The AI Summit London 2023: Omdia Takeaways

The AI Summit London opened its doors in mid-June, bringing together thousands of business leaders, technology vendors, and industry experts at the distinctive Tobacco Docks venue. Attendees to the two-day event had the chance to learn about emerging technology trends, interact with the latest products in the market, and hear about best practices from practitioners as they deploy and scale AI throughout their organizations. Perhaps most important aspect of the event was the opportunity to connect and network with a veritable who's who of the AI market.

So, what were the hottest topics discussed in London? It is probably no surprise Generative AI (GAI) was at the forefront of many conversations given how this transformative technology has captured the attention of much of the world in recent months. At the AI Summit London, GAI was discussed during presentations and panels at the various event stages, in the Omdia suite, and at the VIP-only VisionAIres Village.

Summarizing the vast number of conversations held in London about GAI in just a few sentences is impossible. It is further complicated because it is still early days for GAI and as a result opinions and views on the topic, even among industry experts, often differ. However, one thing that is clear is a wide range of potential use cases, from text generation to automated code development, will be enabled by GAI. The wide reach of these use cases and applications, along with new ones that will emerge as commercialization of the technology grows, suggest virtually every industry and enterprise will soon be impacted by GAI.

Though the excitement around Generative AI was obvious at the AI Summit London, it was coupled with the understanding this must coincide with a shared responsibility to use the technology properly. Enterprise adopters frequently spoke about the challenge of finding the correct mixture of aggressively adopting GAI to gain a competitive advantage while also ensuring they have proper guardrails in place. The Tobacco Docks provided the prefect venue for leaders to speak about this dilemma face-to-face with counterparts facing similar issues and find solutions (and in at times partners) that can help address this concern.

Government agencies are also playing a role in the responsible deployment of AI. In fact, on the opening day of AI Summit London the European Union took a major step in passing a draft law known as the A.I. Act, one of the first major laws to regulate and put restrictions on what are seen as the technology’s riskiest uses. The impact of this regulation, if it comes into effect, and its potential to serve as a model for other policymakers around the world, led to some fascinating conversations at the event that will undoubtedly continue in the months ahead.

As you might expect, all these topics and issues kept our Omdia Research Team quite busy at the event. If you were in attendance in London you had a chance to hear and engage directly with our analyst team at various sessions, panels, and our Lunch & Learn presentations. If you missed the event, below are some of our analyst takeaways from the show and what they mean for AI market.

“Generative AI, and ChatGPT in particular, highlighted the need to move fast with putting in place the necessary safeguards -- aka regulation and governance -- for this disruptive technology. Many of the issues we face today are not new; they include toxic data of questionable provenance, model discrimination and bias, lack of oversight, accountability and transparency (“black boxes”), privacy and copyright infringements, detrimental impact on the environment, and so on. ChatGPT and its cousins (Stable Diffusion, Midjourney) and their descendants exacerbated all of this and raised concerns that this technology can potentially disrupt the sheer bedrock on which our societies, business and commerce rest – that of trust. If we no longer can distinguish the real and true from deepfakes and hallucinations, if we don’t know whom or what to trust, at best we may not be able to fully benefit from this transformational technology; at worse, we may face societal collapse. Lots of conversation at the Summit focused on what the right approach to regulating AI should look like, with examples from the UK, European Union, US, Taiwan and others.” – Natalia Modjeska, Research Director, AI, Omdia

"I spent some time during the conference speaking with AI practitioners, both those solving pressing data-related issues and those exploring future opportunities with generative AI, and to a person, one message reverberated across both ends of this problem/potential spectrum: Doing AI in the real world remains a complex and difficult undertaking, one where there are no best practices, only practices that seem to work for a given situation at a certain time. And what practitioners need most is not the latest shiny, technological gizmo, but rather access to the insights and expertise of their colleagues as a means of becoming either unstuck or better grounded as they traverse their own AI landscape of the as yet unknown." – Bradley Shimmin, Chief Analyst, AI, Omdia

“The AI processor space is resolving itself into a data center market where NVIDIA’s consistent long-term investment in software has given it a hegemonic position, and an edge market that is much more diverse and innovative. The fundamental trade-off is between optimization for efficiency, on one hand, and for flexibility and developer productivity on the other. Since the beginning of the computing industry, developer productivity usually trumps. That said, the power and thermal costs of AI will eventually bite – the most power-intense GPU is now 750W, and the highest power configuration for OAM modules is 1kW.” – Alexander Harrowell, Principal Analyst, AI, Omdia

“There were a lot of interesting sessions on responsible AI and regulation which demonstrates the importance the industry is placing on “getting it right”. In the panel that I chaired, we discussed the importance of the responsible and ethical use of AI, principals such as: safety, transparency, privacy and security, inclusivity, fairness and accountability.  We discussed the importance of these principals as the foundation upon which to build AI implementations within the enterprise and should be a precursor to the development of specific use cases and model development. The panel also highlighted that diversity is key in AI development teams to help mitigate model bias and support inclusivity in model outputs. Furthermore, the panel discussed the importance of including representatives from business units in the early stages of AI projects to facilitate the development of appropriate KPI’s and alignment with organizational goals. I am looking forward to next year’s London Summit for further sessions on these and other topics as the pace of development in AI shows little sign of slowing down.”  - Andrew Brosnan, Sr. Principal Analyst, AI, Omdia

If you want to find out more about the latest trends in the field of emerging technology, then make sure you join us at Applied Intelligence Live! Austin this September 20-21, at the Palmer Events Center, Austin.