Capital, politics and techno-ethics – an unwholesome threesome

When political and business leaders pontificate on ethics, as England’s acting prime minister Theresa May has done today in Davos, you should immediately check for your purse.

Davos, aka the World Economic Forum, is an annual excuse in mutual backscratching and deal-making among the world’s uber-rich (pun intended) and politicians, democratic or otherwise. The democratic legitimacy of politicians, and financial cleanliness of business execs, is largely immaterial given the scale and scope of the wealth and politicking. Still, the world habitually focuses on this alpine Spectacle, as decisions made in Davos affect the lives of billions.

Business folk and policymakers who for the most part couldn’t distinguish a hard drive from a golfing action are suddenly speaking with feigned authority on artificial intelligence (AI) and technology ethics, explaining away the problems of social media, app-based business models, and much else besides. May, for example, talks of improving business regulation and employment law, and this when the prime minister’s political string-pullers regard both as anathema. May has on many occasions spoken of these subjects and promised action, but the policy discussions never seem to progress beyond the green paper stage.

Talking of getting rid of the green crap, another of May’s interventions concerns plastics waste. Here she prostrates herself before the Great Sage David Attenborough, speaks out strongly by a Thameside bog, and then deflects the discussion by turning to more mundane matters. It works every time, with many of us forgetting that May’s government is attempting to block the ambitious recycling targets in the European Union’s 25-year environment plan.

Political machinations notwithstanding, the thing about technology ethics is that the kind of problems referred to third-handedly in recent days cannot be ignored in the long term, as to do so would work against the political imperatives of power and control. If one cuts away the hyperbole, AI continues to make progress, but in the current context the term is misleading. We are not talking about machine ‘intelligence’ in creative, self-replicating terms, but rather sophisticated, task-specific algorithms designed to handle large social datasets (so-called ‘big data’) for commercial and intelligence-gathering purposes.

Environmental policy and vulture capitalism aside, there is an important discussion to be had about AI and journalism. We are only beginning to scratch the surface of what is possible with data journalism, and the activity raises a number of ethical questions.

For one thing, with far more data out there than is possible to analyse in detail by human hands, the use of algorithms to reduce it into a form manageable by journalists is unavoidable. That means coding data manipulating and discarding decision making into complex software outwith the context of particular journalistic investigations, and the result is Bayesian inference run riot. This kind of statistical methodology is powerful, as I found when using it as a space scientist, but it’s not without problems.

The NUJ’s Ethics Council, of which I am a member, is currently looking at AI and journalism. If re-elected to council next month, I hope to contribute further to what promises to be a challenging discussion.