Just enough Internet: Why public service Internet should be a model of restraint
This is more or less the text of a talk I gave at MozFest House, for the BBC Research & Development and PublicSpaces conference: Public Spaces, Private Data: can we build a better internet?
The UK needs a new social contract for the digital age.
It’s time to renegotiate the tacit agreement between the people, the market and the state to take account of the ways that data and technology have nudged behaviours and norms and changed the underlying infrastructure of everyday life.
Digital public services have a vital role to play in this, a different space to hold, distinct from business or regulatory activity and it is time to reimagine and restate what a good public service looks like for 2020 and beyond, when the most important work to do is to take care of people and the planet.
Creating this new social contract requires new norms, new institutions, new laws – and a stronger public sector, that is prepared to hold a different space: uphold different kinds of progress, measure different kinds of success, and celebrate a different kind of innovation to that of big technology platforms and venture capitalists.
The opportunity of the public sector, after all, is to do things for the public.
Writing this talk has made me nostalgic for the ad campaign that went along with the launch of BBC4 in the early 2000s. It was perhaps a little pompous and silly, but it expressed something unique that perhaps only the BBC can do, and which is perhaps needed more now than ever: it showed a public service as created a place to think.
And I’m going to talk this morning about the importance of the public sector creating its own definition of what good technology and responsible innovation look like.
It is time to stop following big business, and think back to the social contract. How can digital public services be part of a strong and resilient society?
If the answer from business is to add data to everything, then perhaps the public sector should be more measured. Perhaps a public service Internet should be a model of restraint; a counterweight to the ballast of peak data, that knows and provides just enough.
I’m going to start by talking a little about measurement, and then cover ways of working. And I’ll finish by talking about some new decisions I’d like you to think about.
I worked on the BBC website in the early 2000s. It was a fantastically exciting time for all sorts of reasons, but one of them was because more and more people were getting online every day. We’d get a spreadsheet of stats for the week every Friday afternoon, and it was a hockey stick that just kept going up and up and up.
It made us feel as if we were magical – that we were making new kinds of irresistible content. But we were also lucky: the number of people online in the UK more than doubled between 2000 and 2003, from 27% of the population to 65%.
And all the way back then, the great big KPI was page views. The big exciting hockey stick was very intoxicating and exciting, and the spreadsheet that went around every week was ranked in numerical order. The BBC “category” with the most page views was the winner.
And so my colleague had the idea of putting webcams that refreshed every five seconds on as many pages as possible.
And we did really well, and got loads of page views, because the pages refreshed themselves – but it’s probably not going to surprise you particularly if I say those webcams weren’t quite the pinnacle of public service value.
One of them was meant to be in a haunted house, so mostly all it showed was a black square – a camera taking and uploading a photo of a darkened room every five seconds.
It was quite literally the abyss.
But the numbers kept going up, so it got recommissioned every year for several years.
And that same spirit of quantity over quality has been a driving force of the attention economy. More is more is more.
It’s almost three years since YouTube passed the milestone of streaming a billion hours of content every day. As the super-excited announcement post puts it:
If you were to sit and watch a billion hours of YouTube, it would take you over 100,000 years. 100,000 years ago, our ancestors were crafting stone tools and migrating out of Africa while mammoths and mastodons roamed the Earth. If you spent 100,000 years traveling at the speed of light, you could travel from one end of the Milky Way to the other
And that volume is important to YouTube because it drives ad revenue. In 2018, those billions of hours of human eyeball time turned into $116bn in ad revenue.
Meanwhile, Netflix – the most successful entertainment company in the world – describe their competitors as basically all human non-work activities: reading, dinner with friends, enjoying a glass of wine with your partner. I love Netflix, but I don’t imagine it is trying to be good for me. In fact, their vision of the world is one in which those little entertainment pods on the Axiom in Wall-E might not be a work of fiction.
Over on social media, Twitter and Facebook are racing one another to the bottom to define free speech. Both platforms have decided that incendiary political content is “newsworthy”, but perhaps “worthy of news” is not quite what they mean: it’s not that the content is necessarily worthy of comment and analysis, it’s that it gives their platforms importance, relevance, publicity and more and more eyeballs in this tumultuous political world.
These companies are giants and their primary purpose is to deliver value to their shareholders.
Now delivering shareholder value has a place in society, but it is not the role of public services. The NHS, the BBC, our education and benefits systems, our libraries are not here to turn a profit and they do not benefit from infinite resources.
None of these services can or should compete with the platform tech companies. Their mission is not to innovate at all costs but to meet the needs of everyone with excellence.
And the fact that neither NHSX or BBC R&D will send a rocket to Mars this year does not mean they are not innovative. It means they are not in the rocket business.
And if Netflix and YouTube want to keep us stapled to the sofa, then perhaps the BBC should take us – metaphorically or in reality – into the fresh air. Being at the cutting edge of technology does not need to involve putting an X on it and progress does not need to be measured in numbers of automated systems.
