff 1964; Bailey 1966).

This is all rather undercooked, but represents an attempt to think about how to design systems of governance and development-promotion that instead of requiring innovation in terms of ways to get around regulations, designs regulations that stimulate 'good' short-cuts and leapfrogs that help point officials towards making government both responsive and responsible. I stand ready to be accused of rank naivety...

Jo

* = Often, of course, it is less romantic things at work, and what is cast as 'enterprising' is instead just about survival or subsistence; that is, adversity and necessity -- not just curiosity or the promise of commercial gain -- are a major source of inventiveness.

** = The other point to note is that much of the most damaging corruption (in Africa and around the world) is not particularly innovative: often it is just a blunt and blatant act of taking (or withholding, for example of tax obligations) that does not require strong entrepreneurial skills to find ways around barriers, it only requires weak systems of oversight and accountability. Moreover, from the perspective of those marginalised from their proceeds or benefits, formalised systems of governance may be seen simply as private enrichment systems dressed up as public order. It depends on one's view of the legitimacy of the state and its processes in any one setting.

*** = A somewhat related question is the opportunity cost of anti-corruption and accountability systems -- some (eg Anechiarico and Jacobs 1996) argue, in effect, that effort should rather be directed to supporting 'good' creativity in governance rather than trying to stamp out 'bad' creativity...