Elon Musk will use tweets to train his new AI bot

Elon Musk has warned of the potential risks of warning AI developers about the «future of the Terminator». Photo: Anadolu agency

Elon Musk said his new AI company will use tweets to train its algorithm.

The CEO of Twitter and Tesla said his new company, xAI, will use public tweets as a training dataset and could work with them as well. Tesla on software.

Mr. Musk announced the creation of xAI earlier this week, saying his goal would be to «understand the universe.»

On Friday in a 90-minute audio chat He told Twitter Spaces he wants to create an AI that is «as curious as possible and as truthful as possible.»

Mr. Musk said xAI will work closely with his other ventures, including scanning tweets to train his programs.

The disclosure could raise concerns among privacy groups and Twitter users about whether Musk has the right to use the data.

AI companies are facing a growing number of legal challenges from publishers in connection with using news articles and other intellectual property. to train your bots.

The owner of the Daily Mail is preparing for a legal battle with Google over allegations that the tech giant used hundreds of thousands of online news feeds to train its AI robot without permission, reports The Telegraph. reported earlier this week.

Learning is the process by which AI is taught to understand and interpret data so that it can ultimately make its own decisions.

0605 AI Impact

ChatGPT, best known bot with artificial intelligence, was trained on a large number of texts from books. , Wikipedia, articles and other written materials, according to reports.

Mr. Musk said that working with Twitter and Tesla could bring «mutual benefits» to all of his companies and could spur Tesla's self-driving car work. .

Despite his new AI business, Musk is more skeptical than other tech entrepreneurs about the technology's prospects.

Earlier this year, the head of Tesla was on a panel of AI experts and some rival tech industry bosses. which called for a six-month pause in AI development.

Mr Musk signed a letter warning of the technology's potential to «flood our information channels with propaganda and lies», «automate all jobs» and, in the worst case, , “risks losing control of our civilization.”

Former Google CEO Eric Schmidt warned in May that AI could harm or kill “many people” in the future.

artificial intelligence? Poll

Last week, Mr. Musk said developers should be wary of a «future Terminator,» alluding to a scenario in which AI would go rogue and try to harm humans.

However, speaking on Twitter on Friday, Mr. r Musk Musk seems to have downplayed the risk. He said: «I think for superhuman intelligence, humanity is much more interesting than non-humanity.»

According to xAI's website, the company is being advised by the director of the AI ​​Security Center, a non-profit organization headquartered in Francisco. which is exploring ways to securely deploy and develop AI.

Artificial intelligence has been in the spotlight over the past year following the success of ChatGPT. The latest version, GPT-4, is said to be showing signs of approaching human-level intelligence.

Rishi Sunak is preparing to hold a global AI summit in London later this year in hopes of developing international rules. governing the use and development of technology.