Everybody is talking about artificial intelligence (AI).
But it is a divisive issue. Either you’ll agree it will help accelerate medical advances, tackle climate change and could even help us understand animal communication, or you’ll fear it will lead to the end of humanity.
Given that AI already benefits many, should it really be perceived as an existential threat?
I would argue not, especially considering we’re living through a very real climate emergency… but that’s another story.
I am clear that no matter the discussions or reports being published that predict the future of AI, we need less conversation and more action right now.
From an AI perspective, in the last month, the EU has taken the bold step of approving its AI Act, and the US has also seen a blueprint put forward in the Senate. In the UK, though, we are very much still in the discussion phase.
Make no mistake, though; swift action is needed to ensure we don’t fall behind as a nation when it comes to AI regulation.
Not acting could be catastrophic and we must take learnings from the Online Safety Bill, which has been slow in its development. All the while, social media has continued to grow and evolve – with some platforms changing their route to market and operational practices completely.
The pace of regular technology change is quick – and the advancement of AI models appears to be even faster. In parallel, organisations up and down the country and worldwide want to understand better how AI can help their operations. Still, they’re crying out for the guardrails of regulation and frameworks.
Professor Shannon Vallor is co-director of the Bridging Responsible AI Divides (BRAID) programme at the University of Edinburgh. During a talk at Turing Fest, she outlined two choices for organisations.
They could either move fast and break things with catastrophic effects (as we have seen with social media), or they can innovate boldly but responsibly (seen in aviation and biosciences, among other sectors).
Right now, the fear attached to AI appears to be down to worries about getting it wrong.
We need some level of regulation to fuel further adoption and responsible innovation of AI – and along with it, develop both the knowledge of those building AI systems and improve the understanding of its users.
I recently returned from London Tech Week, where I heard Prime Minister Rishi Sunak speak about the opportunity AI could bring to the UK. He spoke not only about the technology improving the economic standing of the country’s SMEs, but also about helping improve public services and supporting the advancement of life sciences.
We know that AI can help many sectors. For example, manufacturing and renewables - although not mentioned directly by Sunak - could also hugely benefit from the broader use of AI.
With general elections on the horizon at some point, too, there is an even greater need for regulation to move swiftly or risk getting caught up in the noise of another electoral battle.
Regulations for AI must be flexible to accommodate its ongoing evolution. There is no point in creating rigid guidance that will be out of date before it is published. This flexibility enables technology to continue to evolve and new use cases to emerge.
The development of such regulation requires expert guidance. Having specialists from diverse backgrounds, sectors and industries is vital if any regulation is to have the scope required to guide and benefit organisations and individuals for years to come.
We are in a prime position to support this – it was encouraging to have so many insightful conversations with people who were intrigued by The Data Lab’s model, which brings academia, industry and the public sector together and could see it working internationally.
At the core, it is crucial that we break free from the cycle of stagnant discussions about AI.
We have the perfect opportunity to maximise the future use of AI to benefit humanity. Let’s take the opportunity and ensure that AI’s potential is fully realised.
Brian Hills is chief executive of The Data Lab
