Waiting for the future
“In the 90s, we were dreaming of a future when machines would perform human tasks in a way that was better than humans,” says BrainCreators CEO Jasper Wognum, “But back then, it was all a big fantasy.”
A lot has changed since 1995, when BrainCreators co-founders Jasper Wognum, Gerbert Kaandorp and Peter Eijk met while studying at the then newly launched AI programme at Amsterdam University. When they graduated, they took the theories they had learned and hoped that technology would one day catch up with their algorithms and ambitions. In the meantime, they took jobs in different fields and waited.
Eighteen years passed, and in 2013 the trio began collaborating on LookLive, a shopping platform that utilised AI to identify and sell celebrity fashion. It was a successful business, but to them, it was only a stepping stone.
“We started LookLive to get back into AI,” says Wognum, “but we had bigger dreams for the uses of machine learning. We didn’t want to stick around in fashion; we wanted to apply automation where it would make a bigger impact.”
Time for impact
In 2016, Wognum, Kaandorp and Eijk finally started BrainCreators, the realisation of a dream more than 20 years in the making.
BrainCreators is an AI company, but its brightly lit office along Amsterdam’s famous Prinsengracht canal doesn’t stick to the stereotypical view of the sector. For starters, there isn’t a cyborg in sight. Instead, the window-side desks are occupied by software engineers sipping coffees and tapping quietly on their keyboards. In a glass-walled conference room, a nattily dressed group of clients takes notes as a BrainCreators expert gives a workshop on the applications of AI. “People think AI means creating robots that behave like humans but that’s not what we’re working on,” says Jasper. “We are automating processes to reduce manual labour.”
Rather than focusing on a single industry, BrainCreators helps businesses across a wide variety of sectors. Through seminars, workshops and events, they educate their clients about AI, demonstrating how machine learning can improve systems. Then, using their BrainMatter platform, they help their clients apply AI technology to automate processes.
The backbone of AI is data, and the BrainMatter platform structures datasets to facilitate machine learning. If you need your computer to identify the patterns that differentiate cats and dogs, you would most likely feed it lots and lots of correctly labelled pictures of cats and dogs. If half the pictures of cats were labelled “dog”, the system wouldn’t work. BrainMatter makes sure the data is structured and labelled correctly so that AI systems can learn and operate effectively.
Implications of applied AI
The practical applications of well-structured machine learning are far-reaching. The BrainMatter platform allowed geneticists working in selective pig breeding to reduce the time it takes to make phenotype predictions from years to minutes. It trained an AI model to automatically alert airport security in the event of a possible cyber attack and has helped Tata Steel to more efficiently assess the quality of its steel. “Machine learning augments our lives by taking over the tasks we aren’t good at or aren’t interested in,” says Wognum. “When tasks are very repetitive, we lose our concentration and make mistakes. Machines don’t. ”
Machine-learning in medicine
When you apply this kind of automated data analysis to life sciences, the implications are tremendous. BrainCreators has been working with physicians on a method of analysing brain scans to improve accuracy in diagnosing strokes. “If you can advise the medical experts on what they should be looking at – for example, ‘based on the data this is most likely an ischemic stroke, located in this region of the brain’ – then you prime the medical professional for a better diagnosis,” says Wognum. By using a deep neural network to identify stroke locations in 3D volume, they were able to trim the time of diagnosis from six minutes to less than 30 seconds while also improving accuracy.
BrainCreators is also training an AI model to predict the efficacy of cancer treatments, which Wognum expects to be high impact and, in yet another potential MedTech revolution, they are developing a method that would allow MRI scans to recognise and classify soft tissue and soft edges. This would make an MRI a viable alternative for cases that are currently only candidates for CT scan, a higher-risk procedure.
Moving into the future
When people get accustomed to artificial intelligence, they stop thinking about it as AI. Google Maps, Fitbit, smart-home thermostats: these are no longer novelties, but the white noise of modern life. “We try to minimise the hype by demystifying it,” Wognum says. “Every time there’s been a major technological advance, people have been afraid. But every time it has created more jobs and made life more interesting.”
Wognum predicts that AI will be ubiquitous within five years and believes Amsterdam should be among the early adopters: “In a small but crowded country efficiency matters, to keep traffic flowing, to keep people safe and to compete efficiently on a global scale.”
The right place, the right time
Amsterdam, with its cobblestone streets and historic canal houses, makes an attractive destination for internationals and the talent pool is only increasing in size. There are 11 different countries represented among BrainCreators’s team of 25 employees. A few of the BrainCreators engineers found the company simply walking by the Prinsengracht offices on a trip to Amsterdam and popping their heads in, intrigued by the futuristic signage out front. And it isn’t just adventurous engineers who are looking to set up shop in the Dutch capital. “You don’t have to go to Silicon Valley to meet international companies,” says Wognum. “Uber, Google, IBM, Salesforce: everyone has an office here.”
Amsterdam is home to the world’s largest data-transport hub, and its business culture is “honest and straight to the point,” according to Wognum, who says he would like to see Amsterdam doing even more to compete globally in AI. “When the local government and businesses begin really investing in and incorporating AI, people will see the practical applications and the impact it makes,” he explains. “That’s how you stimulate the whole scene.”
There is no doubt that AI is on the rise. A recent survey shows that the percentage of CIOs employing AI has gone from 4% to 14%, tripling in one year, and half of the organisations in the survey indicated plans to employ AI before the end of 2020. A local example of this growth is Amsterdam Science Park’s Startup Village, which has doubled in size in the past year to accommodate the expanding community of startups working in the field.
“This is only the tip of the iceberg”, says Wognum. “We should be preparing for a time when most of our lives are automated. It isn’t a bad thing. The human role in this process changes from doing all the repetitive work to finding the dependencies and coming up with new solutions. It will free us up to be more human - emotional, creative and conscious - and machines will automate all the rest.”
Find out more about the latest life-science innovations in Amsterdam.