Making America Great Again may not be as simple as the President hopes when it comes to the burgeoning artificial intelligence competition from China. The recent release of DeepSeek has understandably shaken the confidence of US Big Tech giants, especially OpenAI. It has led, according to the BBC, to ‘reputational downsizing’ and potential market difficulties because of DeepSeek’s demonstrated ability to compete with Western equivalents at lower cost and higher levels of efficiency.
DeepSeek illustrates that despite having access to a relatively small number of graphic processing units (GPUs), not even the latest generation, it is possible to compete with more proficient American models. This is largely because of the more efficient management of the software component. In other words, greater intelligence can compensate for brute force.
And here lies the moral. AI is no exception to the general rule that exemplifies the evolution of emerging technologies: initially they rely on brute force, crudely exploiting available resources, and then gradually they optimise their use through experience-based technical innovations and evolution. But the rules of the AI game are still based on the ability to accumulate more and more powerful GPUs.
Nevertheless, growth in computing power is not linear with respect to energy consumption and other variables, such as the ability to invent new software approaches that define what it takes to develop an LLM (large language model). This is at the core of the cost-effectiveness of DeepSeek and why it poses a problem for its competitors.
Despite the emergence of approaches such as so-called ‘frugal AI’ which promotes more efficient and intelligent use of resources, the belief that more computing power equals more artificial intelligence remains widespread. With a nightmarish compulsion to repeat, the AI industry has adopted marketing strategies that differ little from those practised in other technology sectors. Consider, for example, how PCs, cameras and smartphones are promoted: they vaunt the idea that bigger (the processor, sensor, screen etc.) is better.
This is a much easier concept to understand for those who hold the purse strings and need to loosen them, than unique selling propositions based on mathematical or engineering arguments, let alone those related to software, which are complex, hard to summarise in a few catch phrases, and difficult for users, investors and decision makers to understand.
In the case of AI, this strategy has created barriers to market entry. In fact, the common perception is that to play the LLM game requires such enormous investment and infrastructure that it discourages attempts to do so. By maintaining this approach, the giants of the sector succeed in, on the one hand, monopolising financial resources and, on the other hand, keeping out potential newcomers.
However, thanks to DeepSeek, there is the unexpected realisation that it may be possible to ‘do more but spend less’ and that, as a result, it may not make sense to invest in technologies and hardware that are unnecessarily expensive to buy and manage. This is especially true in view of the similar warnings in the chip sector, where new ARM processors promise affordability and performance.
However, the greater appeal of DeepSeek is true if and only if the development costs are as publicly stated, that is, a fraction of those incurred by US competitors. Although, in fact, the approach to the overall design of the Chinese model is clearly efficiency-oriented, it is unclear whether, and if so to what extent, there has been even indirect support from the Beijing Government. This could be in respect of access to the energy and computing power required to train the model, or other forms of support.
If, in fact, the lower cost of DeepSeek’s development was even partly possible because of state aid, it would be legitimate to raise doubts about whether the project is actually more sustainable than its American competitors, and to ask whether, instead, we are not faced with the use of economic leverage to disrupt the market by lowering the value of competitors. If so, it runs the risk of having to chase DeepSeek instead of dictating the pace and the risk of suffering the introduction of new technologies into the market instead of controlling them. Moreover, they would lose their privileged status as suppliers of ‘raw material’ for the rest of the supply chain that develops LLM-based products and services, since DeepSeek is more efficient, cheaper but, above all, ‘open source’.
For some time now, the concept of open source, a generic term that, essentially denotes the right to access the information necessary to understand a technology, and the right to use it freely, has been moving towards losing its original role as a tool that fosters the free circulation of knowledge to become an important component of states’ geopolitical arsenal.
It’s no mystery that DeepSeek was unashamedly developed in compliance with the guidelines set by the Chinese authorities in relation to how to respond to issues involving socialist values and policies. In this sense, such a choice is the perfect match of the ‘ethical constraints’ embedded in proprietary and open source LLMs already available in the West.
The decision to release Deepseek in open source could not only reduce the value of the AI giants, it also risks removing user market share from them. Companies, developers and researchers might, in fact, be interested in accessing sophisticated technologies without having to pay expensive licences or experience other limitations. We would be facing a complementary situation to the one created by the choice made by Huawei to release HarmonyOSNext (Android’s competing operating system) as a ‘free’ version, capable of running on a wide range of devices, from wearables to terminals, potentially enabling the creation of a global technological infrastructure independent of Western technologies. And it is worth mentioning that Huawei also produces AI GPUs optimised for DeepSeek!
With its wider use, DeepSeek could become part of a strategy for spreading ideas that do not necessarily conform to Western values. Lycurgus, writes Plutarch in Parallel Lives, banished from Sparta all foreigners who had no good reason to stay, fearing “that they would spread something contrary to good customs. Foreigners bring foreign words; these produce new ideas; and on these are built opinions and sentiments whose discordant character destroys the harmony of the state”.
This may be (relatively) simple in the case of humans, much more challenging when it comes ‘new’, or rather ‘different’ ideas conveyed by software that can be duplicated, modified and circulated without any effective restrictions. It is hard not to think of the TikTok squabble, and the reasons, real or supposed, that led the US administration to order its forced sale.
Although a direct, immediate, and large-scale impact of this kind is unlikely, it’s not unreasonable to suggest that in the long run ‘heterodox’ ideas may more easily infiltrate mainstream thinking by passing through smartphone screens, and contribute, if not to redefining it, at least to orienting it in a way that is more favourable to China or, which is the same thing, more critical of our own governments.
Andrea Monti and Raymond Wacks are co-authors of Protecting Personal Information: The Right to Privacy Reconsidered; COVID-19 and Public Policy in the Digital Age; and National Security in the New World Order: Government and the Technology of Information.
To join in with the discussion please make a donation to The Daily Sceptic.
Profanity and abuse will be removed and may lead to a permanent ban.