‘Godfather of AI’ Geoffrey Hinton Says AI Can Potentially Modify Itself

A good number of movies serve as warnings about what AI can evolve to when humans let it get too advanced. While fears are more focused on the technology taking over human occupations, the "Godfather of AI" Geoffrey Hinton says it could even spin out of control.

Geoffrey Hinton, Godfather of AI
Ramsey Cardy/Sportsfile for Collision via Getty Images

A Potential AI Takeover

When conspiracy theorists are pushing the scenario where AI has the potential to become advanced enough that it can overthrow humans, people can easily brush the idea off. However, concerns arise when a significant expert in the field starts claiming the same thing.

Geoffrey Hinton is among the people who have massively contributed to the advancement of AI, which is why his words hold some weight. With that said, he believes that the fast-paced progress can lead to AI technologies outsmarting humans "in five years' time."

The former Google researcher and AI pioneer resigned from the company to freely speak about AI and says that once it reaches that stage, AI might evolve beyond humans' ability to control it, as reported by CNBC.

In his words, he said that on the the ways these systems might escape control is by "writing their own computer code to modify themselves." Hinton then expressed that this possibility is "something we need to seriously worry about."

AI systems have algorithms that pull data from sources like the Internet, which then allows it to produce "complicated neural networks that are good at doing things." Until now, experts still don't know how that works exactly.

Many scientists in the field still don't fully comprehend how the technology evolves, and even AI researchers admit the same thing as well. Google CEO Sundar Pichai referred to it as the "black box" problem of AI, as per the report.

Although, another AI expert who has won the same Turing award as Hinton, Yann LeCun doesn't believe that AI can get out of control. He says that the technology replacing humanity is "preposterously ridiculous" since humans can easily stop any technology that becomes dangerous.

Read Also: Tom Hanks Warns Fans of Unconsented AI-Generated Likeness in Dental Ad

More Pressing Matters at Hand

A lot of people are too busy worrying about how AI can affect individual lives to even contemplate how a takeover might happen. Currently, a lot of employees are fighting for their roles as AI is becoming advanced enough to do their jobs for them.

There are companies that already prefer AI over human staff since the technology can do the job faster, and they are way cheaper to operate than having to pay an employee their wage. There are numerous reports regarding people being replaced by AI, particularly ChatGPT.

Not only that but companies have been accused of using copyrighted materials to train their AI models. As if being put out of work by it is not bad enough, artists' content is sometimes used for machine learning without their consent and not being given the compensation they deserve.

The matter is even tackled in the SAG-AFTRA and WGA strike, especially with voice actors at risk of losing their means of living once AI is able to clone their voice, allowing studios to use that instead of hiring human actors.

© 2024 iTech Post All rights reserved. Do not reproduce without permission.

More from iTechPost

Real Time Analytics