Myth #44: The future of AI is in the hands of companies.
Philippe Lorenz und Kate Saslow
Myth: National governments are unable to strengthen their own AI innovation base. Private companies buy up national AI resources. States are unable to develop middle- to long-term policy plans regarding strong national AI approaches because policy makers don’t fully understand what AI actually is or means.
Busted: AI technology differs from previous emerging technologies because private sector companies are now innovators in AI technologies. But states can still influence the creation of AI ecosystems where such innovative firms will thrive. This requires providing companies and universities with the necessary resources to produce cutting-edge AI technologies. Private sector actors and state actors merely pursue different roles.
AI is no magic: The industry inputs necessary to produce machine learning (ML) technologies are data, software, hardware, and talent. ( #43). Understanding ML as a composition of these inputs allows policy makers to grasp its economic relevance and recognize the significance of actors who produce and apply this technology. But the rhetoric around AI in policy today remains inconsistent and vague. A deeper understanding of AI is urgently needed. It will improve the debate among the foreign policy community and thus provide the basis for better monitoring of global trends, which can inform policy making at home.
In the AI field, companies matter, as innovation around ML disproportionately stems from private sector entities. Companies to watch focus their research and development efforts on creating specific core ML products, and have a revenue stream based on ML driven applications. Such are true “AI companies” and they push the development and adoption of AI technologies.
But this does not mean that states have no role to play: Foreign service officers are instrumental in gathering and analyzing political and economic matters with regional expertise and then reporting this information back to the government headquarters. The pace of AI development and deployment is fast; its economic importance substantial. But resources to actively monitor global developments are currently limited. However, policy makers around the world have begun to understand the importance of AI and assert themselves into the debate. This is reflected by numerous international initiatives and fora devoted to AI governance.
But these fora and initiatives lack teeth and governments still lack a common nomenclature to discuss and monitor global developments in AI. Leveraging the international network of embassies and the strength of foreign services – information gathering and analysis – foreign services can significantly help government headquarters shape foreign and domestic policy around AI. This can help raise awareness and capacity for more comprehensive and more effective policy making for AI governance. States thus play an important role by providing an ecosystem conducive to AI innovation, and by tracking strategic developments in AI innovation globally.
Truth: States continue to play an essential role in governing AI. But they must be proactive to know what to do. Policymakers today must consider how states can influence industry inputs necessary to create machine learning technologies (data, software, hardware, and talent) and provide governments with a clear perspective on how to defend and enhance their strategic development of AI ecosystems.
Source: Philippe Lorenz and Kate Saslow, Demystifying AI & AI Companies. What foreign policy makers need to know about the global AI industry (July 2019), Stiftung Neue Verantwortung, https://www.stiftung-nv.de/sites/default/files/demystifying_ai_and_ai_companies.pdf.