And another conference on artificial intelligence? On the 18–19th April, I took part in AI Expo Global, an event hosted at the Olympia National, London, in which more than 10,000 attendees discussed not only AI and Business Intelligence, but also topics like IoT, Big Data & Analytics and the emergence of blockchain.
Speakers including P&G, Wells Fargo, HSBC, Salesforce, EON, Dell EMC, Unipol, Credit Suisse and many more took to the stage to explore the use of AI in marketing, healthcare, energy & utilities, banking & finance, government & public sector and more.
I have followed several panels during a day and a half spent at the conference, and these are my main findings:
- Data scientists: Where do you find them? How do you educate them?
The ‘AI in the Enterprise’ session, where our Executive Director Dennis Curry was one of the panellists, described how Konica Minolta is implementing intelligent solutions to face the challenges from the increase of complexity within our working daily activities, and how data scientists are playing a fundamental role within our organization. “They are one part of the puzzle around the AI. They are cool, and the best talents come when you have a vision on a project. It is not just an aspiration to do that,” said Dennis Curry while replying to Pardeep Bassi, Head of Data Science at LV= who was supporting the concept, “Data scientists do not exist. They have to be created and then educated to tackle a problem.”
On the same page of Dr Curry was Peter Jackson, Chief Data Officer at London Southern Water. He argued that, “The best action an organization could take for engaging data scientists is the upscaling of internal resources.” This is feasible because the data scientist is already an employee of the company and they know the company vision. Furthermore, the intellectual background of a data scientist cannot be limited to a few technological areas: it has to be wider and as multidisciplinary as possible embracing engineering, mathematics, statistic, and communication skills as described in The data Scientist Venn Diagram.
- AI Application fields in ‘good for society’. The world is getting smarter and smarter and AI enables you to make better and more clever decisions. When developing solutions for social areas such as food, healthcare, charity and no profit, discrimination of minorities, education and HR in the workplace, considering the economic, ethical and social implications of artificial intelligence is a mandatory requirement. According to James Hodson, CEO of AI for Good Foundation, in order to avoid biases baked into AI, a set of general rules might be respected:
- involve the whole ecosystem (from end user to Government),
- get people involved in the innovation process,
- define new regulations (as described in the report from the UK Parliament, April 2017),
- understand the environment where AI will be adopted.
- IoT+AI+Security: what about standards? Data generated by IoT devices or data provided on a voluntary basis via a “Data Marketplace” is now considered the fuel of artificial intelligence. Moreover, data analytics is strongly related with the domain of AI. And when the security topic was raised, the most frequent answer was ‘blockchain’, but this seems to be an imcomplete answer for both security issues and also for standards. According to many participants in the event, there are still many doubts and uncertainties. Today there is no common standard for blockchain. Several initiatives are taking shape to identify the pros and cons on the implementation of blockchain and proper guidelines for developers are required as many solutions are already available in this novel market. The risk of recreating issues, similar to the vendor lock-in case that we lived with the first generation of IoT devices, is realistic and cannot be underestimated.
- EU as the best market for AI. According to some of the panellists in the conference, upcoming regulations like GDPR and others the European Commission may release in the coming future for avoiding the abuse of AI are going to set the boundaries in making Europe the best market for AI technologies.
- The key role of Universities for shaping the smarter worker of the future. The secret for Europe to overcome difficulties and barriers in the adoption of AI in the enterprise is the high level of education that European Universities offer to students. “The European university – as an old established institution – is requested to play a big role in accelerating AI adoption and usability for new generations,” said John Shawe-Taylor, Professor of Computational Science & ML at UCL). “AI will not make students more ignorant or too lazy to acquire knowledge. AI is a tool for supporting academic education while presenting the right knowledge at the right time and in the right form for shaping the smarter worker of the future.” Therefore, no worries, AI will not replace humans, but will assist the creation of a new workforce with new jobs.