Will AI Develop A Mind Of Its Own And Escape Human Control? Here's What Bill Gates Has To Say

Microsoft co-founder Bill Gates shed more light on the artificial intelligence revolution that has been making a flutter recently as OpenAI and its ChatGPT shot to prominence.

What Happened: Gates was answering a question by Financial Times' Gideon Rachman as to whether he was worried by the idea that at some point AI might develop a mind of its own, have its own ideas about what to do, and escape human control.

This power of being able to perceive and feel is called sentience.

In the long run, if AI gets good enough, and not just writes papers, questions would be asked about the implications for certain jobs, the value of education, etc., Gates told on the podcast hosted by Rachman. “That’s a long way south in the future,” he added.

The billionaire said he sees some labor displacement but expects the technology to lead to increased efficiency. AI will likely help with the “ability to research articles or have things summarized for you so that you can write better articles,” and help doctors to be more efficient, he said.

See Also: Best Technology Stocks Right Now

GPT4 Wows: “You know, GPT2 was kind of a trick. GPT3 uhh it started to say if some interesting [things], GPT4, it’s like, wow!,” Gates said, referring to the evolution of the Generative Pre-trained Transformer of OpenAI. GPT is a language model, generally trained on a huge volume of text data to generate human-like text, and GPT4 is the next iteration of OpenAI’s large language model that is expected to be released shortly.

“You Know, I challenged them in June to do some things GPT3 couldn’t. And then by early September, they showed me that they could do those things,” the tech entrepreneur said.

“And that was very surprising to me,” he added.

When asked to name some of the things GPT4 did, Gates said passing an advanced placement biology exam. “This was very impressive coming many years before I thought, what I thought we’d have to add explicit knowledge, storage and manipulations as opposed to just scaling up LLMs and then a few clever things, but it’s mostly a gigantic LLM,” he said.

Gates also called for AI to be broadly available and not be controlled by a small group of people.

Read Next: OpenAI Co-Founder Sam Altman Doesn't See Artificial Intelligence Ending Google, Discusses Microsoft Partnership And More

Photo: Shutterstock