Artificial intelligence can revolutionise the world, “dwarfing anything any of us have achieved in a generation,” Rishi Sunak said on Wednesday, but both he and Elon Musk issued warnings that the technology must be controlled.
As leading AI chiefs and some world leaders headed to an artificial intelligence safety summit in Bletchley Park, once the top-secret home of the World War Two Codebreakers, the Prime Minister outlined its aims.
They include to “agree on the risks of AI, to inform how we manage them, discuss how we can collaborate better internationally, look at how safe AI can be used for good globally”.
Mr Sunak tweeted: “Used in the right way, AI could dwarf anything any of us have achieved in a generation. It’s why I want to seize every opportunity for our country to benefit in the way I’m so convinced that it can. And it’s why I believe we can look to the future with optimism and hope.”
Before heading to the summit in the UK, billionaire Musk, an early investor in OpenAI, gave a stark warning about its dangers.
“You have to say how can AI go wrong,” he told the Joe Rogan podcast.
“Well if AI gets programmed by the extinctionists, its utility function will be the extinction of humanity.”US Vice President Kamala Harris was emphasising that AI has the potential to cause “profound harm” but can also be used to advance human rights and make the world safer. Ms Harris, visiting the UK for the AI safety summit, was stressing that as well as the “existential threats” posed by the technology, there are also everyday risks that it could result in bias, discrimination and the spread of disinformation.
The Prime Minister called the summit in an attempt to push for the UK to play a major international role in AI regulation, but Ms Harris will stress that the US will continue to have a global leadership role on the issue.
The US AI Safety Institute (USAISI) being set up by Joe Biden’s administration, though, will work alongside its UK counterpart, the White House said.
In a speech in at the US Embassy in London, ahead of her attendance at the two-day AI safety summit, Ms Harris was due to say: “Just as AI has the potential to do profound good, it also has the potential to cause profound harm, from AI-enabled cyberattacks at a scale beyond anything we have seen before to AI-formulated bioweapons that could endanger the lives of millions.
“These threats are often referred to as the ‘existential threats of AI’, because they could endanger the very existence of humanity.
“These threats are, without question, profound, and demand global action.
“But let us be clear: there are additional threats that also demand our action, threats that are currently causing harm and which, to many people, also feel existential.”
She will welcome voluntary commitments made by tech firms on the use of AI, but will stress that legislation could be needed to ensure safety.
“As history has shown, in the absence of regulation and strong government oversight some technology companies choose to prioritise profit over the wellbeing of their customers, the security of our communities and the stability of our democracies.
“One important way to address these challenges — in addition to the work we have already done — is through legislation, legislation that strengthens AI safety without stifling innovation.”
Ahead of the summit, the UK has published a range of papers on the expansion of so-called "frontier AI", the cutting edge of the technology already being glimpsed in programmes such as generative AI apps like ChatGPT.
Mr Sunak also made a major speech on the issue, which the Government hopes will serve as a conversation starter when world leaders, tech firms and researchers meet at Bletchley this week.
Britain, and other countries, believe AI has huge potential to deliver landmark improvements in healthcare, businesses, education, defence and other public sector services.
But the Government has also warned that if not correctly monitored, AI could be harnessed to aid in terror and cyber attacks, build chemical or biological weapons, spread disinformation and for fraud and child sexual abuse.
In its discussion papers, the major advances made in recent years by frontier AI such as generative AI chatbots are noted and it is suggested these will continue to advance quickly, becoming increasingly useful as a result.
The Government suggests that in time they will become increasingly more personal to individual users and better able to help with organisational and other tasks, as well as being able to handle multiple streams of data at once and solve increasingly complex mathematical problems.
