This is a short sharp summary of how biases affect AI design and what to do about them, reaching the conclusion that government oversight is essential (though not, of course, sufficient). There are interesting parallels with Google’s in house rules for working on AI, so worth reading the two together.
Including this link is slightly indulgently self-referential, for reasons which will be apparent to anybody who reads the last fifth of this article – but there is value in the first four fifths regardless of that. More generally, aggregators of good things from across the web are to be welcomed, and others will get mentioned here from time to time. As Benedict Evans once put it, “All curation grows until it requires search. All search grows until it requires curation.” This is where curation is celebrated.
Not at first sight an obvious entry for inclusion here, but indirectly very relevant to thinking about the impact of change on large organisations. The size and complexity of organisations, it is argued, is a function of the relative costs of internal control and market transactions – whether it’s cheaper to make or to buy. That is in turn in part a function of the cost and availability of information. And so the conclusion is reached that digitally-based organisations will be smaller and fleeter of foot than their predecessors and that large ungainly organisations of the pre-internet era are doomed to extinction, with government in their current form being among those whose demise is inevitable.
In some circles, that conclusion is seen as an obvious one. Farrell’s argument is that something critical is missing from the analysis, the self-interest of individuals, and that when that is factored in, the picture looks very different. That matters not to justify slow change in large organisations, but to explain that power relations are a critical part of understanding the overall situation. And as long as they remain unequal, the pop up employer is likely to remain intriguing at the margins, rather than central to how stuff gets done.
Organisations exist to get things done. They are necessary because they solve problems ranging from communication and co-ordination to moral hazard. But in principle, if we could find other ways of solving those problems, we wouldn’t need organisations any more – at least of the traditional kind. This article gives some examples of where that is starting to happen, by assembling not just project teams, but project organisations, created to meet a specific need and disbanded as soon as that need is met.
The examples given show that it can work, but they don’t and can’t show that it can scale, and there may be good reasons to think that it can’t. But that doesn’t stop the idea being a good challenge, particularly to those in organisations which don’t tend to think about organisational change in quite those terms.
This is a powerfully argued manifesto for service design in a digital age. It’s not good enough just to put a digital layer on top of organisational and service architectures which predate the internet: that results in services which are unstable and unscalable and so unsatisfactory and ultimately unviable.
There is much here which is persuasive and important, but it underplays a critical part of the overall picture. The new organisations and approaches Ben lauds have in common that they are all comfortable with the rhetoric of sharing, but they are all completely conventional market organisations whose relationship with their customers is one-dimensionally transactional. Amazon is the one partial exception to that, but even they do more to fit the institutional model of the pre-internet firm than to challenge it. Public services are – or should be – part of wider conversations. To define them solely through their transactions is to miss something essential.
“All organisations are perfectly designed to get the results that they get” (Arthur Jones, probably)
What government organisations get is hierarchy, slow and often unresponsive decision cycles and a sense that government is done to people, rather than with them, still less by them. There are – or were – some very real strengths in a Weberian rationalist bureaucracy, but Weber was writing a century ago for a very different world. Adaptation is as crucial for organisational evolution as it is for natural evolution – without it, organisations become less and less well fitted to their environment and eventually fail. But that failure can be long drawn out and painful to all concerned.
This essay is about spotting – and encouraging – new approaches to public administration, better suited to new requirements and new contexts.
What’s the best way to arrange the nearly 3,000 names on a memorial to the victims of 9/11 to maximise the representation of real world connectedness?
Starting with that arresting example, this intriguing essay argues that collection, computation and representation of data all form part of a system, and that it is easy for things to go wrong when the parts of that system are not well integrated. Focus on algorithms and the distortions they can introduce is important – but so is understanding the weaknesses and limitations of the underlying data and the ways in which the consequences can be misunderstood and misrepresented.
A useful ladder of intervention for policy makers, which refreshingly treats legislation as the last possible intervention, not the first. As with other Policy Lab products, its value comes from prompting better questions rather than from providing direct answers, so the ladder may seem more ordered than the real world of policy development tends to be – which doesn’t stop it being good food for thought.
But the post doesn’t really answer the question very firmly posed by its title, Paul Maltby’s post may be a better place to start for that.
