The Many Races That Will Shape AI


There are two different prisms for investigating the relationship between technology and society. The technological prism views technology as a determining factor in society’s evolution. This prism assumes that once a new technology has been introduced, it will send multiple ripple effects through society impacting power relations, class struggles, geopolitical competitions and even norms, values, and laws. The factory, for example, altered society in a dramatic fashion including mass urbanization, the evolution of a working class, the introduction of statistical models to anticipate and insure against loss of profits due to injuries and an overall reduction in the cost of living. Yet the factory also impacted art with artists such as Andy Warhol transitioning to an assembly line of production. Disease and pandemics in tenement houses led to the first public health initiatives while novels and plays soon focused on the mind-numbing work of those employed in factories. The factory also introduced a new type of human activity – that of surveillance. Floor managers “watched” over assembly line workers like Big Brothers.

The second prism asserts that society develops technologies to accommodate its needs and desires. This societal prism believes that technology will not necessarily reshape society as social institutions can stifle creativity and innovation or move to regulate new technologies. In other words, technological advancements can be restrained. One interesting case study is that of Napster; a peer-to-peer file sharing application launched in 1999. Napster soon became extremely popular as it allowed users to share music files for free. Rather than buy a new record, users could download it from a peer’s computer. However, Napster’s technological innovation was soon curbed by US courts which ruled that Napster violated the copyrights of music companies.  By July of 2001, Napster shut down its network.

Thus far, debates about the possible impact of AI on diplomacy have adopted the technological prism investigating the many ways in which generative AI may reshape the practice and conduct of diplomacy. Some have argued that diplomats will save time and effort as generative AI will draft mission statements, speeches, and UN addresses. Others have argued that through generative AI, diplomats will be able to access vast amounts of data enabling them to better manage crises and gather information on unfolding events across the world. Still others have noted that generative AI will pose a threat to diplomacy given the fact that generative AI will transform users into a troll farm able to create all manners of false information. From fictitious images and videos to false reports and doctored memos, ChatGPT will enable users to mass produce false information on a scale unseen. Finally, there are those who have noted that AI-powered chatbots may be useful in consular diplomacy or even in conversing with social media users.    

While these observations are all insightful, they neglect to embrace a societal prism, one that may account for how social institutions, power dynamics and actors may impact the development of AI-related technologies. If one adopts the societal prism, it becomes evident that there will be four races that will determine if and how AI-related technologies are developed and introduced to consumers.

The first race will be between AI companies, or tech giants. This race has already started. On the one hand, the race is marked by AI companies such as Open AI which developed ChatGPT. Open AI is in no way alone or close to dominating the AI market. In fact, the market is rapidly being saturated with generative AI applications ranging from textual applications that summarize documents to image-based applications that allow users to create new images and artworks. Everyday brings with it the consecutive launching of new AI-powered tools. As AI companies vie over dominance in this growing marketplace, it is unlikely that they will be tempered by ethical considerations or will take time to closely examine the impact of their tools on society. Indeed, it’s fair to assume that today’s AI start-ups will be tomorrow’s Apple, Meta, and Google.

Yet the age of the current giants has not ended and all tech giants including Apple, Meta, Amazon, Google, and Microsoft have either announced mass investments in AI technologies or already launched some version of an AI-based tool. These giants find themselves engulfed in two separate races as each giant tries to outperform its peers while also ensuring that no new companies come to dominate the tech marketplace. In other words, Google is in an AI race against Meta, Apple, Amazon, and Microsoft while also competing with Open AI. With billions of dollars at stake, today’s giants will also rush to the market while launching new AI tools with awesome powers.  Given that these tech giants have already proven a disregard to mitigating the negative consequences of their activities, the tech AI race may lead to truly terrifying outcomes including the demise of reality as we know it. Within a short time, we will not be able to trust any online source, any online image and any online individual as all texts, images and videos may be AI generated. Internet users will also never be sure if they are conversing with a human or a Chatbot.  

The second AI race is already being waged between the geopolitical giants- China and the US. Yet much like the tech-marketplace, today’s giants may be overshadowed by new giants, or countries that will rise to prominence by effectively harnessing the power of AI. Technology has always impacted power relations be it the ability to chart maps and navigate the seas or the atomic bomb. It’s only reasonable to assume that some countries will rise to giant-like status thanks to AI. Here again, ethical and societal considerations will be no match for geopolitical struggles. Countries will invest hundreds of billions of dollars to win the AI race. While this race may result in some positive outcomes, with nations using AI to better health services, upgrade education systems and narrow income gaps, much of this race may have negative outcomes. Especially if nations invest resources in military and intelligence AI applications. The weaponization of AI will make drone armies seem obsolete and primitive.

And yet, there are two other races that may help temper the rapid advancements in AI. One of these is the regulation race. Some governments may choose to regulate technologies as a means of bettering their image and reputation. Indeed, as more and more voices demand some form of regulation on AI development, countries, and blocks such as the EU, are trying to position themselves as ‘regulatory leaders’, the ones ensuring that technology is not allowed to roam free. Even the multilateral system has sought to reinvent itself through AI regulation. UNESCO, for instance, recently published a roadmap to ethical AI while the UN Secretary General has also called for an international approach to AI regulation.  Indeed, by focusing on AI multilateral bodies and institutions may find a new raison d’etre after sustaining many injuries from populist leaders, nationalists, and extremist political parties.

Finally, there is the public relations race. In todays’ world, public opinion still matters. Companies and tech CEOs that are vilified lose money and prestige. This is the reason why generative AI tools suffer from commercial biases. ChatGPT and the like skirt sensitive issues with parent companies fearing bad press and negative headlines such as “ChatGPT blames Jews for Covid19 pandemic”.  This might explain why some AI CEOs are currently in the midst of world tours assuring leaders, governments and parliaments that they are ‘responsible tech actors’. The courtroom of public opinion may thus also limit, to some extent, AI development.

These different races demonstrate how societal actors and structures can influence technological development. Thus, as is the case with all technological advancements, assessing AI’s impact on diplomacy or international relations demands a dual approach. One that analyzes technological affordances on the one hand, and societal agency and existing power relations, one the other.  

Read More