{"id":808,"date":"2026-08-17T14:31:56","date_gmt":"2026-08-17T13:31:56","guid":{"rendered":"https:\/\/blog.lboro.ac.uk\/vice-chancellor\/?p=808"},"modified":"2026-08-17T14:33:23","modified_gmt":"2026-08-17T13:33:23","slug":"the-age-of-ai-societies","status":"publish","type":"post","link":"https:\/\/blog.lboro.ac.uk\/vice-chancellor\/2026\/08\/17\/the-age-of-ai-societies\/","title":{"rendered":"The Age of AI Societies"},"content":{"rendered":"\n

By Nicholas R Jennings<\/p>\n\n\n\n

\"Vice-Chancellor<\/figure>\n\n\n\n

Receiving the IJCAI Research Excellence Award is an enormous privilege and I\u2019m truly humbled to appear alongside some of the most influential AI researchers from the last 40 years.<\/p>\n\n\n\n

The award has given me an excuse to reflect on my research journey. And what is apparent is that many of the issues that now dominate discussions about contemporary AI are precisely those that the multi-agent systems community has been wrestling with for decades. So, while the technology of agentic AI is certainly new, the underpinning science is not.<\/em><\/p>\n\n\n\n

That may sound like a curious claim when almost every week brings another announcement or story about AI agents. OpenAI, Microsoft, Google, Anthropic, NVIDIA and many others are investing heavily in agent platforms and capabilities. Analysts predict an adoption curve unlike anything we’ve seen before. Now, I understand this excitement as I\u2019ve literally spent my entire professional career making this point [1, 2]. But I also think much of today’s debate has a significant blind spot. It assumes that once we have sufficiently capable individual AI systems, the rest is largely an engineering problem.<\/p>\n\n\n\n

I don’t believe that is true.<\/p>\n\n\n\n

The next breakthrough in AI will not come from smarter agents. It will come from understanding intelligent societies.<\/p>\n\n\n\n

In entering this Age of AI<\/em> Societies, the field needs to move on from its current preoccupation with cognitive intelligence: making individual systems better at reasoning, learning and planning. Intelligence alone is not enough to transition AI from answering questions to accomplishing meaningful tasks in the real world.<\/p>\n\n\n\n

Hardly any interesting problems involve a single agent working in isolation. An autonomous procurement agent must negotiate with suppliers and banks. A healthcare assistant must coordinate with clinicians, patients and insurers. A logistics planner must interact with warehouses, customs systems and transport providers. Intelligence is not about solving such problems by bringing everything together into a centralised oracle\u2014it is about interactions between agents representing different participants that possess different information, pursue different objectives and retain autonomy over their own decisions. Single-agent AI is the exception. Solipsism is an edge case and we must focus much more on social intelligence.<\/strong><\/p>\n\n\n\n

Once you view the world in terms of multiple autonomous decision-makers, entirely new phenomena appear. Cooperation and competition. Incentives and institutions. Trust and reputation. Coordination and negotiation. Emergent behaviour that cannot be understood by studying one participant in isolation.<\/p>\n\n\n\n

What strikes me, looking back, is that every important problem we tackled hinged on interactions rather than intelligence. We designed automated negotiation algorithms that allowed independent organisations to coordinate maintenance, logistics and supply chains [3]. We built human-agent partnerships that helped emergency responders to understand an evolving situation and allocate rescue teams after natural disasters [4]. Although these projects addressed very different domains, they rest on the same scientific foundation. The challenge was never simply building a single clever agent. It was to engineer effective interactions between autonomous participants that have both individual and collective goals.<\/p>\n\n\n\n

The distinction between cognitive and social intelligence matters because the next phase of AI will be shaped less by improvements in individual models and more by advances in organising collections of models. This shift from individual capability to collective organisation has many precedents. Human civilisation did not emerge because one individual became exceptionally intelligent. It emerged because large numbers of people learned how to cooperate, specialise, establish institutions, resolve conflicts and build trust. Collective intelligence mattered at least as much as individual intelligence.<\/p>\n\n\n\n

I believe AI is approaching a similar transition.<\/p>\n\n\n\n

For many years, increasing the intelligence of individual systems was the dominant challenge. Today, even though that challenge is far from solved, it is no longer the only show in town. As soon as AI systems begin working together\u2014and alongside people\u2014the organisation of those interactions becomes equally important. The fundamental unit of AI is shifting from individual systems to societies of interacting agents.<\/p>\n\n\n\n

That changes the questions we need to ask. For example, we need to consider how collective responsibilities should be distributed to a team, how we establish trust between agents that have never previously encountered one another, and how humans remain appropriately involved as systems become increasingly autonomous. These are not peripheral issues. They are central to the very future of AI.<\/p>\n\n\n\n

This also explains why I have never found the narrative of humans versus AI particularly compelling. The most successful systems we built were never those that removed people from the process. They were those that combined complementary strengths [5, 6]. Humans bring judgement, experience, contextual understanding and accountability. Agents bring speed, persistence, scalability and the ability to process enormous quantities of information. The goal should not be to replace people with AI. It should be to create multi-agent systems in which humans and AI agents work together to achieve things that neither could effectively accomplish alone.<\/p>\n\n\n\n

