ChatGPT creator Sam Altman’s AI reality check
Business
Utpal Siddhartha
24 JUL 2025 | 10:52:26
Are 20 lakh jobs under threat in India? If you work in customer support, this might be your wake-up call. Because according to AI pioneer Sam Altman, the threat isn’t coming, it’s already here.
''AI customer support is incredible”
Speaking at the Integrated Review of the Capital Framework for Large Banks Conference in Washington DC, the OpenAI CEO pointed to customer service as one of the roles likely to be soon replaced by AI. He said, “Some areas, I think, will be totally gone. For example, AI customer support bots are incredible.”
Further explaining how AI was transforming the sector, Altman said, “Now you call one of these things and AI answers. It’s like a super-smart, capable person. There’s no phone tree, there are no transfers. It can do everything that any customer support agent at that company could do. It does not make mistakes. It’s very quick. You call once, the thing just happens, it’s done.”
AI in customer support: India impact
Now, if Sam Altman’s take on this holds true, the impact on India could be massive. We’re a global hub for customer service roles, with anywhere between 20 to 30 lakh people currently estimated to be employed in the sector.
Will Doctors survive the AI threat?
Altman, the architect of AI platform ChatGPT, also talked about the impact of AI on some other sectors. He underlined that though ChatGPT today can often provide a better diagnosis than most doctors worldwide, he himself wants a human doctor. Talking about people’s experiences with ChatGPT, in the context of medical advice, Altman said many people share stories of ChatGPT saving their lives by diagnosing rare diseases that doctors missed. But, he added, that he doesn’t want to entrust his medical fate to ChatGPT with no human in the loop. He also wondered if anyone in the audience would rather be diagnosed by ChatGPT than a doctor, even if it’s better, calling it "quite an interesting question.”
AI concerns that worry Sam
Sam Altman also talked about the things about AI that keep him up at night, and worry him. One of his key concerns regarding AI is a bad actor with super intelligence first. He warned an adversary, maybe a hostile nation, develops super intelligent AI first and misuses it—designing bio weapons, taking down the power grid, breaking into financial systems, etc. This, Altman said, would be unimaginable without super intelligence, but possible with it. He added that because we don’t have such defences, this is a big risk. Altman underlined that AI’s bio and cybersecurity capabilities were growing rapidly, but the world was not taking the warnings seriously enough.
What if we lose of control to AI?
The OpenAI CEO also talked about the loss of control incidents. He said the sci-fi scenario where AI refuses to be turned off or disobeys humans is less concerning to him than the first category, but remains a serious worry. Therefore, he added, OpenAI works extensively on model alignment to prevent this, but as systems become more powerful, the risk increases.
Accidental takeover by AI
Sam also talked about accidental takeover by AI through over reliance. He said there could be a scenario where models don’t wake up or become malevolent but become so ingrained and smarter than us that we rely on them heavily without fully understanding their operation. Society’s decisions could increasingly depend on AI.
''AI customer support is incredible”
Speaking at the Integrated Review of the Capital Framework for Large Banks Conference in Washington DC, the OpenAI CEO pointed to customer service as one of the roles likely to be soon replaced by AI. He said, “Some areas, I think, will be totally gone. For example, AI customer support bots are incredible.”
Further explaining how AI was transforming the sector, Altman said, “Now you call one of these things and AI answers. It’s like a super-smart, capable person. There’s no phone tree, there are no transfers. It can do everything that any customer support agent at that company could do. It does not make mistakes. It’s very quick. You call once, the thing just happens, it’s done.”
AI in customer support: India impact
Now, if Sam Altman’s take on this holds true, the impact on India could be massive. We’re a global hub for customer service roles, with anywhere between 20 to 30 lakh people currently estimated to be employed in the sector.
Will Doctors survive the AI threat?
Altman, the architect of AI platform ChatGPT, also talked about the impact of AI on some other sectors. He underlined that though ChatGPT today can often provide a better diagnosis than most doctors worldwide, he himself wants a human doctor. Talking about people’s experiences with ChatGPT, in the context of medical advice, Altman said many people share stories of ChatGPT saving their lives by diagnosing rare diseases that doctors missed. But, he added, that he doesn’t want to entrust his medical fate to ChatGPT with no human in the loop. He also wondered if anyone in the audience would rather be diagnosed by ChatGPT than a doctor, even if it’s better, calling it "quite an interesting question.”
AI concerns that worry Sam
Sam Altman also talked about the things about AI that keep him up at night, and worry him. One of his key concerns regarding AI is a bad actor with super intelligence first. He warned an adversary, maybe a hostile nation, develops super intelligent AI first and misuses it—designing bio weapons, taking down the power grid, breaking into financial systems, etc. This, Altman said, would be unimaginable without super intelligence, but possible with it. He added that because we don’t have such defences, this is a big risk. Altman underlined that AI’s bio and cybersecurity capabilities were growing rapidly, but the world was not taking the warnings seriously enough.
What if we lose of control to AI?
The OpenAI CEO also talked about the loss of control incidents. He said the sci-fi scenario where AI refuses to be turned off or disobeys humans is less concerning to him than the first category, but remains a serious worry. Therefore, he added, OpenAI works extensively on model alignment to prevent this, but as systems become more powerful, the risk increases.
Accidental takeover by AI
Sam also talked about accidental takeover by AI through over reliance. He said there could be a scenario where models don’t wake up or become malevolent but become so ingrained and smarter than us that we rely on them heavily without fully understanding their operation. Society’s decisions could increasingly depend on AI.
