The 51-minute long love-in between tech mogul Elon Musk and Rishi Sunak aired last night. Advertised as a “fireside chat“, the PM played a Jimmy Fallon-esq eager chatshow host, with the two giggling and agreeing on almost all of Sunak’s soft ball questions. Guido watched it so you don’t have to…
They agreed that rogue robots should have an “0ff-switch” (they’ve both seen The Terminator), and Musk praised Sunak’s decision to invite China to the summit and hailed London as a global power in AI. Even Elon’s bleak prediction that robots will make better friends than humans got no push-back from the PM. There was a brief hint of protest from Sunak when Elon claimed that AI means “there will come a point where no job is needed”, to which Sunak weakly responded “but work gives you meaning“, without really making much of a defence of this core Tory value. Still twenty points behind…
Rishi has announced a “landmark” agreement with other countries at the AI safety summit that will “tip the balance in favour of humanity” and prevent a dystopian Terminator-style nightmare. The agreement will supposedly mitigate against the robot takeover by introducing government-led safety checks and tests for new AI models…
“Like-minded governments and AI companies have today reached a landmark agreement. We will work together on testing the safety of new AI models before they are released. This partnership is based around a series of principles which set out the responsibilities we share and it’s made possible by the decision I have taken, along with Vice-President Kamala Harris, for the British and American governments to establish world-leading AI safety institutes.”
He’s set to host Elon Musk for a fireside chat later this evening, although it will be prerecorded, not streamed live. He dodged the question of whether that’s because he doesn’t trust Musk to go off-piste…
Nick Ferrari gave Oliver Dowden a rinsing on LBC this morning over his inability to name any possible recipients of £38 million in taxpayer cash gifted to “accelerate development in world’s poorest countries using AI“. 8 African research labs are said to be receiving funding, though Dowden couldn’t name where any of them would be. Hardly specific and targeted…
Ferrari said: “If I was DPM and I was handing out £38 million worth of voters’ money I’d be bloody certain which countries it’d be going to in Africa, but that’s just me“.
The FCDO has announced this morning that the UK is combining with Canada and the Bill and Melinda Gates Foundation to promote a “boost in AI programming to combat inequality and boost prosperity on the continent“. Because only UK taxpayer funds can unlock “the benefits of AI to the 700 million people who speak 46 African languages”…
Rishi’s warnings about human extinction this morning were spooky – Westminster’s think tanks are cautious on his Bletchley Park safety festival and the government’s AI policy. It’ll be a strange event… Kamala Harris is attending and Ursula von der Leyen’s apparently on the fence about turning up…
Matthew Lesh at the IEA says Rishi is “right to highlight the immense opportunities“, although “fear mongering risks setting fire to a nascent industry that has the potential to help solve some of our most intractable issues“. Don’t set up the guillotine too early…
Matthew Feeney of the CPS says Sunak “tried to give some reassurance that the government won’t rush to regulate AI which is welcome… However, businesses and entrepreneurs will need some certainty if they are to fully unlock the benefits of AI“.
The ASI’s Mimi Yates says the “announcement of more investment in British computer power will help the UK on its way to becoming the most agile place for researchers and start-ups to develop their businesses, as long as we also create favourable market conditions”. The invisible hand will do its work…
Allan Nixon from Onward, which has a report out today calling for tax breaks for AI startups, says the government “is right to put the UK at the forefront of global AI safety” with the caveat; “we must remember that AI is also a fantastic force for good if we harness it right“. Even the wets are playing the same tune…
For 30 years in the 19th century newly-invented motor cars were limited to 2mph and had to have a minimum of two operators plus a man waving a red flag 20ft in front to warn other road users. The wonks are telling Rishi not to make the same mistake…
Sunak is off to a cheery start this morning warning about humanity “losing control of AI completely” to a “superintelligence” that could result in “extinction“. Someone tell the civil servants…
Apart from warning the public about Skynet, Sunak announced a new AI safety institute that would take testing out of the hands of private companies, an expert panel to produce a report, and various investments in quantum computing and healthcare AI. All eyes on Bletchley Park next week as world leaders decide whether to turn up or not…
Dustin Moskovitz, Facebook co-founder and CEO of the multi-billion dollar firm Asana, has called on Rishi to leverage Brexit freedoms to position the UK as a global AI titan. Before Sunak’s big AI Summit in November, Moskovitz hailed London as a “massive magnet” for tech firms, citing EU regulatory hurdles and GDPR conflicts with AI platforms like ChatGPT. Moskowitz told The Times:
“The UK is talking about AI in more appropriate ways … The EU could ban large language models completely. They’re not compatible with GDPR [General Data Protection Regulation] and the right to be forgotten — so you either ignore that law or you need a new one”
He added “London is a huge gravitational force“, and the EU will take time to catch up if the UK takes this seriously. Something Brexiteers can agree with… provided Whitehall doesn’t rely on AI too much to do their jobs.
