The Need For Rules To Make AI Safe And Transparent

Experts say the UK needs rules to make sure AI is safe and clear. This will help British businesses and schools feel confident about using AI. A new report from the University of Cambridge suggests that the UK government should give tax breaks to companies that create AI products and services or use AI in their work. The aim is to make the UK more productive.

The researchers argue that the UK doesn’t have the right technology and money to quickly build advanced machine learning models like big US companies such as Google, Microsoft, or OpenAI. Instead, they suggest focusing on using AI for practical things like new medical tools and solving the shortage of software engineers. This could boost the UK’s economy.

New protections

But the researchers also warn that without new laws to control AI in the UK, these plans might not work. People and businesses might not trust AI like ChatGPT enough to spend time and money learning how to use it.

“Generative AI will change the nature of how things are produced, just as what occurred with factory assembly lines in the 1910s or globalised supply chains at the turn of the millennium,” the researchers explain. “The UK can become a global leader in actually plugging these AI technologies into the economy.”

Generative AI is a powerful technology that can rapidly create original and high-quality text, images, audio, or video. This year, there was a lot of talk about ChatGPT, which generates text, and Midjourney, which can create different styles of images in just seconds.

Tech requirements

To make these AI models work, we need clusters of specialized computer hardware called Graphics Processing Units (GPUs). These GPUs handle the enormous amounts of data needed to train these machine-learning models. For instance, running ChatGPT alone can cost around $40 million each month just in computing expenses. Earlier this year, the UK chancellor announced £100 million to establish a “Frontier AI Taskforce” aiming to develop AI in the UK that can compete with giants like Google Bard.

But here’s the catch: The report points out that the supercomputer promised by the UK chancellor won’t be up and running until 2026. Meanwhile, none of the three big tech companies in the US – Amazon, Microsoft, or Google – have set up GPU clusters in the UK yet.

“The UK has no companies big enough to invest meaningfully in foundation model development,” the researchers say. “State spending on technology is modest compared to China and the US, as we have seen in the UK chip industry.”

Building on strengths

The report advises that the UK should leverage its strengths in fin-tech, cybersecurity, and health-tech to create software – the apps, tools, and interfaces – that can make everyday use of AI a reality.

This generative AI technology has the potential to speed up coding by about 55%, which could be a solution to the UK’s chronic shortage of software developers. Interestingly, it can even assist individuals who aren’t programmers in building advanced software.

Furthermore, the UK boasts world-class research universities that are well-positioned to address various challenges in the AI field, from improving data center cooling to detecting AI-generated false information.

However, at present, UK organizations lack sufficient incentives to adhere to responsible AI practices. According to the researchers, the current approach to regulating generative AI in the UK relies on vague and voluntary principles that touch on security and transparency. To truly unlock the economic potential of AI, the report stresses that meaningful legislation and regulation are imperative to instill trust in the technology.

In addition to these new AI laws, the report proposes a range of tax incentives. This includes an improved Seed Enterprise Investment Scheme to boost the availability of funding for AI start-ups. It also suggests tax credits for all businesses that incorporate generative AI into their operations. To stimulate innovation from within organizations, the report recommends launching challenge prizes to discover creative uses of generative AI.

Facebooktwitterredditpinterestlinkedinmail