Just this weekend, someone from the BBC was tweeting about BBC Sounds, benchmarking its resources against those available in Silicon Valley and positioning it as a home-grown alternative. Finding podcasts is not a problem that only the BBC can solve, but for whatever reason it is very committed to trying.
And I’m going to say something difficult – because frankly hindsight is a wonderful thing. But the totem of UK digital public service is still the Government Digital Service, and at its birth in 2012 the founding team set out to do something revolutionary.
But it was perhaps not revolutionary enough. Tom Loosemoore said at the time, “We wanted to build GDS the way Google built Google and the way Amazon built Amazon” – now, this was a technical claim, not a business one, referring to using Agile and open data and user-centred design, which are still the gold standard for many product teams.
But tax-payer funded public services need different affordances to Amazon and Google; they should not feel magical, they should be data light, they should meet social needs not just user needs; they should work in context, not simply fulfil a job to be done. Sometimes they should be compassionate.
Philip Alston, the UN Rapporteur on Extreme Poverty, published his report just last week into the worst excesses of digital public services, a world of automation, surveillance and targetting. I will not dwell on it here – you should read his report – but Alston warns of the risk of “stumbling zombie-like into a digital welfare dystopia”. His signal recommendation is to avoid technology for its own sake and to double-down on purpose. Specifically he says:
The report recommends that instead of obsessing about fraud, cost savings, sanctions, and market-driven definitions of efficiency, the starting point should be on how welfare budgets could be transformed through technology to ensure a higher standard of living for the vulnerable and disadvantaged.
So what does good look like? Or, to steal a phrase from my friend Alex Mecklenburg, what would be good enough?
It might look like collecting less data. Just what you need at any given time.
It might look like automating with care, and sometimes not at all.
It might look like zooming out from the task at hand, and understanding the wider social consequences of your decisions.
It might look like a policy of Sufficient Technology: just enough to make it work better for the people that count.
Which might also be the best thing for the planet.
It is likely that, outside of the military, a UK public body will never have the staff, resources and computing power to match those of big businesses. As Philip Alston and before him, in their fantastic books, Cathy O’Neil, Safiya Noble and Virginia Eubanks have shown, doing this badly is not good enough.
And it should not be an option.
A badly crafted data-driven service is like a lethal weapon in the wrong hands. For instance, the Metropolitan Police’s Gangs Violence Matrix has been shown by Amnesty International to operate on assumptions about the behaviour and interests of young black men. Boys as young as 12 are put under suspicion of gang membership, based on often inconsistent and often racially biased evidence. Jobcentres, schools and housing associations all have access to the matrix, leading to school expulsions and evictions – scuppering the life chances of already vulnerable young men, based on the music they like or who they follow on social media.
To paraphrase Kate Raworth, our current data economy is extractive, not regenerative. As individuals, we are data sources; we do not have parity with either governments or businesses: and although we have the right, we do not have the time to have our data explained to us, or the resources to aggregate, use and analyse it.
In the private sector, our data has a hard monetary value – for instance, we’re each worth $34 a year to Facebook – but the terms of our relationship to our public sector data have not yet been set.
Decisions about healthcare, transport, credit, job opportunities, child safety, access to bail and more are made on the back of data we may or may not know we have created or given away. How we walk, how we spend, who we like and how often are all inputs to decisions we have little ability to repeal.
Moreover, Mary Meeker predicts that by 2020, only 16% of the world’s data will be structured enough to be useful.
Or to put it another way, 84% of the data collected about people and things is sitting unused. Using up servers, eating electricity. To quote a recent blog post by AI Now Institute:
Researchers Lotfi Belkhir and Ahmed Elmeligi estimate that the tech sector will contribute 3.0–3.6% of global greenhouse emissions by 2020, more than double what the sector produced in 2007 (Belkhir and Elmeligi, 2018). The estimated 2020 global footprint is comparable to that of the aviation industry, and larger than that of Japan, which is the fifth biggest polluter in the world. Data centers will make up 45% of this footprint (up from 33% in 2010) and network infrastructure 24%.
The carbon cost of collecting and storing data no one can use is already a moral issue.
So before you add another field, let alone make a new service, can you be sure it will make enough of a difference to legitimise its impact on the planet?
And can you be sure it will serve the public – all of the public – in a way that aligns to your organisation’s mission and values?
Public services must change as the world changes, but they should develop their own path not simply follow the one created by business.
A public service internet must cut its cloth accordingly; rather than competing with global businesses, it should set its own, unimpeachable standards.
We have not yet done a good job of defining what good digital public service really looks like, of creating digital charters that match up to those of our great institutions, and it is these statements of values and ways of working – rather than any amount of shiny new technology – that will create essential building blocks for the public services of the future.
Just enough Internet might be all the planet can afford in the future. It’s something that digital public services could champion and do differently, and perhaps in time technology businesses might catch up.