Tuesday, April 25, 2023 at 5:15 am
The latest version of ChatGPT makes last year’s wildly popular AI chatbot look like its dumb little brother, but the UK still hasn’t figured out what to do or how to use artificial intelligence. wrote Tom Westgarth.
It’s easy to look back clearly at the pandemic. At this distance it is clear that the signs were all there. China’s lockdown, Lombardy’s hospital ward surge. January 2020 and he even February’s exponential growth data should indicate the need to act now.
The exponential growth of AI is a similar story. GPT-3, a breakthrough large-scale language model (LLM) that can predict the next sequence of words in a sentence, was created by Open AI in June 2020. Since then, the open source community has improved the model faster than many could have imagined. The new era of “generative AI” has produced more than just poetry and art. New tools and businesses have helped him design antibodies, produce new music, and even create his Linux terminal. With the release of his GPT-4 in the last few weeks, ChatGPT, which went viral in late 2022, looks like a dumb little brother. Time moves fast in the AI era.
Despite a respectable system in place during the pandemic, the response was still unsuccessful. And I misunderstood the basics such as masks and expanding inspections. Even with all these institutions in place, governments are still unable to cope with the crisis.
Still, few health security agencies have the equivalent of AI, let alone the WHO, and AI’s office is ill-equipped to meet the new challenges of the next generation.
Take the British approach. The government’s AI strategy was highly praised by experts for its purpose, but was not even funded. Its AI department conducts policy and research, but has no legal regulatory authority and arguably lacks the political priorities it deserves to coordinate its response to AI.
Technologists and entrepreneurs are eagerly awaiting the release of the UK’s ‘promoting innovation’ regulatory framework for AI this year. But with only a few dozen AI employees in the office, the team has limited work to anticipate and respond to emerging challenges and create an environment for the market to mature. .
I already have a big question that needs an answer. We want people to be able to use AI tools for creative purposes. But how can we allow people to create AI music while ensuring that the artist’s IP is protected?Original music created from AI trained with tools that collect data from the internet If so, is the person infringing copyright?
UK-based Stability AI, the latest potential AI leviathan behind Stable Diffusion, is currently facing lawsuits for scraping artists’ work without their consent. The outcome of this lawsuit is important for the future of the UK AI market. Will relevant UK government departments plan to enact legislation aggressively, or will judges decide the course of AI copyrights in the UK? not.
We need better institutional capacity to provide this kind of information to ministers and officials. As models and the dilemmas they pose become more complex, so must our capabilities.
One possibility is for the Office of AI to lead a “whole-of-government” perspective approach to understanding how to profit from AI disruption. The office of AI should go to all government departments and ask them to consider all possible ways public services could benefit from new generative AI tools, as well as foreseeable issues.
In the most impactful scenarios, we need to put in place a framework that adapts to different stages of AI development and use. A similar framework was provided as a blueprint by Stanford University’s Cyber Policy Center and his OpenAI to address emerging disinformation threats.
This tactic should be part of a new set of responsibilities assumed by expanded government AI agencies. To say it’s AI-ready is cheesy. If you don’t have the talent, foresight and ambition to be agile with new risks and opportunities, you can follow pandemic-era advice and stay home.
As it stands, there are few political priorities. So-called horizon scanning to understand emerging AI is being done in some departments, but in different ways. This role requires a dedicated team, but it can only be achieved if the Ministry of Science, Innovation and Technology provides adequate support.
AI isn’t currently on the list of ‘most important issues for the public’. But tomorrow, a nationwide scandal (e.g., a massive cyber hack into a public database backed by GPT-4) could make this area a major concern for voters.
This is part of a series of essays published today by The Entrepreneurs’ Network.
