OpenAI claims to take proactive measures to prevent AI from ‘going rogue’

OpenAI, the company that developed ChatGPT, announced on 5 July 2023, that it would allocate a sizable amount of funds and establish a new research team, which ensures that its artificial intelligence is secure for use by people.

Ilya Sutskever, a co-founder of OpenAI, and Jan Leike, the director of alignment, said in a blog post that, "The vast power of superintelligence could lead to the disempowerment of humanity or even human extinction. Currently, we don’t have a solution for steering or controlling a potentially superintelligent AI and preventing it from going rogue."

Arrival of superintelligent AI systems

According to the blog post, superintelligent AI systems are intelligent enough to surpass humans. Humans will need better techniques than those currently available to be able to control the superintelligent AI, hence the need for breakthroughs in so-called alignment research, which focuses on ensuring AI remains beneficial to humans.

The report says that 20 percent of the computational capacity that OpenAI has acquired will be used to address this issue over the next four years. A new team, known as the ‘Superalignment Team,’ is also being formed by the company to focus on this project.

What is the team's objective?

The objective of the team is to develop an AI alignment researcher that possesses a level of intelligence comparable to humans. This researcher will be trained through a three-step process i.e., utilising human feedback to train AI systems, training AI systems to assist in human evaluation, and ultimately training AI systems to independently conduct alignment research.

OpenAI aims to further amplify this capability by leveraging significant computational resources. In an interview, AI safety advocate Connor Leahy criticised the plan, stating that it had a fundamental flaw. He said, "You have to solve alignment before you build human-level intelligence; otherwise, by default, you won’t control it. I personally do not think this is a particularly good or safe plan."

Meanwhile, concerns regarding the risks associated with AI have garnered significant attention from both AI researchers and the general public.

In April, several industry leaders and experts signed an open letter advocating for a six-month pause in the development of systems surpassing OpenAI's GPT-4. Moreover, as per a poll conducted in May, which revealed that over two-thirds of Americans are worried about the potential adverse impacts of AI, with 61 percent believing that it has the capacity to pose a threat to civilisation.