Generative artificial intelligence (GAI) and the potential future it could generate have been hot topics ever since OpenAI released ChatGPT into the wild in late 2022. The technology is viewed as a net good by supporters of capitalism; they see it as the logical continuation of the digital age, which has helped to create vast wealth… for a chosen few. Meanwhile, skeptics recall the best science fiction from the 1980s and worry that we could soon develop our own versions of HAL, SHODAN, Ultron, SkyNet, and GLaDOS.
The vocal minorities are those. Most people are aware that technology is just a tool without a will of its own when they are presented with the possibilities given by generative artificial intelligence. Users are expected to “do good” with it. And if that is not possible because “good” is inherently subjective… then democratic governments need to step in and regulate.
The best way to go about doing this is still up for dispute. The proposed AI Act was introduced initially by the European Union. It is a rough first draft, but it has the advantage of representing a genuine effort to manage a very disruptive technology as opposed to letting tech millionaires take the lead. The proposed law is summarized here, along with the advantages and disadvantages of such regulations.
What the EU’s AI Act contains
The AI Act centers the conversation on risk: “The new regulations set forth requirements for users and providers based on the degree of artificial intelligence risk. Even though many AI systems present little risk, they nevertheless need to be evaluated.
AI that poses “unacceptable” levels of risk (such as social scoring, real-time and remote biometrics, and behavioral manipulation) will be outlawed.
Law enforcement, education, and immigration-related high-risk AI systems will be evaluated both before and during the course of their development. Limited-risk AI systems will need to “comply with minimal transparency requirements that would allow users to make informed decisions.”
