I joined yesterday for the first time this event in Berlin, not (only) based on the location (some of my readers are aware of my partiality for this vibrant city), but mostly on my curiosity and deep interest for the rapid progress of Artificial Intelligence and its potential impact on business, information systems and more generally on our lives.
While taking my very early flight from Paris, I was unsure whether I would eventually bring back new and fresh insights, be lost in a gathering of Geeks, or just comfort my current ideas. As you can guess from the picture, also the audience was including some Geek-like profiles, most of the participants looked very much business or academics. The content included a mix of inspiring keynote speeches on future of AI, its technical and societal implications, but also practical sessions proving that AI is becoming “real” and has materialized in a number of solutions and application cases presented at the conference.
On my way back, and despite a surprising lack of performance of Berlin airport security resulting in a very none-German mess at the airport, I felt that my trip has met and exceeded my expectations. Based on the presentations and conversations, I am more than ever convinced that AI is the next big revolution and a key topic for businesses, academics, and citizens. As only 600 participants got the chance to attend this great event, I decided to take the time to write down and share some of my insights in this article in the form of few quotes from some of the speakers.
” Out of the 7500 companies claiming to do AI, we have identified 3,465 Artificial Intelligence startups, 40% are based in the US, only 700 in Europe” stated Fabian Westerheide, Rise of AI organizer, based on a new Research by ASGARD and ROLAND BERGER. China is making massive investments and catching up quickly. This was also confirmed by Mark Schmitz, a partner at Lakestar, stating that China surpassed the US in AI funding in 2017. From Mark’s perspective, key ingredients for successful AI companies are sustained and ideally proprietary data access, allowing continuous learning and defensibility, talent access – in particular engineering and data sciences, and a clear commercial application combining creativity and know how in monetization.
” We are entering the wave of AI. This 4th wave will show different winners. We will see the emergence of the portable AI, personal and decentralized, less data-driven and replacing apps and digital interfaces“, predicted Charles-Edouard Bouée, CEO of Roland Berger , present at the conference on his birthday (Happy birthday and thanks for a great presentation!). One of my favorite slide below.
” Simple problems can be solved by software, complex by humans, what about the middle? Sporadic problems is the space of AI, this is what AI can bring to business applications” explained Dc Stefan Wess. From his perspective, today’s killers for real world AI is 1/ data availability and data quality 2/ explainability and comprehensibility – as business leaders are typically not comfortable with “black box” solution.
“ For which relevant problem can you find a superior man versus machine work split?” is the question that Georg Wittenburg, recommended to ask yourself to identify promising AI application. If you believe like Andrew Ng that “AI is the new electricity“, then recommended Christian Nagel, from Earlybird VC, ” AI needs to be intelligently applied to create value, like the next light bulb or electric motor“.
“So far we had to adapt at the speed of competition – at the speed of human brain. Now if your competition is using machine intelligence, you have to adapt at the speed of the machine” challenged Antoine Blondeau, from Alpha Intelligence Capital as an introduction to a very interesting session on the value of AI to make better decision (versus only predictions) and the requirement to design evolving algorithms learning at every loop.
“The Robot doesn’t judge me” explained Dc Ben Goerzel from SingularityNet, commenting on an experiment where participants got a better experience of mediation when guided by a robot than by a human. Are we doing to be surpassed by robots not only in maths but even in human communication?
Quite reassuring was Michael-Maria Bommer, who still see some value in us. ” Bots often fail. So human agents must “tango” with AI“. For instance in call centers implementing bots, successful agents can become bots managers and guide the learning of the bots. As per Dc Aljoscha Burchardt, who ” never wants his decisions to be dominated by a stupid machine” (but are we not already each time we do a search?), chatbots need to collaborate with users to build a knowledge base. The amount of data they need to consume to feed their intelligence is quite impressive – for 7K labels for simple FAQ answers to 40K labels to detect intent for instance. “Machines read on the lines, humans read between the lines“.
And I will conclude this short summary by the words from Irakli Beridze, leading United Nations UNICRI center for AI and Robotics : “AI is now potentially the most powerful technology we have ever created. Today AI is known to enhance crime, produce lethal autonomous weapon systems and enhanced technological unemployment, not to mention the existential risk of a potential super intelligence. But AI can also contribute to good wealth and well-being, for instance, predict Ebola outbreak, quality education, and peace and justice, for instance when analyzing crime data and predicting solutions”.
If you found those quotes inspiring, watch out for Rise of AI 2019 – and register early, as the seats will be limited again.