EGG conference: what did we learn about data?

After all businesses in all industries are trying to understand how, and when to dive into AI, and this year’s conference in London provided insight and cutting-edge data products and solutions which allows you to stay ahead of the competition.

Sue Daley of TechUK was the first speaker to start the day, and went straight in at the deep end by explaining that AI is not new and has actually been around for a while, but what is new is the power we have to make AI happen.

Daley said: “The power of AI and what we are moving towards is a combination of technologies such as 5G, and what these technologies can do together. They have the potential to be amazing and it’s going to affect every aspect of our lives.”

Are we brave enough to embrace the power of AI? Daley believes we need two things for this; confidence and vigilance. We need to get this right for the UK, Daley explained: “The day-to-day basic sciences are so important, and we need to realise that we have a digital skills gap in this country.”

To address this gap, the key skill the UK is missing is confidence – we need to talk about what these technologies can do for our country and need to encourage them more.

Leading on from this Florian Douetteau, CEO at Dataiku, explained that machining learning is a team sport. For example, just one superstar on their own won’t save the day, you need to work together as a team. But do we need a superstar data scientist to lead the way? The answer is we are all one.

Douetteau used a great metaphor comparing machine learning to football: “Even if you don’t have the best player in the world, a team playing together is better than any Neymar or Ronaldo alone.”

Motivation is the 12^th player on the team, and as a team Douetteau emphasised it is important to be able to keep each other motivated. “You never know how it’s going to end, so you need to be prepared for extra time, to keep working. When it comes to machine learning, you need to keep pushing projects until the very end.”

He finished with stating that at the end of the day only results matter: “However arguing about it can be fun! Machine learning is about making it work but arguing afterwards.”

As the day and talks progressed we moved on to combining the technology with data culture. What is more important culture or strategy? Amongst all the talks the consensus was clear that data culture is not entirely new, but still not spoken about that much and is hard to create and change.

There is this massive stigma now, especially with the new GDRP laws, around data and keeping hold of our data private and watching what we expose and where. However, a number of people at EGG said when it comes to data, sharing is caring, and trust is so important.

We were introduced to the ‘data circle of life’ which starts at co-creating, moves onto collecting, sharing and accessing, and finally culminating at the end-model. That is all that data is needed for, and then it can go around again and again. We need to not be as scared to share our data, as data can be used in so many useful ways, but the data culture we currently have is centred on privacy and being scared of data sharing.

How do you build a data science team?

• Create focus
• Build what your customers need
• Create production ready products
• Measure success

Working together
Data scientists and machine learning engineers can work together to make a team. There seems to be a bit of a stigma separating the two, but in fact as a few speakers at EGG mentioned, by pulling together and working as a team, they can teach other.

Kim Nilsson, CEO at Pivigo gave a passionate speech about data, and she strongly believes in its power and how it can revolutionise our lives. There is no doubt that the future, in which data is exploited to its full potential, can be a bright and wonderful place to live.

Nilsson explained the steps that businesses and professionals need to take towards data maturity, which started with collecting data and access, but then picking apart this data to see how good it is, and how many people can reach and control it. She said: “A lot of people miss these vital beginning steps and then are doomed from the start.”

Collecting data, and collecting quality data certainly doesn’t happen overnight and people need to remember that it may take months and years - you need the right team with the right skills to succeed.

Moving forward: the data revolution
Outra is a company that aims to help brands connect their data to their customers. Simon Hay, CEO of Outra, started his speech by outlining that as data and insight grows every customer interaction needs to be personal and relevant to build sales and loyalty.

Hay does some work for the Brain Tumour Charity, and said: “We are going to cure cancer with data skills more than we are with medical skills.” He asked the audience to think about the data revolution, “We are often capped by the past and using languages and techniques that are dated.”

Hay said: “The industry is pulled back because we are using the same ideas just through different channels. They don’t always work and are not the most effective. We need to think outside of the box, and think about data.”

It wasn’t all bad news Hay explained, as GDPR is helping breakthrough some of these historical failings. Data science challenges people and acts as an agent of change.

Taking a look at the future, we can see a glimpse of it when we consider that there has been a 30,000% increase in IoT devices since 1992. That is a lot of data, so how do we deal with it all?

It can create problems don’t get me wrong, but what we learnt at the EGG conference is that if we utilise it correctly, this data can be so beneficial - at the moment one of our biggest downfalls is that a lot of data collated doesn’t get analysed or even looked at.

Artificial Intelligence (AI): the risks and opportunities
A hilarious and powerful statistic given to us by Captain Tim Hulme from the UK Ministry of Defence: “One third of Britons fear robots, but yet 17% of us are still willing to have sex with them!?”

Hulme went on to talk about the use of AI being limited by imagination. “When AI does work it is amazing, but it is not infallible.”

It is very easy to pass the blame to people Hulme discussed. It’s not machines that fight wars – people fight wars and they use their minds to do so.

To conclude the day, one speaker Dejan Petelin, Head of Data at startup Gousto, gave us some top tips for startups in the data world:

• Chose projects wisely
• Focus on the impact you will have
• Start simple, iterate fast
• Own your own algorithms
• Be accountable for your own work.