Receiving this award also reminded me that research itself is a profoundly collective endeavour. All the contributions I have been involved with happened because of exceptional students, collaborators and colleagues. Multi-agent systems has always been an unusually multidisciplinary community, drawing ideas from economics, social psychology, game theory, operations research, and software engineering. That breadth sometimes made it feel a slightly interstitial field that didn\u2019t fit neatly within conventional boundaries. Ironically, I now believe it may prove to be one of its greatest strengths [7].<\/p>\n\n\n\n

There is a tendency in AI to assume that every technological breakthrough creates an entirely new scientific discipline. History tells a different story. Progress often comes from combining new technological capabilities with scientific ideas whose time has finally arrived. Agentic AI feels exactly like such a moment.<\/p>\n\n\n\n

So, while building intelligent agents has been the first half of the AI story, let\u2019s make building intelligent societies the second.<\/p>\n\n\n\n

\n\n\n\n

References<\/strong>:
[1]\u2002N. R. Jennings, Cooperation in industrial multi-agent systems. World Scientific, 1994. 10.1142\/2257
[2]\u2002M. J. Wooldridge and N. R. Jennings, \u201cIntelligent agents: Theory and practice,\u201d Knowl. Eng. Rev., vol. 10, no. 2, pp. 115\u2013152, 1995. 10.1017\/S0269888900008122
[3]\u2002N. R. Jennings, P. Faratin, T. J. Norman, P. O\u2019Brien, and B. Odgers, \u201cAutonomous agents for business process management,\u201d Applied AI, vol. 14, no. 2, pp. 145\u2013189, 2000.
[4]\u2002S. D. Ramchurn, T. D. Huynh, Y. Ikuno, J. Flann, L. Moreau, N. R. Jennings, et al., \u201cA disaster response system based on human-agent collectives,\u201d J. Artif. Intell. Res., vol. 57, pp. 661\u2013708, 2016. 10.1613\/jair.5098
[5]\u2002S. D. Ramchurn, S. Stein, and N. R. Jennings, \u201cTrustworthy human-AI partnerships,\u201d iScience, vol. 24, no. 8, p. 102891, Jul. 24 2021. 10.1016\/j.isci.2021.102891
[6]\u2002N. R. Jennings, L. Moreau, D. Nicholson, S. Ramchurn, S. Roberts, T. Rodden, et al., \u201cOn human-agent collectives,\u201d Commun. ACM, vol. 57, no. 12, pp. 80\u201388, 2014. 10.1145\/2629559
[7]\u2002I. Rahwan, M. Cebrian, N. Obradovich, J. Bongard, J. F. Bonnefon, C. Breazeal, et al., \u201cMachine behaviour,\u201d Nature, vol. 568, no. 7753, pp. 477\u2013486, Apr. 2019. 10.1038\/s41586-019-1138-y<\/p>\n","protected":false},"excerpt":{"rendered":"

By Nicholas R Jennings Receiving the IJCAI Research Excellence Award is an enormous privilege and I\u2019m truly humbled to appear alongside some of the most influential AI researchers from the last 40 years. The award has given me an excuse to reflect on my research journey. And what is apparent is that many of the […]<\/p>\n","protected":false},"author":727,"featured_media":776,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"lboro_blog_alternative_thumbnail_image":"","footnotes":"","_links_to":"","_links_to_target":""},"categories":[10],"tags":[],"class_list":["post-808","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-articles"],"_links":{"self":[{"href":"https:\/\/blog.lboro.ac.uk\/vice-chancellor\/wp-json\/wp\/v2\/posts\/808","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/blog.lboro.ac.uk\/vice-chancellor\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/blog.lboro.ac.uk\/vice-chancellor\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/blog.lboro.ac.uk\/vice-chancellor\/wp-json\/wp\/v2\/users\/727"}],"replies":[{"embeddable":true,"href":"https:\/\/blog.lboro.ac.uk\/vice-chancellor\/wp-json\/wp\/v2\/comments?post=808"}],"version-history":[{"count":6,"href":"https:\/\/blog.lboro.ac.uk\/vice-chancellor\/wp-json\/wp\/v2\/posts\/808\/revisions"}],"predecessor-version":[{"id":816,"href":"https:\/\/blog.lboro.ac.uk\/vice-chancellor\/wp-json\/wp\/v2\/posts\/808\/revisions\/816"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/blog.lboro.ac.uk\/vice-chancellor\/wp-json\/wp\/v2\/media\/776"}],"wp:attachment":[{"href":"https:\/\/blog.lboro.ac.uk\/vice-chancellor\/wp-json\/wp\/v2\/media?parent=808"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/blog.lboro.ac.uk\/vice-chancellor\/wp-json\/wp\/v2\/categories?post=808"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/blog.lboro.ac.uk\/vice-chancellor\/wp-json\/wp\/v2\/tags?post=808"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}