Blair shared his views on the future of AI in a thought-provoking speech delivered at the opening of the Tony Blair Institute’s (TBI) Future of Britain conference early last week. Though attempting to portray a brighter, more technologically advanced future, Blair first forecast a grim alternative marked by dire economic and social challenges the likes of which are currently highlighted by the "triple whammy" of high taxes, heavy debt, and poor public service outcomes.

Blair proposes jumping head first into a world not just aided by AI, but fully beholden to and dominated by it - or at least jumping into a version of this world he believes already exists. Referencing the significant decline in economic growth and productivity over the past decade, he proposed that Artificial Intelligence (AI) is the “game changer” that could “turbo charge” Britain’s economy with an impact “at least as big as that of the Industrial Revolution.” 

Blair asserted that “with the technologies and the digital infrastructure we have today, up to £40 billion can be saved each year with the technology as it exists now.” More than this, he proposes that there would be cumulative savings of around £150 billion by the close of the annual fiscal year’s second term. These numbers are not only dubious due to their size, but also because in his organisation’s paper, “Governing in the Age of AI: A New Model to Transform the State”, released ahead of their conference, TBI’s economists forecast a net annual savings of only £34 billion by 2034.

The difference between TBI’s speech and report is mirrored by the difference between their understanding of the current power of AI and what is truly available. Understandably, with AI washing raging rampant, Blair and co have fallen victim to thinking that AI and the digital infrastructure we have today will be able to surrogate for and advance beyond the current processes we have in place, with little to no work.

On the same day as the Future of Britain conference, Blair’s own think-tank published a separate report: The Economic Case for Reimagining the State. In this they suggest that by adopting AI across the UK public sector they could save approximately 20-40% of workforce hours. Repetition is key to Blair’s statements as he reiterated this exact quote in his speeches at the Future of Britain conference, on BBC Radio 4’s programme ‘Today’, and in discussion with William Hague on Governing in the Age of AI. Repetition, however, is precisely what limits today’s AI tech and more specifically the technology such as ChatGPT that Blair has mentioned using. 

These large language models (LLM’s) are not the same as artificial general intelligence and have not been trained to complete the tasks in question- they are not capable of reaching the nuanced understanding and reasoning that is necessary for it to progress beyond generating text and repeating it back to us as they currently do. The possibility of training these models to do so is a hotly debated topic in its own right, but the time it would take to develop and roll out this tech moves it further and further away from the technology Blair and co insist we have right now - tech he already calls “personalised and human-centric”.

Speaking to the over-exaggeration of AI’s ability to replace humans is prominent MIT economist, Daron Acemoglu, who argues the tasks Blair wants AI to do “are multi-faceted and require real-world interaction, which AI won’t be able to materially improve anytime soon”. Directly contradicting Blairs assertion that the government should run a “hundred day plan” for scaling up AI, Acemoglu insists that truly transformative changes won't happen quickly and that few - if any - will likely occur within the next 10 years. Despite this, Blair believes that AI displacing humans in the workforce will be a natural (and speedy) development, announcing that he foresees a loss of about 1.15 million public-sector jobs in the next two decades of his technological revolution. 

Rather than viewing AI as an instant replacement for jobs, as a magic wand already ripe for wishing away our efficiency problems and spontaneously boosting the economy all in one, we should take stock and view it as it really is.  With the constant AI-washing in play there has been an overreach of expectations surrounding what AI can achieve, and how quickly it can be implemented. This overestimation has led to a miscalculation of finances, and has instilled a panic that AI will steal jobs.  Additionally it has undermined the great talents and efforts that are constantly fuelling the development of such an expansive cosmos of possibility - and the time that this takes.

Ironically, in his panel speech alongside William Hague, Blair said “the world is changing so fast, one of the biggest risks for Britain is introspection and to not understand the way the world and AI is developing.” We propose that instead of replacing jobs, AI can redistribute power and time - acting as a copilot to give people back their time to focus on higher priority tasks, or those that require the undeniable human spirit. Instead of stepping out of one boat before lining up another, we should collectively use AI in conjunction with our current processes- for the betterment of them both. In turn this will lead to a more steady and gradual switching to AI as it develops, (without the tendency to oversell what AI can achieve and the proclivity to believe it is always the answer), focusing less on a future that is AI led, and more on one that is AI enabled.

‍