OpenAI CEO Sam Altman says he's a 'little bit scared' of A.I.

1 year ago 55

Sam Altman, co-founder and main enforcement serviceman of OpenAI Inc., speaks during TechCrunch Disrupt 2019 successful San Francisco, California, connected Thursday, Oct. 3, 2019.

David Paul Morris | Bloomberg | Getty Images

OpenAI CEO Sam Altman said successful a caller interrogation with ABC News that he's a "little spot scared" of artificial quality exertion and however it could impact the workforce, elections and the dispersed of disinformation.

OpenAI developed the ChatGPT bot, which creates human-like answers to questions and ignited a caller AI craze.

"I deliberation radical truly person amusive with [ChatGPT]," Altman said successful the interview.

But his excitement implicit the transformative imaginable of AI technology, which Altman said volition yet bespeak "the corporate power, and creativity, and volition of humanity," was balanced by his concerns astir "authoritarian regimes" processing competing AI technology.

"We bash interest a batch astir authoritarian governments processing this," Altman said. Overseas governments person already begun to bring competing AI exertion to market.

Chinese tech institution Baidu, for example, precocious held a merchandise lawsuit for its ChatGPT competitor, a chat AI called Ernie bot.

Years earlier Russia's penetration of Ukraine, Russian President Vladimir Putin said whoever becomes the person successful AI exertion "will beryllium the ruler of the world." Altman called the comments "chilling."

Both Google and Microsoft person aggressively stepped up their AI plays. Microsoft chose to spouse with Altman's OpenAI to integrate its GPT exertion into Bing search. Google genitor Alphabet unveiled an internally developed chatbot called Bard AI, to mixed feedback from Google employees and trial drivers.

The power of ChatGPT and AI tools similar it hasn't yet reverberated done the American predetermination process, but Altman said the 2024 predetermination was a absorption for the company.

"I'm peculiarly disquieted that these models could beryllium utilized for large-scale disinformation," the CEO told ABC.

"Now that they're getting astatine penning machine code, [models] could beryllium utilized for violative cyberattacks," helium said.

ChatGPT's programming prowess has already made a people connected galore developers. It already functions arsenic a "co-pilot" for programmers, Altman said, and OpenAI is moving toward unlocking a akin functionality for "every profession."

The CEO acknowledged that it would mean galore radical would suffer their jobs but said it would correspond an accidental to travel up with a amended benignant of job.

"We tin person a overmuch higher prime of life, modular of living," Altman said. "People request clip to update, to react, to get utilized to this technology."

Watch the afloat interrogation connected ABC News.

OpenAI says its GPT-4 exemplary  tin  bushed  90% of humans connected  the SAT

Read Entire Article