From Generative AI to Changing Lives: The AI Summit London 2023 Recap

After a whirlwind of half the year filled with the most mainstream attention AI has ever had, it is no surprise that the conference “where Commercial AI comes to life” was absolutely buzzing last week. During London’s first heatwave this Summer, the AI Summit London 2023 brought together visionaries, thought leaders, and industry pioneers to the beautiful Tobacco Docks venue to discuss the transformative potential of artificial intelligence (AI). 

Let’s dive into some of the key highlights and trends that emerged from this influential AI conference.

Generative AI 

One of the most intriguing and widely discussed topics at this year’s AI Summit was generative AI, specifically using Large Language Models (LLMs). Generative AI refers to artificial intelligence’s ability to create “original” content in the form of words, images, music, and much more. The field of generative AI has been transforming businesses across several industries, from art and entertainment to customer service to healthcare. 

At the moment, we are even using generative AI to create more efficiency across several different departments at Seldon, from the tech room to the sales office. It is a technology that is trending both for our customers as well as our colleagues.

During the panel discussions and featured speaker sessions, the recurring mention of ChatGPT as a prime example of generative AI’s incredible potential. In the last six months, ChatGPT, an advanced LLM developed by OpenAI, has garnered worldwide attention as it makes the use of AI and natural language processing (NLP) more accessible to the everyday user.

I attended a Masterclass titled, “The AI Factor-How AI and Generative AI Is Transforming Our World” by Asha Saxena, Founder and CEO at Women Leaders In Data & AI. Asha presented on the impact AI and generative AI have had on several different industries.

She emphasized that AI’s aim has always been to assist and improve human life. She noted that there’s rapid progress in LLMs like ChatGPT evolving into artificial general intelligence (AGI) however we are far from having AI reach that level of consciousness.

Some notable use cases included Netflix successfully creating original content based on data (leading to the fall of Blockbuster), Dominos considering themselves to be a technology company delivering pizza (surpassing Pizza Hut in sales), and robots saving loneliness felt by nursing home residents.

Asha’s talk focused on both use-case examples as well as practical tips AI practitioners can take home with them. She introduced the power of the “data-readiness framework” and stressed the importance of making platforms easily consumable for customers, referencing Amazon’s one-click purchase as an example. Any product development process needs to be scalable, repeatable, and adaptable to work. This is a key concept that we’ve been implementing since our Series B funding round back in the Spring. 

The talk concluded with Asha highlighting that to be successful in the AI field, you need to be committed to continuous learning. Learning doesn’t stop. She encouraged us to invest in lifelong learning to keep up with the rapidly evolving AI industry.

Scalability

This year, there was a stage at the AI Summit titled, “AI at Scale”, an area in which our customers come to us for guidance. Scalability emerged as a top priority and a key topic of discussion among industry experts and attendees alike. Scalability in AI refers to the ability of an AI system to handle increasingly large workloads, data volumes, and user demands without significant degradation in performance or efficiency.

During the summit, speakers and panelists highlighted the challenges and opportunities associated with scaling AI initiatives. They emphasized how important it is to develop scalable AI models, infrastructure, and systems to ensure reliable and efficient model deployment in real-world environments.

One of the most interesting panels was titled, “Data-driven Disruption in the Digital Media and Streaming Industry: Is the Industry Ready?” The talk included Gilbert Hill, Privacy Technologist and CEO at Data and Marketing Commission, Alberto Rey Villaverde, Executive Director of Advanced Analytics and Data Science at Virgin Media 02, Matthew Healey, Lead Machine Learning Engineer at Bumble, and Brian Leonard, Head of Engineering – Production & Workflows at IMG Media.

AI pioneer Andrew Ng said, “Data is food for AI.” He recently has been advocating for data quality and emphasizing its importance over the actual ML models themselves.

The panel here all agreed with Andrew’s sentiment and spoke about data quality being a top priority at their respective companies. 

The challenge for a company like Bumble is to stop all the bad actors of bad data. Especially when looking to scale ‌machine learning capabilities, a big factor for whether you’ll be successful is the amount of trust users have in your company.