If machine learning is not the same as human learning, and if machine learning can end encoding the weaknesses of human decision making as much as its strengths, perhaps we need some smarter ways of doing AI. That’s the premise for a new Google initiative on what they are calling human centred machine learning, which seems to involve bringing in more of the insights and approaches of human-centred design together with a more sophisticated understanding of what counts as a well-functioning AI system – including recognising the importance of both Type I and Type II errors.
Artificial intelligence is more artificial than we like to think. The idea that computers are like very simple human brains has been dominant pretty much since the dawn of computing. But it is critically important not to be trapped by the metaphors we use: the ways in which general purpose computers are not like human brains are far more significant than the ways in which they are. It follows that machine learning is not like human learning; and we should not mistake the things such a system does as simply a faster and cheaper version of what a human would do.
What counts as minimum viable competence for public servants (or indeed anybody else) in the modern age? This post is a robust challenge to the false modesty of digital incompetence, which is heard much less often than it used to be, but is still too often not far below the surface – and still reinforced by working environments which have yet to break free of the twentieth century.
Most people think that if they had a basic income, they would do something creative and constructive with it. Most people think that if other people had a basic income, they would laze around. Understanding which is right – and understanding whether there is any meaningful distinction between them – is an important element in planning for the future of work.
— RSA Events (@RSAEvents) July 7, 2017
Jakob Nielsen has been writing about usability since the dawn of the web. His approaches seemed to go out of fashion for a while, but there has always been a lot of evidence-based common sense in his approach. This post uses OECD data to demonstrate just how limited digital skills are, even in the most advanced countries and draws out the critical point that people who design, build, or even vaguely think about online sites and services are massively unrepresentative.
It is though curious – presumably as a consequence of the way the research was done – that the test tasks described are very work focused. It would be interesting to know if task context and familiarity at least partly countered task complexity.
Continuing the theme of how digital and policy people are more powerful when aligned than each is when operating separately, this post has lots of insights about how to make that happen in practice. Empathy, vulnerability and culture are as important as communication and collaboration. And as the company Harry leads has deservedly has just won an award for being the best digital SME workplace, his thoughts come with some authority.
One of the misunderstanding which often crops up between people who approach problems from a digital service design perspective and those who come more from a traditional policy development perspective is around transactional services – whether fundamentally we are designing services for users or outcomes for society. This post comes from a very different perspective of deliberative policy development, drawing out very clearly that people see their role as citizen more broadly than their role as consumer, even when they are being both simultaneously.
Big organisations have things which aren’t quite anybody’s job to do, so they don’t get done. Small organisations tend to solve that problem partly by making everybody’s role much more fluid, and partly by reducing the overheads of the collective action problem. Big organisations find that hard because they manage complexity through structure – which is fine for things which go with the grain of the structure but can be very difficult for things which cut across it. That can lead to situations where – in a neat phrase from this post – ‘the indecision is final’. The solution advocated here is a simple one: if spaces are unoccupied, occupy them.
If we are going to get smarter about how things get done in government, and in particular are serious about melding the strengths of the different tribes into a whole which is stronger than its parts, how do we make that happen.
This post lists seven principles for building a one team government community – and then invites people to sign up to join that community.
Does the power of big data combined with location awareness result in our being supported by butlers or harassed by stalkers? There’s a fine line (or perhaps not such a fine line) between being helpful and being intrusive. Quite where it will be drawn is a function of commercial incentives, consumer responses and legal constraints (not least the new GDPR). In the public sector, the balance of those forces may well be different, but versions of the same factors will be in play. All of that, of course, is ultimately based on how we answer the question of whose data it is in the first place and whether we will switch much more to sharing state information rather than the underlying data.
Is user-centred digital government unstoppable? At one level the answer must be yes, we live in an increasingly digital environment and government is not and cannot be immune to that. Digital is well on the way to just being the substrate for how things get done. But just as twentieth century governments could be good at managing paper without necessarily being good at using it to communicate clearly and efficiently, so doing things digitally does not immediately imply doing them well. This post argues the positive case, that we are beyond the tipping point where ideas about rapid, responsive service design have a sufficient life – and strength – of their own to transcend the vagaries of individual leaders and services.
This is a good simple and succinct description of what user research is and why it matters. It draws out the critical point that user wants are not necessarily a good indicator of user needs, not least because what matters in the end is not (or is not just) the immediate interaction, but the success of the underlying service in delivering the right outcomes.