AI: a hyper-intelligent flywheel
Aiir Innovations started out three years ago as a university project by five students of the University of Amsterdam’s Artificial Intelligence master’s programme. The airline KLM had asked them to research whether AI could help make the inspection of airplane engines faster, more thorough and more reliable. ‘Within four weeks, we developed a prototype for a digital AI assistant, which KLM’s engineers tested with positive results,’ said Miriam Huijser, alumna of the AI master’s programme and co-founder of Aiir Innovations, in an interview with the student newspaper Folia.
With the help of an encouragement grant by the Mainport Innovation Fund and organisational support from the ACE Venture Lab, Aiir Innovations was born. The startup is now in the final development phase for its self-learning system that autonomously detects material anomalies or weaknesses, reducing the need for mind-numbing and thus error-prone manual inspections.
After a successful first few years, Aiir Innovations expects to expand the customer base for its smart software solution to other airlines. The UvA students have become AI entrepreneurs.
The power of smart algorithms
The term Artificial Intelligence (AI) denotes applications that use data to train machines to be self-learning, take decisions and carry out routine actions autonomously. AI renders possible the personalisation of products, supports prevention of errors, and helps us take more informed decisions. According to Maarten de Rijke, professor of Information Retrieval at the UvA, the importance of AI cannot be overestimated: ‘AI is today’s most significant technological innovation and promises enormous changes in our society.’
Indeed, the possible applications for AI are manifold. Smart algorithms support medical professionals with diagnoses and enable personalised medicine. Virtual assistants can answer people’s questions, while microtargeting makes tailor-made marketing possible. Autonomous transport can provide solutions for traffic congestion. ‘Smart cities’ will become even smarter. AI can also provide a significant impulse to goals such as sustainable living, housing and working.
However, our increasing reliance on AI also raises new questions. We need to rethink organisational and income models, products and services, laws and privacy regulations and data ethics. There will be shifts in the labour market. Education programmes will need to be adapted, and lifelong learning is more important than ever.
The economic landscape is changing, too. Increasing numbers of organisations will become footloose, requiring only a location that ensures their data streams can be stored, processed and shared securely.
Amsterdam as an AI hub
Considering the huge influence of AI – now and in the future – it’s crucial that the Netherlands, with the Amsterdam Metropolitan Area at its helm, can offer an intelligent, accessible and innovative ecosystem for the development and distribution of AI applications. Fortunately, the country is in a great position to provide just that.
In international comparisons, the AI hub Amsterdam scores highly in all essential aspects. The Amsterdam Area has Europe’s highest concentration of data centres. AMS-IX – one of the world’s busiest internet exchange points – and the country’s broadband network ensure secure and stable data traffic. Amsterdam is home to an internationally oriented network of creative spirits and business services, and the city is a magnet for – often all-too-rare – AI talent. In addition, there is an accessible academic and research ecosystem, leading to numerous scientific advances. No other European city is further in anticipating the possible effects of AI and what measures are necessary to ensure we can profit from its many applications in a socially responsible way.
The so-called ‘flywheel function’ of science is crucial for successful developments in AI, says Natali Helberger, professor of Law & Digital Technology (with an emphasis on AI) and part of the UvA’s AI team. As part of a unique interdisciplinary approach, Helberger and her colleagues research the impact AI has on its users and our society. ‘AI still tends to be perceived as a technological challenge. But more importantly, it’s a question of integration: how can we incorporate AI into all aspects of our lives in the best possible way, and how will this change our way of living?’
Professor Natali Helberger
According to Max Welling, a strong sector is needed to retain human control over AI. Welling is vice president technology of the chip manufacturer Qualcomm, professor of Machine Learning at the UvA and board member at the European AI research network ELLIS (European Laboratory for Learning and Intelligent Systems). ELLIS is an association of European universities and research institutes with the aim of establishing ethical frameworks and training a new generation of data scientists. In an interview with the Dutch business newspaper Het Financieele Dagblad, Welling said: ‘We – Europe – want to take part in deciding the ethical guidelines for the use of AI. In order to do that, we must ensure we are involved as a fully-fledged partner’.
Research and healthcare organisations, local authorities and businesses also work together in Amsterdam to better coordinate supply and demand for AI research and developments and to help bring promising projects by startups to market.
The Innovation Center for Artificial Intelligence (ICAI) is a shining example of how this can work in practice. Founded in 2018, the ICAI network brings students, scientists and practitioners together, enabling them to cooperate in industry labs to develop AI-related solutions. This is how the basic concept for an AI-supported form of medical image recognition was created, which is now being brought to market by the AIM Lab startup. It is accessible initiatives such as this that make Amsterdam a frontrunner in the development of AI applications for healthcare and life sciences; startups such as Medvice, PacMed and Aidence are also working on innovations in this segment.
Marcel Worring from ICAI
ICAI partner and retailer Ahold Delhaize is participating in an AI lab in which seven students of the AI master’s programme are working on designing algorithms to target customers’ individual needs and to improve the management of the flow of goods. At the launch of the programme, Ahold board member Frans Muller said: ‘We want to learn how AI can help us to serve our customers better and more sustainably.’
Another ICAI spin-off is Scyfer, which develops software for machine learning. The startup was acquired by Qualcomm in 2017. Like so many of the others, it’s located in the Startup Village at the Amsterdam Science Park, the heart of Amsterdam’s AI ecosystem. And just a few doors down, you’ll find Aiir Innovations, with their international breakthrough firmly in their sights.
Read more about AI in Amsterdam.