OpenAI boss Sam Altman says DeepSeek did ‘nice work’ with AI chatbot

OpenAI boss Sam Altman has acknowledged that Chinese AI firm DeepSeek did some “nice work” in the creation of the chatbot now rivalling his firm’s ChatGPT.

Mr Altman said he was going to meet the competing AI firm, but the meeting would not take place during this week’s AI Action Summit in Paris, which he will attend on Tuesday.

The sudden emergence of DeepSeek last month shook the AI sector and sent US tech stocks tumbling after the Chinese chatbot gave a performance to rival ChatGPT.

That performance was reportedly at only a fraction of the billions of dollars spent by OpenAI on its own platform, raising questions about whether the seemingly untouchable US-based AI firms could in fact be challenged.

Speaking to The Times Tech Podcast, Mr Altman said the timing of DeepSeek’s appearance took him by surprise, even if the appearance of a new challenger did not.

“We knew that there were going to be, at some point, we would get more serious competitors and models that were very capable, but you don’t know when you wake up any given morning that that’s going to be the morning,” he said.

“So it’s like, it is not surprising at all that it happened. Surprising it happened that day, I guess.”

He added: “They did some nice work. And I think there’s also some nice pieces of product work, like showing the chain of thought was clearly something people wanted.

“Mass availability in the free tier was clearly something people wanted. Research wise, it’s not a big update to us, although they did a few nice things there as well.”

The two-day AI summit in Paris, hosted by French President Emmanuel Macron, is seen as an opportunity for world leaders and the largest tech firms to find some common ground and a global approach on the development and governance of AI.

But security and safety concerns have been raised about the nature of China-based AI development.

Mr Altman acknowledged that said regional differences in AI products was inevitable, given current geopolitics, and that AI services would likely “operate differently in different countries”.

“I think you can imagine a sort of hierarchy of what AI has to do globally. Here’s what AI has to do with this country, this country, and then individual users,” he said.

“One thing that I think is not so easy to resolve is, are we heading towards, like, more authoritarian AI or more democratic AI.

“I’m very much on the democratic AI side, but that comes with some trade-offs, and that will come with some bad things in society too, like we have never empowered individuals this much as we’re about to, and I think that’s the only way forward.

“But it is, you know, it is a different thing. So, my hope is that we can find what we can agree on, have some rules, and the technology operates differently in different countries.

“I’m sure that, you know, there’ll be different models in China versus the west, but you know, we’ve all got a shared interest in  the world continuing on.”

The ChatGPT boss also discussed his firm’s latest innovation, Deep Research, a tool designed to be capable of independently discovering online information and carrying out complex, multi-step research tasks on behalf of users.

Mr Altman said he believed the tool could have a significant impact, and that he thinks it can do “about 5% of all tasks in the economy today”.

He added that AI was already changing people’s daily lives, and predicted this would continue at an even faster rate.

“A lot of people, even recent AI sceptics, were saying things like, ‘I can now do things that would have taken me many days or even weeks of work’,” he said

“‘AI can do it in 20 minutes, and it can do a bunch of these tasks in parallel. It’s going to change the way I do my job. It’s going to change the way my scientific field works’.

“If you project this same rate of change forward another two years, or 10 years I think that’s going to mean people can do incredible things.

“I’m not sure what I would ask the AI to do for me at that point. I think that we will be quite literally limited by the questions we can come up with and we’ll have AI to help us with that too.”

Mr Altman’s comments came as OpenAI separately released new statistics about ChatGPT and OpenAI usage in the UK.

The US firm said the UK was in the top three countries globally for paying ChatGPT subscribers, as well as in the top three for developers using OpenAI’s tools to build their own programmes, and for paying business customers.

The US firm noted that the UK Government was among its customers, and was using its GPT-4o model as part of a test of a chatbot to help small businesses navigate the Gov.UK website.

The AI chatbot has more than 300 million weekly active users.

On the UK’s AI adoption, Mr Altman said: “The UK has played a pivotal role in the development of AI and is now a world leader in its adoption.

“Millions across the UK are experiencing how AI can improve how they live, work and learn, whilst businesses of all sizes are boosting productivity, creativity and competitiveness at scale.

“As AI capabilities advance, we expect even more people, businesses and start-ups in the UK to benefit from this transformative technology.”