Introduction
Web Summit is the world’s largest tech conference and something Lynchpin have been keen to attend and experience. Finally, a few weeks ago, we were able to head across to sunny Lisbon and visit Web Summit 2023. We were keen to understand the latest trends and conversations and to understand how that impacts the world of data and analytics.
One of the first things you notice is the scale of the conference – around 70,000 delegates packed into the Altice Arena overlooking the sea in Lisbon. That’s a lot of people, matching the crowd we witnessed the day before as we luckily attended the Lisbon football derby contested by Benfica and Sporting Lisbon – a thrilling game. As football fans, our curiosity was piqued when a number of famous footballers from the past, including world cup winners, were presenting at the show! We both experienced a thrilling finale to the game, so our hopes were high for the conference.
Football was not the main draw of the event, however. Across multiple pavilions and one arena, there were a staggering amount of talks, presentations and exhibitors spanning around 27 distinct tracks – from content creation strategies to the evolution of finance to future societies there was certainly enough content to suit all tastes! Where else can you get such diverse talks from ‘LLM managing your supply chain’ to ‘the future of urban air mobility’ and just about everything in between…
The shape of the industry
As is always our way we can look to the numbers to help tell the story of the conference. With over 70k attendees from across 150 countries and 43% of female representation, diversity seemed assured, however a lot of the talks still highlighted the reality of this diversity struggle in our everyday lives and in particular within a tech-driven industry.
In a ‘Women in Tech’ driven conversation on “Capitalising on diverse perspectives as a competitive advantage”, the frustration was evident – we’ve been talking about equality for so long “why hasn’t the needle moved yet?”.
With an estimated 30% of women across the world working in tech it is still a stark contrast of representation in leadership roles. Less than 10% of founders are women and only 2% of VC money go to businesses driven by female founders. Tara Chklovski – Founder and CEO at Technovation shared that there are 11 million men in technology but only 3 million women who are technology professionals. The inequality is not just with the gender gap, representation from cultures across the globe is still poorly lacking too.
Working with AI
- AI maturity
With the introduction of Chat GPT into the mainstream, it came as no surprise that AI was the standout topic at Web Summit 2023. Such a vast amount of talks, roundtables and exhibitors were leading with an AI message.
We’ve been to many events, many of which discuss AI and machine learning, but it did really feel like a tipping-point has been reached where we moved past the bloated hype and unrealistic expectation and focussed on more practical concerns and challenges – especially around bias, privacy, regulation, automation and jobs.
During a talk with Albert Wenger – Author of ‘The World After Capital’ and Managing Partner at Union Square Ventures, the concept of AI being part of our evolution was discussed readily.
The worry of machines taking over our jobs is not a new challenge, long gone are the days when we spent most of our time working the fields, and now is the time to embrace what AI can bring us to enhance our lives. Using AI for automation and freeing us up to spend time on the things we love and enjoy, to be more creative and spend time on more rewarding tasks. While some of Albert’s opinions may be considered unpopular, this particular sentiment did seem to be common across a few talks in relation to how we use AI. Manny Medina – Co-founder & CEO of Outreach noted: “technology is supposed to buy us time so we can spend more time doing human things”, however, we need to be careful because it can also be disruptive and distracting, and as many of us are aware, detrimental to mental health.
- Data quality / Model quality
Many talks focussed on data and how that fuels AI systems. A lot of discussion centred on data quality and how important that is to developing good models. Whereas you can buy compute power, algorithms are often public (via open research) – the need to train models on good quality data was a consistent theme and a key differentiator. Synthetic data came up in many talks, particularly around its role in improving bias in real-world data. Douwe Kiela, Co-founder and CEO at Contextual AI, noted that the use of synthetic data opens up the possibilities to deprogram the bias from our inherent data.
- AI in practice: Customisation and adaption
A lot of companies are re-purposing general AI models for specific use-cases – success for those were impacted by quality of data and smaller models but with focussed evaluation. In one talk from Christina Kosmowski, CEO at Logic Monitor, we heard about the importance of coverage, context, and correlation (orchestration of data) as key enablers, with some great football analogies to compare success of AI solutions and football team success with Messi compared to a generative AI system, taking prediction one step further to provide recommendation.