The EU AI Act was an upcoming regulation that was mentioned in several talks, and this one was no exception. The panel spoke about the EU AI Act’s requirement for human intervention.

Human intervention, especially during the polarizing hype for AI at the moment, is crucial for trust to be built. AI is helping humans understand humans more, but the question is, do humans actually want to be understood by the technology?

If they do, it opens up a world of possibilities for users in the digital media space.  Conversational AI could be an opportunity to stop the hours of trying to choose what film or show to stream on a weeknight. More intentional profiles could be built on dating apps. And the way these digital companies communicate with their customers could become hyper-personalized. 

User sentiment will undoubtedly be used as a compass for organizations that are scaling their AI capabilities. Lose user trust and the speed of scale can be slowed down or even stopped altogether.

Ethical AI

Ethical AI refers to developing and deploying AI systems that adhere to ethical principles and values. It means designing AI algorithms, models, and applications in a way that considers the potential impact it could have on both individuals and society‌. 

It was a talking point that was heard on several different stages, which shows that the community‌ has ethical AI considerations at the forefront of their minds.

From a regulatory point of view, AI is a full time job to keep on top of. One of the panels that was focused on the EU AI Act was titled, “How Should Businesses Respond to the EU’s First-ever AI Law?” and included Coran Darling, Associate (Intellectual Property & Technology) at DLA Piper, Marco Mendola, Community Builder and Co-host at The Law of Tech, Kerry Sheehan, Head of Service Development and Innovation at UK Civil Service, Var Shankar, Director of Policy at the Responsible AI Institute and Racheal Muldoon, Barrister at Maitland Chambers.

The panelists emphasized how important it is to start planning and preparing as a response to the new EU AI Act even though it has not been fully finalized yet. They referenced the Google AI Pact, which indicates that industry leaders are already taking steps to address AI ethics and accountability. 

The AI Pact is a set of voluntary rules to be put in place while the formal EU AI Act regulations for applying AI are still a work in progress. Google’s CEO Sunar Pichai is working with lawmakers in Europe on this initiative.

There are yet to be details on what the AI pact would include; however much like any self-regulatory arrangement, it won’t have any legal consequences for failing to meet the voluntary commitments.

The EU certainly wants to be “proactive”. The panelists discussed how a standardized approach across countries and regions could lead to more certainty, stronger consumer trust, and the legitimization of the AI industry worldwide.

Ethics in Healthcare

Ethics were also a topic of discussion at a very different talk held towards the end of the first day of the AI Summit. The talk was titled, “AI to Beat Ear Disease in Rural and Remote Indigenous Australian Children” and presented by Narinder Singh, Chief of Otolaryngology at Westmead Hospital, Sydney.

With personal experience growing up in a disadvantaged community, Narinder emphasized the critical role of ethical AI in improving the lives of those challenged with ear disease.

While the potential for AI in healthcare is immense, there’s still several barriers to its adoption. Privacy regulations are a serious challenge, as are ethical considerations and liability concerns. 

In order to overcome these obstacles, healthcare clinicians and data scientists need to collaborate, which is a task that’s not always straightforward. Narinder urged interested data scientists to explore this field, as the ability to make a societal impact is enormous.

By starting with the problem and working their way backward, Narinder’s team strives to create an inclusive and ethical approach that uses the power of AI to improve health outcomes for disadvantaged communities.

Looking Forward to the Future of AI

It’s hard to say what was hotter–the topics being discussed at the AI Summit London 2023 or the temperatures. The conference provided a vibrant platform for exploring the transformative potential of artificial intelligence. The AI Summit explored key trends such as generative AI, scalability, and ethical AI.

Talks on generative AI highlighted the real-world impact that LLMs like ChatGPT have in various industries. Experts stressed the need for AI systems that can handle increasingly complex workloads and user demands. Plus, it is very clear that AI practitioners are focusing on developing and deploying AI systems that follow ethical principles and values. It was such a well-organized event, we can’t wait to attend the AI Summit 2024.

The last six months were hugely transformative for AI. Where will the next six months in AI take us?