The AI Summit London showcased the leading global innovators pushing the boundaries in artificial intelligence for both business and societal outcomes.
A centerpiece event in London Tech Week, huge advances in technology across artificial intelligence were brought centre stage, underpinned by a celebration of knowledge sharing, networking, and breaking insights into technology trends.
According to Omdia 2022, AI adoption is sitting at a 25% scaling on the diffusion of innovation curve but it was clear from this summit, especially announcements from Intel® with whom I am honored to be an ambassador, that the trajectory, scale and scope of adoption is accelerating across multiple sector verticals. This includes rising relevance for FinTech, Insurtech, and eCommerce companies with the rapid acceleration in scrutiny regarding how an individual’s data is used.
As a judge for the AI Journal Awards, I was especially impressed with the focus on student outreach, with a superb Hackathon showcasing youth talent in AI and Data Science alongside John Lewis & Partners sharing about their innovative apprentice program for AI and Machine learning – brilliant activities! It was also excellent to see the attention afforded to issues and risks around privacy, especially with the imminent demise of third-party cookies, plus AI ethics and the potential for inherent model bias. This included highlighting the role of synthetic data to help address these challenges alongside negating AI data scarcity for training Artificial Intelligence models as currently experienced by some organisations. Indeed, Gartner predicts that 60% of all data used in the development of AI will be synthetic rather than real by 2024.
Beyond this, a number of critical announcements were made at the AI Summit event by Intel® and I believe these are key to help move beyond ‘data to insights’ and from businesses investing in AI, to ultimately reaping its full benefits, deployed in production environments. From a hardware perspective these included Intel’s evolution of the GPU line and expansion of Xeon® to integrate even greater AI capabilities, through to bigger steps still in software. This includes the ‘SoftwareFirst’ Initiative underpinned by 6 pillars of innovation, all new Open Source Industry Reference Kits and new Industry Frameworks – in particular, Intel® is now default in Google TensorFlow giving millions of developers 3 x or more performance gains - without making any code changes.
Reflecting on the Open Source Industry Reference Kits in more detail, I believe these are critical to the democratisation of AI, making it both more straightforward and easier to deploy. Built on the OneAPI parallel programming framework which is open, cross-industry, standards-based, unified, multi architecture and multi-vendor by design, these kits are targeted at specific verticals notably utilities, healthcare and manufacturing. Built by Intel® in collaboration with Accenture as part of Project Apollo, these kits, which are available now on GitHub, also contain code and data repositories, benchmarking results and solution briefs to reduce time to deployment, whilst also improving performance, seamlessness and Total Cost of Ownership.
Another example of Intel® and Accenture in partnership was brought to life at the Intel® Booth, namely Computer Vision and AI-based scalable scratch detection, using Intel® ’s oneAPI SW toolkit and Intel® NUC products. Typically defect defection remains highly manual when building cars and other products within the manufacturing sector - and this can risk issues being missed. Earlier identification means a higher production yield at the highest possible quality. And with Intel® processors found everywhere, this does not necessitate a brand new GPU because AI and Machine Learning workloads can be processed using extant Intel® CPUs alongside vino technology to drive production line improvements early. Or in other words, a scratch can be scanned, observed and flagged as new during a manufacturing process and all using your existing hardware in real-time – flexible, sustainable, efficient and low cost!
With collaboration and the power of the ecosystem so key to AI innovation at scale, it was also superb to see the announcement about the partnership between Hugging Face - the reference open source in Machine Learning - and Intel®. With more than 5,000 organizations already using Hugging Face today, I love how this is bringing the latest iterations of Intel® Xeon® hardware and Intel® AI software to the Transformers community, through open-source integration and integrated developer experiences.
Again, the democratization of Artificial Intelligence comes to the fore, empowered by Intel® and Xeon being ‘already everywhere’ affording End to End Openness, Performance, Security, Productivity and Partnership – by design. Intel® has a unique capability to help build and deliver the full range of AI tools and solutions in the market, from language and model optimisations, embedded security, open-source development tools, right through to full AI solutions or business partnerships.
I discuss this and more in a podcast special with Dr Wei Li – Vice President and General Manager of AI and Analytics (AIA) at Intel® - entitled ‘Accelerating AI Everywhere with Software Solutions - and which is now available here. And with a shared passion for all things ‘Tech For Good’ we also share key considerations around ethical AI development, addressing diversity in tech talent gaps, and applying AI for Social Impact at Scale, with many tangible examples. To find out more about Enabling AI Everywhere by Accelerating the Open AI Software Ecosystem, further information is also available here.
Additionally, discussions and presentations at the AI Summit helped bring the role of Federated Learning centre stage to address AI data challenges, especially in sectors such as Medicine, Defence and Telecommunications. As brilliantly explained in this video, Federated Learning is an approach to Machine Learning in which the training data is not managed centrally but rather in distributed and private datasets.Data is retained by the participating parties in the process and not shared with any other entity, something which may be prioritised for a privacy, security, regulatory or practical optimization rationale.
Federated Learning heralds a step change forward in negating the challenge of working with sensitive data, data that is too valuable to share and data silos, alongside working across various parties to develop a shared Machine Learning algorithm, and the increasing need to be able to adapt data in real-time to optimize conversions automatically. Given this context, it then perhaps comes as no surprise that the Global Federated Learning Market size is poised to reach some $198.7 Million by 2028 (KBV Research 2022).
More information on how Intel® SGX helps protect data in use provides further background, with details on how Intel® OpenFL and Intel® NUCs can perform Federated Learning in a secure and easy- to- deploy manner accessible here. And there are already many live examples of its benefits in use, including the healthcare partnership between Intel® and Penn University, and advancing the Federated Tumour Segmentation Initiative. An amazing collaboration for good!
I was also delighted to interview Walter Riviera the AI EMEA Technical Lead at Intel® in a podcast special on ‘Federated Learning: Redefining AI at Scale’ which can be viewed here. Filled with more insights, future thinking on the AI and Federated Learning trajectory ahead and more Intel® impact examples, including with NASA, I hope the collective knowledge sharing will help inspire your next steps in Artificial Intelligence and Federated Learning and the impact this can bring for business and broader society.
As we have seen, AI is a software problem but hardware remains an essential component of the solution too — as Intel® exemplifies, we must cater for both alongside the needs of developers and ecosystem to truly democratise its benefits and I am excited about the next iteration of innovation yet to come!
All feedback and follow up questions most welcomed, Thank You Sally
Prof. Sally Eaves is a highly experienced chief technology officer, professor in advanced technologies, and a Global Strategic Advisor on digital transformation specializing in the application of emergent technologies, notably AI, 5G, cloud, security, and IoT disciplines, for business and IT transformation, alongside social impact at scale.
An international keynote speaker and author, Sally was an inaugural recipient of the Frontier Technology and Social Impact award, presented at the United Nations, and has been described as the "torchbearer for ethical tech", founding Aspirational Futures to enhance inclusion, diversity, and belonging in the technology space and beyond. Sally is also the chair for the Global Cyber Trust at GFCYBER.
Dr. Sally Eaves is a highly experienced Chief Technology Officer, Professor in Advanced Technologies and a Global Strategic Advisor on Digital Transformation specialising in the application of emergent technologies, notably AI, FinTech, Blockchain & 5G disciplines, for business transformation and social impact at scale. An international Keynote Speaker and Author, Sally was an inaugural recipient of the Frontier Technology and Social Impact award, presented at the United Nations in 2018 and has been described as the ‘torchbearer for ethical tech’ founding Aspirational Futures to enhance inclusion, diversity and belonging in the technology space and beyond.