What’s the role of AI startup firms in the European Union?
One year more the European Commission celebrated the ICT Proposers’ Day on November 8th, 2017. This year, the Hungarian Ministry of National Development was the hosting nation and their organization was outstanding. The event focuses on the 2018 Calls for Proposals of the Horizon 2020 Work Program in the field of Information & Communication Technologies and Future and Emerging Technologies (FET). The European Union is at the forefront of research in the world. In this event, the selected projects for the year are discussed extensively together with experts from a variety of related fields.
The European Commission is leading the world when it comes to initiatives that build bridges between AI technology, experts and startups (an important link for this is the Startup Europe Club). On this occasion, I had the opportunity to moderate a double networking session organised by Startup Europe and the AI and Robotics Services of the Commission, that included two panels”. The first one was about ‘The State of the Funding of Artificial Intelligence (AI) for Startups’ and the second one was ‘The Role of Startups in the European Artificial Intelligence (AI) Landscape’. These are the key points of each panel:
The panel on funding for startups was particularly insightful. Populated with prominent VCs and leading startups, in this panel the speakers described what was the motivation of their firms and their specific ‘sweet spot’ for investing in companies. While VCs asked for clear signs of traction early on; entrepreneurs emphasized that the key to getting funding lies in providing immediate value for consumers.
Things heated up when the discussion gravitated towards understanding if VCs really know the difference between AI and just ‘a very sophisticated algorithm’. Some investors mentioned that when neural networks scale, that this is really AI. Instead, entrepreneurs described AI as a form of application of intelligence into practice, especially if this technology uses Big Data intensively. This was a point where everyone seemed to have their own unique understanding of what is AI. Which is probably not the best way to introduce this technology into the market. This additional point, the introduction of AI into the market had instead a quite unanimous response.
Artificial Intelligence: A global industry?
The panel clearly described AI as a global industry. After all, if a computer is better at image recognition than a person, this has nothing to do with the geographical location. An image is an image anywhere in the world. In this respect, the panelists described some private initiatives, such as France is AI, that is trying to build an ecosystem for AI in some countries, but lamented that there was no pan-European initiative that would somehow make this easier. They posited that if the European Union would have such initiative it would have a serious shot at gaining and maintaining an edge on AI. Still, panelists were hopeful, as they dutifully remarked that these government-wide initiatives to promote AI are not only absent in Europe at the moment but also anywhere else.
When asked if Europe had a shot at being a leading force for AI in the world panelists unbundled the challenge in three parts. First, data as row material for the AI industry. In this respect, they feared that regulation on data was too burdensome and inefficient in Europe, which may prompt other economies (especially Asian economies) to indirectly benefit from that and end up becoming the repository of AI data in the world. Second; ‘number crunching’ where Europe is supposed to have an advantage. Turns out that the difficulty in financing ‘number crunching’ projects and the ability of the Asian countries to process data is again becoming a threat. Third, building a business ecosystem to make the AI industry more economically attractive. There was a general agreement that better exits would lead to better funding, which would eventually lead to better projects.
The role of startups in the European landscape panel redirected the audience’s attention to where and how can European AI firms become a dominant force in Europe and the world. Populated again by VCs and prominent entrepreneurs the dominant discussion of the panel revolved around focusing efforts in vertical industries where the development of AI expertise can give European companies an edge. For instance, there are funded companies in Europe that are applying AI to car manufacturing machines that would identify – based on a collection of symptoms – if a machine is about to malfunction and will subsequently halt operations. Generating several millions of euros in production costs.
Another example came from the Hyperloop entrepreneur. He explained how the Hyperloop company is divided into four divisions and each has had a different strategic approach. The first one is ‘inboarding’. At the Hyperloop they are developing an entire system of tracking passengers so they can monitor them in the distance and have them enter the Hyperloop without any sort of check-in. The ‘Executive Assistant’ division is geared toward making the trip as comfortable as possible. The ‘Collaboration’ division, a clear example of Open Innovation, has over 20,000 collaborators that are actually improving in a private ecosystem the entire company on a monthly basis – this division is placing a particular emphasis on automation. Finally, there is a ‘Systems’ division which is focused on adding and/ or removing components of the system platform to make the company better, more efficient, and more connected to the environment. This division is trying to make the entire process of booking a trip seamless and effortless.
Another fundamentally critical point where the panel was particularly critic and recommended urgent action is regulation. They claimed that in Europe the amount of regulation is unusually high, especially the paperwork that needs to be taken care of to build a fully operating business. They lamented that this regulation, far from helping entrepreneurs, is actually deleterious for both the European Union and for making industries evolve. A key final point came at the end. The panel acknowledged that while having a successful business might be a problem in some economies – because the government or established firms might want to intervene too much – that this was not the case in Europe. They all unanimously considered the European Union a place of fair trade where the regulator does its job at protecting new firms that are aggressively scaling up and revitalizing industries. Indeed, a very positive note to end the discussion.
In my view, this high- level discussion left us with some work to do: First, we should congratulate the European Commission for creating such an event. Showing that one cares is a great way to start. Initiatives like this are not at all that common in the rest of the world. But the panelists also left us with some challenges that we should be concerned about. Same as with many entrepreneurial initiatives, the process of entrepreneurship is still too informal and lacks focus. This not only delays industry transformation but also makes promising companies fail. There should be a ‘protected ecosystem’ that must be developed at the pan-European level and that could shelter entrepreneurs from regulation and local constraints in the first years in the life of new firms. After all, one ecosystem can disrupt another, but it is rare that one company disrupts an entire ecosystem. Finally; regulation. We understand that regulation has a place and a reason, but too much of a good thing can also be destructive in the short term. I am sure that there must be an optimal solution that satisfies the European Commission’s fundamental requirements and does not have these deleterious side effects. In my opinion, it would be wise for the European Commission to take these recommendations seriously and take action. After all, we rarely receive such a clear message from people of this caliber.
Post written by PhD. Juan Pablo Vázquez, strategy, EADA’s Leadership and People Department professor.