OpenAI and Google Deep Mind staff warn about AI risks

Employees of OpenAI and Google Deep Mind believe that the companies do not take sufficient security measures in the development of artificial intelligence, focusing more on the rapid development of products, as well as on competition with each other, Reuters reported.

"Current and former OpenAI and Google Deep Mind researchers are warning that the companies are doing too little to protect the world from potentially catastrophic consequences from building self-improving artificial intelligence systems that can outpace their human control," the agency said.

As Reuters notes, in video messages collected for the agency by Palisade Research, employees of these companies said that their concerns about the risk of artificial intelligence are sincere, and this is not a marketing ploy.

These calls were part of the frominside.ai project, which is an attempt to share their concerns directly with the public of those concerned about the development of artificial intelligence.

Thus, according to Reuters, in one of the videos, a Deep Mind researcher said that the likelihood that artificial intelligence could lead to the extinction of humanity is at least 10%. He described this figure as “unjustifiably high,” the agency adds.

In another video, a research engineer from OpenAI said that in matters of developing artificial intelligence we need to act “very carefully,” but instead the laboratories are competing with each other.

“Nobody knows how this will end: either we will cure cancer, or we will lose all our jobs, or perhaps we will all die,” Reuters quoted him as saying.

Artificial intelligence researchers have been working on security issues for years, but the general public has become especially concerned since OpenAI hacked the Hugging Face platform.

Since then, the debate over how to balance security and progress in this area has divided the technology industry and become a global political issue, the agency emphasizes.