Data is the fuel for AI, so let’s ensure we get the ethics right
Data is the propulsive energy behind the fourth industrial revolution – playing the same role as coal, oil and electricity did in the previous revolutions.
Artificial intelligence (AI) is the new engine, just as the railways connected north and south, roads and cars connected our cities, and electrification lit up our homes and allowed domestic appliances to free women to go to work.
As when agriculture was mechanised or mass production became mainstream, some jobs will be lost, even as other jobs will be generated. But to harness AI’s job-generating power, we need to ensure that it has the flows of data crucial to its success.
Read more: Questioning AI ethics doesn’t make you a gloomy Luddite
Without data, there is no AI. Since the potential for the abuse of data is huge, solving the ethical conundrums surrounding this is becoming the number one AI issue.
I welcome the emphasis on data and ethics in the industrial strategy published last week. With the commitment to create a Centre for Data Ethics and Innovation, the government is putting ethics at the heart of how this country develops and harnesses the power of AI and automation.
It is important to remember that technology is essentially neutral. Robots do not decide to go after anyone’s jobs – it’s the people who write the algorithms and shape this brave new world.
We need more public trust to free our data to generate the growth and welfare. Data sharing is a social benefit. It is the way value is created in the AI era.
In fact, data sharing should be part of the “social contract” between each person and their society. Thus, just as we require a driving licence and car insurance to drive a car, so users of a driverless car should be required to share their data for everyone’s collective good.
For this to happen, the Big Innovation Centre argues that the government needs to establish a data charter with stakeholders on what can be done with personal and business data, so that everyone will know how their data is used and not used.
This will in turn increase trust and create incentives to allow data to be shared. The US is already moving ahead with “user rights”. Britain needs to introduce a “fair use” principle too.
The UK could lead this entire move towards the global rules, norms and standards in the governance of personal data, boosting our exports, and getting to grips on the trade-related aspects of AI.
We didn’t lead the e-commerce world, but could we lead the world of AI-commerce via chatbots, the internet of things, data exchanges, and blockchain. Our professional and public services could be newly dynamised.
This is an opportunity that last week’s industrial strategy did not seize. However, there were other positive steps made. These include the strategy’s commitment to create an AI council – short of our call for a minister for AI, but is an advance nonetheless.
The strategy’s commitment to investment in talent, skills, and infrastructure is also laudable, in particular earmarking £250m per annum by 2020-21 for higher education innovation funding. We must ensure that we have enough AI teachers teaching the teachers.
The public sector must take a lead on the fourth industrial revolution. Public services in health, energy, finance and social housing have a vast opportunity to create markets at scale, boosting innovation and entrepreneurship and readiness for export from our growing businesses.
Finally, AI and big data should also inform the implementation of the industrial strategy. The focus should not just be on addressing stagnant productivity, but on fundamentally changing the way we work.
Sadly, the big ideological mistake here is to apply outdated methods to twenty-first century industrial performance.
Instead, we should be identifying the challenges ahead: we need transparent markets for intellectual property, entire supply chains made innovation ready, and every system supporting entrepreneurship – from talent to finance – in all our regions made mutually reinforcing.
To find the answers to these urgent questions, there needs to be a big data platform with diagnostic tools to identify the country’s strengths. That’s why we at the Big Innovation Centre set up a national innovation audit, to give us a real-time online assessment of the skills base and innovation capabilities of the UK’s regions.
We look forward to using our insights to help inform government policy and direct support for this vital area in the future.
But this industrial strategy, foregrounding how public activism can stimulate AI, is a welcome and long overdue beginning.
Read more: Automation will be a bigger shift than industrialisation