KubeCon China – Open Source a '$9 trillion' asset, says Linux Foundation's Zemlin. But what about AI?

KubeCon + CloudNativeCon is always a lively event – and KubeCon China doubly so. This year, the joint conference moved over the water from the Kowloon side of Hong Kong to the center of the city itself, and into a brand-new five-star hotel and conference center on the edge of its highest point: Victoria Peak (aka The Peak).

The perfect metaphor for Open Source's ascension in software development? And for China's relentless rise to be the world's number two economy and second largest contributor to the Open Source community? Perhaps. But a key question was conspicuous by its absence. More on that later.

Either way, the buzzy event has presented its characteristic mix of enthusiastic young developers, high-level themes, and technical 'lightning talks' on a diverse mix of subjects, including Large Language Models (LLMs) and robotics. That said, it is a more emphatically Chinese affair than last year's KubeCon in Hong Kong, with noticeably fewer Western faces and just one from Africa. A reflection of new political realities, perhaps, though it remains welcoming to everyone.

But of course, the opening keynote could only have one topic: Artificial Intelligence (AI).

A $9 trillion price tag 

Linux Foundation Director Jim Zemlin – the leading light at KubeCons, wherever they are in the world (this year, London, Tokyo, Hyderabad, and Atlanta, as well as Hong Kong) – suggested that Open Source has become a $9 trillion asset. That is the amount of stored value in the world's ongoing projects, he explains, or (if you prefer) the putative saving from proprietary software programs.

That said, the development world still has the capacity to shock, with this year's biggest Open Source story coming from China: the arrival of DeepSeek, which triggered a 17% collapse in NVIDIA's share price in January. But fear not, explains Zemlin: NVIDIA is once again the most valuable company on Earth (leapfrogging Apple to a valuation of over $3.4 trillion this week, roughly the equivalent of the entire UK economy).

Quite who the "fear not" remark was directed at is a mystery, given that nearly all of the audience were Chinese. That aside, Zemlin says of AI, Open Source, and DeepSeek's disruption of the sector:

It really showed that Open Source is a force multiplier, that sharing of ideas accelerates innovation. And it kicked off this trend we're seeing now, where every couple of weeks, a new open-source model comes out that's incredibly performing.

This was such a big deal that it even caused folks like Sam Altman at OpenAI to say things like 'We're on the wrong side of history'. And shortly after Altman said this, OpenAI announced that it would be releasing an Open Source model later this year.

Altman's comments came back in February, though they were expressed with several caveats. Speaking in an online interview, the OpenAI CEO said:

I personally think we have been on the wrong side of history here and need to figure out a different Open Source strategy; not everyone at OpenAI shares this view, and it's also not our current highest priority.

Even so, the point stands about a company that, after all, was named 'OpenAI'. Then Zemlin adds:

Work in Open Source AI continues in the United States, with Meta leading that path. Meta was the creator of things like [deep learning framework] PyTorch, and it is also home to the Llama set of models. And I think you're going to see more innovation come out of organizations and labs like Meta in Open Source and be surprised at how performant and amazing that technology is.

Maybe so, but we are in China, Jim! Finally remembering where he was, Zemlin adds:

In China, folks like [Beijing-based Taiwanese computer scientist] Kai-Fu Lee have been talking about how Open Source is really the winner. And of course, I agree with that when it comes to the basic infrastructure of Artificial Intelligence. Just in the last few months, what we're seeing is this amazing trend where Open Source AI is catching up in terms of evaluations, in terms of performance, with the leading proprietary frontier models.

So, the big question is, how did this happen? Why was everyone so surprised when DeepSeek released this model – though for folks in this room, I suspect it was a little less of a surprise, because the tech industry bet on Open Source a long time ago.

Going against open source is a bad idea

But rather than immediately get into China's role in this, or into DeepSeek's history and strategy, or into the vast amount of funding that the Beijing government has been pumping into tech, and into AI in particular, Zemlin defaults to talking about the United States:

As long ago as the 90s, there were Open Source pioneers like IBM and Red Hat that were creating alternatives to proprietary technologies. In the early 2000s we saw huge billions of dollars of investment in Open Source enterprises, start-ups, and the like. And now, of course, Open Source is needed in innovation in areas like Artificial Intelligence, telecommunications, automotive, and such:

But even more interesting for those who have been watching China for as long as I have – and I have been coming here for over 20 years – is that China also has a long history of investing in Open Source, so DeepSeek didn't come out of nowhere. It came out of the adoption of Open Source in the 90s [did he mean in the United States?!].

Plus, a strategic approach, where companies in China like Alibaba, Tencent, Huawei and others, became leading contributors and adopters of Open Source, cloud technology and other tech.

A late save, of sorts, from the Linux Foundation Director. Though we should add that the Chinese Government has been directly endorsing and adopting DeepSeek technology throughout the administration this year.

But Zemlin was just getting into his stride on the community he fronts from San Francisco:

One other lesson I think is important to understand, on why Open Source is so critical to things like AI, is there's a long history of companies who have gone against Open Source and tried to hold onto their proprietary incumbency. But they ended up either being acquired at a low valuation or simply going out of business.

And companies that initially spoke against Open Source, that didn't go out of business or get acquired, have totally changed their strategy. Microsoft is probably the best example of this. They were early critics against Open Source but shifted, under the leadership of Satya Nadella, to being one of the most pro Open Source companies in the world. And one of the largest contributors of Open Source projects around the world.

So, going up against Open Source is generally a bad idea.

My take

Aka if you oppose us, you will fail – a bold sentiment, indeed, given that there are countless US software companies whose proprietary status has not stopped them from becoming vastly wealthy. But his point about Microsoft was a good one, and we can detect the presence of Nadella in Altman's recent musings.

But the unanswered question in Hong Kong – and elsewhere – remains.

With vendor after vendor, including some Zemlin mentioned, claiming that AI will take the place of all coding, and soon, surely the distinction between proprietary and Open Source is, in a sense, becoming irrelevant. Many AI vendors have simply scraped the Web for code of every kind. Where did some of it come from? Well, who really knows? And soon, who will even have the skills to find out for themselves?

Answers please, on a postcard from Hong Kong.