By Tom Fowdy
Last week, Chinese Artificial Intelligence (AI) company Moonshot released the third edition of the Kimi app, known as K3. The founder of the firm, Yang Zhilin, had pursued his technology career in the United States and worked for big magnates such as Google, that was before he returned to China amidst a McCarthyist witch-hunt against Chinese scientists and technology professionals. Like they say, the rest was history.
AI is deemed one of the most critical technological resources of humanity's future. Despite the backlash against it in the creative sphere, it isn't going away, but is hastening a permanent economic revolution in the same manner industry itself did, and the internet. While I respect the authenticity of human talent in aspects such as art and film, those who witch-hunt "AI Slop" are akin to people in rowing boats complaining about the invention of steamships. The world is not going to hold itself back just to suit these sensitivities.
As a result, a fundamental aspect of American foreign policy over the past few years has been to try and contain China's rise in Artificial Intelligence capabilities and to ensure that they maintain a sufficient lead in the technology. AI is not about talking to chatbots, it is increasingly becoming the engine for policy and strategy. The Trump administration infamously plotted the capture of Venezuela's Nicholas Maduro using a private version of Claude, while they are also doing their best to strongarm OpenAI into supporting the US Department of Defence. Thus, Washington has attempted to try and arrest China's approach by depriving it of the critical chips created by boom company Nvidia, to ensure they can't reach the same level of sophistication. Similarly, apps such as ChatGPT have been banned from being accessed from China and Hong Kong based Ips.
Somehow, this isn't holding China back. Again and again, Chinese companies have repeatedly held their own in producing equal, or even more powerful AI applications despite having weaker chips.
The first shock came from DeepSeek, a chatbot which, according to outlets has advanced reasoning capabilities which rival its US counterparts, its initial release causing market shock in the United States for as the BBQ reported: it "requires a fraction of the cost to train and run" compared to US models, proving America's chip embargoes are not having an effect. Now it is launching its fourth edition.
The second came in the form of ByteDance's SeeDance 2.0. SeeDance 2.0 is an AI video generating application that is integrated with software such as CapCut. SeeDance is capable of generating ultra-realistic cinematic scenes on a whim, suppressing anything any Western AI video generator has ever made, and does not fall victim to the obvious telltale AI signs shown with applications such as OpenAI's Sora, Kling AI or Grok imagine.
Clips soon went viral of users having prompted the platform to create a convincing alternative ending to Avengers Endgame in 2019 where Thanos apologises.
Then the third shock cam Kimi K3. Kimi K3 is an absolute juggernaut whose coding and organizational capabilities easily outgun ChatGPT and Claude. Its ability to code applications from nothing is unprecedented. As one example, I tried out the paid subscription on Saturday and within just hours I created a comprehensive and completely functional working computer game engine, one which was also capable of correcting any detail or issue on a whim.
Posts subsequently went viral on X showing how the application purportedly recreated a Windows XP or MAC OS capable of playing commercial games such as GTA on it (although this should be treated with some scepticism), amongst others. The app has now had to suspend new subscriptions due to overload. Kimi K3 is getting close to giving the average person an almost godlike level of creativity on a computing level, showing that China is nowhere near being contained in the "AI race." It is likewise fanciful to resort to the cliché that "China steals intellectual property" if they are repeatedly capable of producing products which are equal to or better their Western counterparts. One of the much-touted strengths as to how Chinese companies are managing this is through the creation of Open AI models that are collectively used, as opposed to being strictly enforced corporate secrets like in the US.
These developments might teach us that the global race for AI supremacy is not going to be won by containment, embargos, or "choose us or them," policies, but innovation, genius and pure competitiveness in and of itself, ironically in the way America would in theory endorse. The global market will not choose artificial intelligence apps just on pure geopolitical considerations, even if some countries are forced to, but they will naturally seek out the best and most powerful products, regardless of the nationality or ideology of its creator. The US can forcefully deny that to itself, and perhaps some of the usual suspect countries too, but most nations who want "best of both worlds" won't be so naïve.
AI is firmly here to stay; we are entering a new historical revolution now, one will produce rapid social and economic change, along with new ethics and challenges alike. China is not trailing this game, they're leading it, and while we might call its AI brilliance a "Sputnik Moment," it's quite obvious the old playbook of Cold War containment, the American assumption that China "can't innovate," is outdated and badly in need of revision. Let's see where it all goes.
The views do not necessarily reflect those of DotDotNews.
Read more articles by Tom Fowdy:
Opinion | The US retreats over Hong Kong
Opinion | The Iran war Merry-go-round continues, when will it end
Comment