OpenAI CEO Warns Against Putting Blind Trust in ChatGPT: 'AI Hallucinates'

TEMPO.CO, Jakarta - CEO of OpenAI, Sam Altman, urges users not to place blind trust in the AI chatbot ChatGPT due to potential inaccurate or misleading information it may provide. 

"People have a very high degree of trust in ChatGPT, which is interesting, because AI hallucinates," Altman said in the inaugural episode of the official OpenAI podcast as reported by India Today on Wednesday, July 2.

Altman's statement sparked widespread discussion among the technology community and daily users, as many rely on ChatGPT for various purposes such as writing, research, childcare, and more.

However, Altman's message was clear. He said ChatGPT is just like other large language models that can provide convincing answers but are actually wrong or misleading, and thus should be used wisely.

ChatGPT operates by predicting the next word in a sentence based on patterns from its training data. However, since it lacks a human-like understanding of the world, ChatGPT sometimes produces inaccurate or entirely fictional information. This phenomenon is known as "hallucination" in the AI world.

Altman emphasized the importance of being transparent and realistic about users' expectations. "It’s not super reliable,” he said. “We need to be honest about that.”

Despite its limitations, ChatGPT is still widely used by millions of people every day. Altman acknowledged this popularity but also warned of the danger if users rely too much and accept answers without verification.

Altman also touched on the importance of transparency and the challenges in maintaining the reliability of AI technology. This was voiced amid increasing legal pressure on OpenAI, including copyright lawsuits from major media outlets such as The New York Times.

He reiterated the company's commitment to remaining open about the limitations of the AI system they developed.

Editor’s Choice: MIT Researchers: ChatGPT Use May Lead to Cognitive Decline

Click here to get the latest news updates from Tempo on Google News

Mark Zuckerberg Launches Meta Superintelligence Labs Division

17 jam lalu

What is the Premier League Companion Created by the English League and Microsoft?

1 hari lalu

New Facebook Feature Uses AI to Edit Your Camera Roll Photos

1 hari lalu

Google Launches Doppl, an AI-Based Virtual Trial App for Clothing

3 hari lalu

Four OpenAI Lead Researchers Leave for Meta, Slack Profile Deactivated

5 hari lalu

LinkedIn CEO Admits AI Writing Feature Struggles to Attract Users

8 hari lalu

MIT Researchers: ChatGPT Use May Lead to Cognitive Decline

8 hari lalu

Apple Reportedly in Talks to Acquire AI Startup Perplexity

10 hari lalu

Tempo Joins Forces with Multiple Media to Bolster Sahabat-AI

12 hari lalu

Tempo Readies Two AI-Based Services for Businesses and Institutions

12 hari lalu