Digital Catapult, the not-for-profit body that focuses on helping UK businesses to scale up, has selected the first start-ups to join its Machine Intelligence Garage program.
The Machine Intelligence Garage Program at Digital Catapult aims to support the ambition of making the UK a global centre for artificial intelligence (AI) development – and several start-ups have just been signed up for the program.
The companies selected are digital manufacturing start-ups Intellisense and Predina; digital health start-ups Cambridge Bio-Augmentation Systems and GTN; and the natural language processing start-up Bloomsbury AI.
“Members of the first cohort have been carefully selected for their innovative use cases in AI, ethical values and ambition,” said Dr. Marko Balabanovic, the chief technology officer at Digital Catapult.
“We’re confident that their journey with the Machine Intelligence Garage will provide them with the support, expertise and compute resource they need to reach their full potential,” he added.
Early access to machine intelligence
Simon Knowles, CTO and co-founder of AI chip company Graphcore said the start-ups will get early access to his company’s Intelligence Processing Unit, which was been designed for machine intelligence.
“Its unique architecture means developers can run current mchine learning models orders of magnitude faster,” he said.
“More importantly, it lets AI researchers undertake entirely new types of work, not possible using current technologies, to drive the next great breakthroughs in general machine intelligence,” he added.
Azeem Azhar, senior advisor at professional services company Accenture, a partner of Digital Catapult, said the program would help “ to level the playing field” by giving start-ups in the UK access to some of the best resources and expertise.
“I’m excited to see where the programme’s first cohorts are by the end of the year,” he said.
High barriers to entry
The program aims to give small and medium-sized enterprises (SMEs) access to cloud-based and physical computational power for AI. According to Digital Catapult, deep-learning techniques incur extortionately high computational costs – a single training run for a ML system can cost upwards of £10,000. This can be a serious barrier for UK innovators and researchers.
Start-ups that are developing products or services that use ML or AI can apply to join the Machine Intelligence Garage program. The last Open Call went live on 23 January 2018, with applications to roll every six weeks thereafter.
Applications will be assessed based on a number of criteria including strength of idea, technical implementation plan, availability of data, ethical use of data, and the immediacy of the need for computation power.