Microsoft, Meta stare down DeepSeek threat, making big reveals

DeepSeek will not derail Microsoft and Meta spending a combined $US145bn on AI this year, with Mark Zuckerberg steaming ahead to build a data centre almost the size of Manhattan.

DeepSeek will not derail Microsoft and Meta spending a combined $US145bn ($232.3bn) on artificial intelligence this year, with Mark Zuckerberg steaming ahead with plans to build a data centre almost the size of Manhattan.

China has created an AI model almost on par with America’s best despite not having access to the world’s most advanced chips — and which it says is much cheaper to train and run —, sparking a frenzy across global markets.

Microsoft and Meta were among the first big tech companies to report their quarterly earnings following DeepSeek’s launch, providing a glimpse of how Silicon Valley will react to the Chinese start-up.

Microsoft chief executive Satya Nadella said the arrival of the Chinese model was “good news” and he feels “very good about the investment we are making”, which will total about $US80bn this financial year.

Mr Nadella said DeepSeek had delivered “some real innovations”, which would fuel more demand for AI models. And he said this was good news for Microsoft, which has reportedly invested $US13bn in ChatGPT maker OpenAI.

Facebook owner, Meta, which also reported its quarterly earnings on Thursday morning (Australian time) was also upbeat about DeepSeek, with Mr Zuckerberg reiterating plans to build a data centre “so large it would cover a significant part of Manhattan” as part of a $US65bn spending spree.

Mr Zuckerberg was speaking as Meta agreed to pay about $US25m to settle a lawsuit from Donald Trump after the social media platform suspended his accounts following the attack on the US Capitol in January 2021.

AI is getting cheaper

Mr Nadella said before DeepSeek’s launch Microsoft was experiencing efficiency gains with its AI models to make them cheaper and increase demand.

“What’s happening with AI is no different than what was happening with the regular compute cycle. It’s always about bending the curve and then putting more points up the curve,” Mr Nadella said.

“I think DeepSeek has, you know, had some real innovations. Now that all gets commoditised, and it’s going to get broadly used. And the big beneficiaries of any software cycle like that are the customers.

“That means people can consume more and there’ll be more apps written. This is all good news, as far as I’m concerned.”

Microsoft chief executive Satya Nadella says the arrival of DeepSeek is “good news”.

‘Very happy’ with OpenAI

Mr Nadella said he was “very happy” with Microsoft’s partnership with OpenAI, and the company had been investing to make its AI models cheaper to run to spur broader adoption.

“One of the key things to note in AI is you just don‘t launch the frontier model. If it’s too expensive to serve, it’s no good, right? It won’t generate any demand.

“So you’ve got to have that optimisation so that inference costs are coming down and they can be consumed broadly. That’s the fleet physics we are managing.

“And also remember, you don’t want to buy too much of anything at one time. Because of Moore’s Law, every year is going to give you 2x, your optimisation is going to give you 10x. So you want to continuously upgrade the fleet, modernise the fleet, and at the end of the day, have the right ratio of monetisation and demand-driven monetisation, to what you think of at the training investment. So I feel very good about the investment we are making — and it’s fungible. It just allows us to scale more long-term business.”

Microsoft’s cloud computing business Azure experienced a slowdown in growth last quarter. Mr Nadella said constraints on data centre supply limited the company from meeting soaring demand for AI.

Azure’s revenue increased 31 per cent, compared with analysts estimates of 32 per cent. Microsoft’s overall revenue for the quarter beat analysts’ expectations, soaring 12 per cent to $US69.6bn. Net income jumped 10 per cent to $US24.1bn — also ahead of analysts’ forecasts.

Why AI needs so much investment

Microsoft is planning to spend $US80bn on AI this financial year. This compares with it spending $US55.7bn last financial year.

“Enterprises are beginning to move from proof of concepts to enterprise-wide deployments to unlock the full ROI (return on investment) of AI,” Mr Nadella said.

“Our AI business has now surpassed an annual revenue run rate of $US13bn — up 175 per cent year over year.

“We ourselves have been seeing significant efficiency gains in both training and inference for years now. On inference, we have typically seen more than 2x price performance gain for every hardware generation, and more than 10x for every model generation due to software optimisation. And as AI becomes more efficient and accessible, we will see exponentially more demand.”

Mark Zuckerberg is spending big on AI.

Natural evolution

Mark Zuckerberg said Meta was still “digesting” DeepSeek but said its arrival was a natural evolution.

“That’s part of the nature of how this works, whether it’s a Chinese competitor or not. I kind of expect that every new company that has an advance event, that has a launch, is going to have some new advances that the rest of the field learns from and that’s sort of how the technology industry goes,” he said.

“The field continues to move quickly. There’s a lot to learn from releases, from basically everyone who does something interesting, not just this, not just the ones over the last month, we’ll continue to kind of incorporate that into what we do, as well as making novel contributions to the field ourselves. And I continue to think that investing very heavily in capex and infrastructure is going to be a strategic advantage over time.

“It’s possible that we’ll learn otherwise at some point, but I just think it’s way too early to call that. At this point, I would bet that the ability to build out that kind of infrastructure is going to be a major advantage for both the quality of the service and being able to serve the scale that we want to.”

Originally published as Microsoft, Meta stare down DeepSeek threat, making big reveals