'Godfather of AI' Geoffrey Hinton says he trusts OpenAI’s GPT-4 chatbot more than he should
Even the “Godfather of AI” admits he puts too much faith in his favourite chatbot, despite knowing better.
- May 20, 2025,
- Updated May 20, 2025 9:16 AM IST
Geoffrey Hinton, widely regarded as the “Godfather of AI,” has confessed to placing too much trust in artificial intelligence, specifically OpenAI’s GPT-4, even when the technology falls short. In a recent CBS interview, Hinton, who was awarded the 2024 Nobel Prize in physics for his foundational work in machine learning, candidly reflected on his reliance on the chatbot for everyday tasks.
“I tend to believe what it says, even though I should probably be suspicious,” Hinton admitted, acknowledging the subtle risks of becoming too dependent on AI tools. During the segment, Hinton presented GPT-4 with a straightforward riddle: “Sally has three brothers. Each of her brothers has two sisters. How many sisters does Sally have?” GPT-4 incorrectly answered two, when the correct answer is one, Sally herself.
“It surprises me. It surprises me it still screws up on that,” Hinton said, highlighting the lingering limitations of even the most advanced AI models. He characterised GPT-4 as “an expert at everything. It’s not a very good expert at everything,” pointing to its broad capabilities but occasional blunders.
Despite this, Hinton remains optimistic about progress in AI, suggesting that upcoming versions will continue to improve. “Yeah, I suspect,” he said when asked if GPT-5 would solve the riddle correctly.
Notably, after the interview aired, users on social media reported that the latest versions of OpenAI’s technology, such as GPT-4o and GPT-4.1, handled the riddle accurately, indicating rapid ongoing advances in the field.
OpenAI launched GPT-4 in 2023, quickly making it an industry benchmark for advanced reasoning and language tasks. Since then, it has released GPT-4o, which boasts faster, more versatile performance, and has continued iterating with models like GPT-4.5 and GPT-4.1. Meanwhile, Google’s Gemini 2.5-Pro currently tops the independent Chatbot Arena leaderboard, though OpenAI’s offerings remain close behind.
For Unparalleled coverage of India's Businesses and Economy – Subscribe to Business Today Magazine
Geoffrey Hinton, widely regarded as the “Godfather of AI,” has confessed to placing too much trust in artificial intelligence, specifically OpenAI’s GPT-4, even when the technology falls short. In a recent CBS interview, Hinton, who was awarded the 2024 Nobel Prize in physics for his foundational work in machine learning, candidly reflected on his reliance on the chatbot for everyday tasks.
“I tend to believe what it says, even though I should probably be suspicious,” Hinton admitted, acknowledging the subtle risks of becoming too dependent on AI tools. During the segment, Hinton presented GPT-4 with a straightforward riddle: “Sally has three brothers. Each of her brothers has two sisters. How many sisters does Sally have?” GPT-4 incorrectly answered two, when the correct answer is one, Sally herself.
“It surprises me. It surprises me it still screws up on that,” Hinton said, highlighting the lingering limitations of even the most advanced AI models. He characterised GPT-4 as “an expert at everything. It’s not a very good expert at everything,” pointing to its broad capabilities but occasional blunders.
Despite this, Hinton remains optimistic about progress in AI, suggesting that upcoming versions will continue to improve. “Yeah, I suspect,” he said when asked if GPT-5 would solve the riddle correctly.
Notably, after the interview aired, users on social media reported that the latest versions of OpenAI’s technology, such as GPT-4o and GPT-4.1, handled the riddle accurately, indicating rapid ongoing advances in the field.
OpenAI launched GPT-4 in 2023, quickly making it an industry benchmark for advanced reasoning and language tasks. Since then, it has released GPT-4o, which boasts faster, more versatile performance, and has continued iterating with models like GPT-4.5 and GPT-4.1. Meanwhile, Google’s Gemini 2.5-Pro currently tops the independent Chatbot Arena leaderboard, though OpenAI’s offerings remain close behind.
For Unparalleled coverage of India's Businesses and Economy – Subscribe to Business Today Magazine