Moving forward with AI
- Using your best judgement
In essence, there was a consensus about the positive contribution for AI as we head into 2024, with some core lessons as they relate to privacy, security, bias and where the models might fit into business decision-making processes, but also an acceptance that AI, as a term, can mean many things today, and that’s ok. That reflection helps us to manage the hype and the expectations that often come with these hype cycles and there is a very clear recognition that AI is not the magic bullet, it does not know all and will not solve all problems. In fact Cambridge Dictionary has updated its definition of ‘hallucinate’ to account for the new meaning within AI – “When an artificial intelligence (= a computer system that has some of the qualities that the human brain has, such as the ability to produce language in a way that seems human) hallucinates, it produces false information.”
But perhaps we will keep moving forward at an incredible pace with Google’s new AI model Gemini claiming to have advanced “reasoning capabilities” to “think more carefully” when answering hard questions, perhaps!
Decision-making and the role of AI was discussed a lot during the conference. Cassie Kozyrkov, CEO at Data Scientific and former Chief Decision Scientist at Google, talked about the importance of human judgement in decision-making – what skills are needed for decision leadership in the AI era. She gave some good advice, if you ask the question: ‘what would it take to change your mind?’ you’ll likely be better at making decisions. Good questions to ask AI systems are: ‘what does success look like?’ and ‘what are the examples we are supposed to base our answers on?’. Important considerations when thinking about ethical and effective applications.
- Centralisation / decentralisation
In a broader sense, some interesting keynote talks where the role of AI in society was discussed. Meredith Whittaker, co-founder AI Now Institute, expressed some fears about the fact that the technologies involved are often re-purposed from those developed by a small number of very large companies and that those companies are focussed on shareholder outcomes rather than the social good. Some concerns too about the re-enforcement of a ‘surveillance business model’ and the need to have good privacy regulations to protect people. As AI becomes more impactful within our world, such sentiments become increasingly consequential. Food for thought!
Conclusion
Looking back, it was difficult to get to all the talks we wanted – even just getting from location to location was a challenge given the volume of people and the size of the venue. With so much content, it’s all too easy to spend time in analysis paralysis looking at the schedule.
Planning ahead of time is definitely recommended but the opportunity to hear from a wide range of perspectives, the opportunity to network and to be immersed in the conversation was definitely worth it. So be clear about your expectations, there is no way you can get through all that content – choose carefully!
And from a Lynchpin perspective it was very rewarding to get a little time out from the day-to-day and come together on some fresh perspectives about the themes from the conference and how we can apply some novel thinking to our clients and their challenges in a holistic sense. And Benfica has two new fans!
As we head into 2024, we are mindful of how AI / machine learning will increasingly influence decision-making in relation to data and analytics strategies across all verticals. But I leave you with a quote from Roberto Carlos – “My philosophy has always been to solve problems with a smile on my face” – let’s see if AI can rise to this challenge from the great man himself.
About the authors
Anne Sargeant
With 20+ years of analytical experience, Anne Sargeant heads up Lynchpin’s Data Science team.
Anne has a degree in Maths and Physics and a master’s degree in Statistics. She is an expert in data mining, predictive analytics, and statistical modelling – with a passion for driving business success from their effective deployment.
Prior to joining Lynchpin in 2013, Anne worked as a senior analyst for Experian, Symantec, Avis and SPSS.
Gary Douglas
Gary Douglas has extensive experience in consulting to build data and analytics strategies and solutions for leading organisations across the UK, EU, and US.
Gary has helped businesses to transform their existing data and build new processes and workflows to enable businesses to scale their ambition through a data-driven mindset – increasing sales, reducing costs and supporting improvements to the customer experience.
Having worked both agency and client-side, and with significant experience in Financial Services organisations in the UK, Gary brings a practical knowledge of successful data and analytics structures.
Building on a PhD in particle physics, Gary understands the challenges from a C-level through to a practitioner-level view.
About the author
Lynchpin
Lynchpin integrates data science, engineering and strategy capabilities to solve our clients’ analytics challenges. By bringing together complementary expertise we help improve long term analytics maturity while delivering practical results in areas such as multichannel measurement, customer segmentation, forecasting, pricing optimisation, attribution and personalisation.
Our services span the full data lifecycle from technology architecture and integration through to advanced analytics and machine learning to drive effective decisions.
We customise our approach to address each client’s unique situation and requirements, extending and complementing their internal capabilities. Our practical experience enables us to effectively bridge the gaps between commercial, analytical, legal and technical teams. The result is a flexible partnership anchored to clear and valuable outcomes for our clients.