By Sascha O'Sullivan
We want to be world leaders in artificial intelligence, but government is only willing to pay tech staff half of their market value. If we can’t hire top minds after the layoffs in the industry, we’ll never get ahead, writes Sascha O’Sullivan.
Last month, the Treasury published an ad for a senior role, the Head of Cyber Security. If you’re in charge of a small company’s finances, it’s an important job. If you’re in charge of the country’s, it’s a question of national security.
It was a fairly stock standard job description looking for a “highly motivated individual” keen to work at the “heart of government in a time of momentous change.” What was unusual was the salary: £57,000 per annum. The national average for a head of cyber security is exactly double that at £114,000. If you were to look at companies like Google, Meta or some of the big banks, you can only expect to be able to double it again, at the very least.
Many of the tech companies have a handy trick of hiring people straight out of regulators, places like Ofcom in the UK or the Federal Trade Commission in the US. To do the opposite, and get governments to poach from these companies is nigh on impossible because of the salary differentials.
Thanks to inflationary pressures, tech companies across the world have bled staff like a haemophiliac with an open wound. Over the weekend, one of the top artificial intelligence pioneers, Geoffrey Hinton, quit Google over fears of the speed of the development of the new technology.
“It is hard to see how you can prevent the bad actors from using it for bad things,” Dr. Hinton told the New York Times.
But clearly, Dr Hinton is hoping he can have some influence in mitigating the worst possible consequences of artificial intelligence. Instead of being inside pissing out, he wants to be outside pissing in, and whoever hires him next will benefit from his years of experience as the “godfather of AI”.
After OpenAI, one of the leaders in the artificial intelligence space and owner of ChatGPT, released an updated version of its chatbot in March, more than 1,000 tech leaders and researchers signed an open letter calling for a six-month moratorium on any new systems because of the “profound risks to society and humanity”. Only days later, 19 current and former leaders of one of the top academic institutions of artificial intelligence, the Association for the Advancement of Artificial Intelligence, wrote their own public letter about the risks of AI. One of the signatories was Eric Horowitz, the chief scientific officer at Microsoft, which has integrated OpenAI into its products.
In the UK, there have been some vague noises from the Department of Science, Innovation and Technology about how we go about regulating artificial intelligence. It includes plans to send out “practical guidance” to organisations and businesses, while also developing “pro-innovation” regulation. Legislation to that effect will be introduced, if there is parliamentary time (hint, there probably isn’t).
A report by the think tank Onward has suggested we should even build our very own “Great Britain GPT” to give us control over the technology. Previous attempts for the government to be up with the latest thing have varied from clunky to cringe. Britcoin is evidence of both. The name undeniably ticks the cringe box, while the consultation, set to go until 2025, might just outlast the rise and fall of five new types of digital currency. We certainly have no hope of doing any better with an AI equivalent unless we have people who actually know what they’re doing.
This isn’t just to give the government a hard time. It’s a difficult task, which requires an incredibly high degree of technical expertise. The good news is that there is a list of a thousand or so people worried about AI, with that level of experience, officials could go through and try and hire. They could sit down tomorrow and work their way through the list until they find someone willing to take it on, for the right reasons and the right price.
We routinely talk about how broke we are in the UK, but we’re not completely skint. There is money going around and at least some of that should be put towards hiring the best people. Many will point to the highly restrictive Whitehall pay bands as the problem, and the lack of public popularity for highly paid bureaucrats and political staff. These are thorny, but not insurmountable roadblocks. It shouldn’t take a genius at the top of politics to explain that with very technical roles, it requires salaries to match. Just look at the more than 400 managers across the NHS earning above £150,000.
With the current state of the tech market, the public sector has the chance to bring some of the top minds on board to work on any forthcoming legislation.
In March, the tech community knew just six months could be enough to change the face of AI a few times over. If we’re hoping to understand it and bring it under some control with salaries the equivalent of £57,000, we hardly stand a chance.
The post Britain badly needs a godfather of AI regulation, but we’ll never be able to afford him (or her) appeared first on CityAM.
