UK must avoid complacency and seize AI opportunity


A new report from The Entrepreneurs Network think tank by former Office for AI Adviser Séb Krier sets out eight key recommendations to make the UK the best place in the world for AI innovation.

Artificial intelligence is already solving our most pressing problems. UK founded DeepMind solved protein folding, one of the biggest problems in biology, and more than $14bn was invested in AI drug discovery last year.

The report argues we must not become complacent. While the UK is a world-leader in AI development, other countries are catching up fast. Over 50 countries have or are in the process of implementing national AI strategies, according to data from Stanford University’s AI Index.

Unless the government embarks on an ambitious policy programme, there is a risk the UK will fall behind and be slow to capture the full benefits of AI to transform the economy. If the UK is to continue to produce AI unicorns (private companies with $1bn valuations) such as BenovelentAI, Graphcore and Improbable then new measures to tackle skill shortages, strengthen the research ecosystem, and foster public support for AI are necessary.

The report backs Google’s recent call for the creation of a British Data Cloud to allow researchers to work on socially beneficial AI applications, scrutinise commercial models, and undertake long-term research. This would also incentivise some researchers to stay in academia and mitigate concerns about a hollowing-out of the talent pool for public interest AI research. This matters because high profile departures from AI faculties have negative effects on students’ specialised knowledge, which is a crucial determinant of the success of AI startups.

A national cloud-based infrastructure would provide researchers and students across scientific fields and disciplines with access to compute resources, co-located with publicly-available, AI-ready government and non-government data sets, and a research environment with appropriate educational tools and user support.

Although the UK leads the world in open data, there is still room for improvement. The Government should commit to reviewing all existing non-public datasets and releasing as many as possible. When Transport for London opened up its data, it allowed a number of start-ups like CityMapper to thrive, adding up to £130 million a year to London’s economy. The author suggests this could involve the creation of an open address system which would help reduce costs on small businesses, who have to pay every time a user searches an address from a postcode.

Access to talent will be key to supporting the long-term growth of the UK’s AI sector. But while investment in the AI skills pipeline is necessary and valuable, it will take years to pay off. In the short term, making the UK more attractive to STEM graduates will have the largest impact.

Recent moves to make it easier for highly skilled migrants from across the world to move to the UK are welcome, but insufficient. To support the AI sector directly, the government should actively fund scholarships for international postgraduate students in AI-related fields. In the US, the majority of AI PhDs are born overseas and more than four-fifths stay in the US once their PhD is completed.

The uptake of AI technology will depend in part on the public’s acceptance of algorithmic decision making. Recent scandals such as the A-levels grading fiasco and the Home Office’s visa-sorting algorithm have understandably damaged trust. To prevent future problems the report recommends making the Centre for Date Ethics and Innovation fully independent and using it to focus on and audit the public sector’s use of AI.

Full List of Recommendations:

  1. Create a pool of cloud compute credits for the UK R&D ecosystem.
  2. Upgrade public data infrastructure and open up datasets.
  3. Explore innovation-friendly regulatory markets.
  4. Work closely with the EU to review and improve GDPR.
  5. Lower barriers to immigration and actively attract foreign talent through targeted scholarships.
  6. Shun protectionism and proactively lead global AI governance efforts.
  7. Foster public trust in the public sector’s use of AI by giving the CDEI greater independence and allowing it to audit public sector algorithms.
  8. Ensure the UK’s intellectual property regime is fit for AI by creating an exemption for for-profit data and text mining.

Séb Krier, author of the report and former Head of Regulation at the Office for AI said: “The UK is a global leader in the development and commercialisation of AI, hosting leading companies such as DeepMind and BenevolentAI, but more needs to be done to cement this position.

“The UK can bolster its position by undertaking ambitious and innovative policies, which will strengthen the country’s general infrastructure, its research ecosystem, the private sector’s dynamism, and public sector accountability.

“Amidst a global retreat towards digital nationalism and protectionism, the UK should push for an open and global technology ecosystem without compromising on safety and security.”