Lessons from the implementation and evaluation of the Greater London Authority "Grow with AI" programme
Anna Valero, Capucine Riom, Juliana Oliveira Cunha
The “Grow with AI” programme was a randomised controlled trial that sought to test barriers to adoption of “tried-and-tested” Artificial Intelligence (AI) technologies in London’s retail and hospitality sectors. Light-touch market-convening events were compared with more targeted caseworker support and a voucher, and with a control group that received only written information. Due to recruitment and implementation challenges, the trial did not obtain a large enough sample to enable robust econometric evaluation. However, adoption-related activity was observed to be higher in the two treatment groups than in the control group. Firms in all groups reported having a better understanding of the costs and benefits by the end of the trial, but attitudes towards AI worsened. The difficulties with recruitment and take-up suggest that the support offered in this trial did not meet the priorities of the businesses targeted. Interviews with a sample of participants highlighted that the AI technologies being promoted were not as “shovel-ready” as had been expected. However, interview responses also suggest that SMEs understand the importance of technological upgrading, and are willing to adopt new tools as long as they prove to be cost-effective.
Keywords: SMEs, technology adoption, artificial intelligence