5 December 2023

Without investment, AI will remain science fiction here

OpenAI launched ChatGPT one year ago. Since then, developments have been moving fast across the ocean - with all the consequences. As a guiding country, can the Netherlands develop AI on European values?

On 6 September, the cabinet had an update on the growing impact of AI on society. There are concerns about the negative side effects that such a powerful technology can cause, but also possible missed opportunities for our economy. For instance, will we still be market leaders in horticultural technology in 10 years' time, if China develops an AI-controlled greenhouse?

With the arrival of ChatGPT one year ago, AI permanently nestled itself in our consciousness. ChatGPT is an example of a large language model that trains itself - based on a huge corpus of text data - to complete sentences. This simple recipe leads some to call ChatGPT a "statistical parrot". But the magic of modern AI is that this simple recipe, if followed on a large scale with lots of data, lots of computing power ánd large models, leads to a form of general intelligence: the ability to reason logically about a wide range of mundane issues.

This form of AI will have a huge impact on our lives. Take the example of a chatbot that can listen in on a conversation between doctor and patient, make a report of it, fill in all the forms ánd make a diagnosis. This is not science fiction, but science reality. AI will integrate into all the capillaries of society as a system technology, just like the internet and electricity.

Huge investments

Training AI models is very expensive. Microsoft recently invested more than $10bn in OpenAI to develop the next-generation ChatGPT. Among other companies, it is code red to catch up.

In the Netherlands, even ASML is struggling to develop this technology in-house. It takes huge investments and the right ecosystem to join this race. The Silicon Valley tech giants threaten to develop a monopoly that cannot be caught up with, as happened earlier with search technology and social media. To avoid falling behind, we are forced to use American AI technology. A technology developed with dubious data, not always transparent and not according to European values.

Since AI is the foundation on which the future economy will be built, it is crucial that we also develop the technology ourselves. There is a unique opportunity for the Netherlands to do this together with government, knowledge institutions and industry. Apart from the fact that we already have experience with this kind of 'triple helix' collaboration in, for example, the Innovation Centre for AI (ICAI), there is also a technical reason behind this opportunity. The new form of generative AI - of which ChatGPT is an example - is trained on large and diverse datasets, which are then specialised on specific tasks. A model trained on holiday photos can be fine-tuned to recognise cancer in tissue scans. Exploit this phenomenon to spread the problem of the huge investment cost of a competitive AI lab across a broad coalition of companies and governments.

Ethical values

What are the ingredients for a successful AI lab in the polder? First, bringing together and keeping the top talent in this field. Besides competitive salaries, these researchers and engineers should be able to conduct free research and publish in the top journals in their field. They should also have access to plenty of computing power. This is where the government could take a leading role by funding an AI computing centre.

How do we make this AI lab interesting for companies to invest in? We pair researchers with companies to formulate research questions together. Researchers then work on a potentially abstracted version of the problem. Investing companies get a 'first right of refusal'. And a shell of tech transfer by software engineers takes care of implementing, testing and transferring the developed models, together with the respective AI labs within the participating companies.

Europe wants to play a leading role in developing AI technology according to ethical values. This is important, but Europe cannot just stand by as a referee shouting. Unclear regulations around AI now still create a barrier to experimenting with AI. It is precisely an AI lab in the polder that can develop expertise on the ethical and legal aspects of AI. It advises companies and governments, and helps curate datasets so that they do not result in discriminatory models.

'We can only catch up with Europe's backlog in AI if we invest in it and join forces'

The AI lab in the polder will be a hotspot of startups and scaleups where researchers are encouraged to bring technology to the market, within the legal frameworks around intellectual property. Only then will the new Booking's and Adyen's of this world be born in Europe.

AI is going to change the world drastically and permanently. We will have to join forces to remain economically competitive. Because of the Dutch culture of collaboration and the nature of generative AI, there is a unique opportunity for us, preventing Dutch companies as well as the government from falling behind in the development of AI.

Max Welling is Professor of Machine Learning, Cees Snoek is Professor of Computer Vision and Maarten de Rijke is Professor of Information Retrieval; all are affiliated to the University of Amsterdam.

Read the full article on Het Financieele Dagblad's website (in Dutch).

Vergelijkbaar >

Similar news items

>View all news items >

6 September 2024

CuspAI Introduces Itself at LAB42 >

On September 5, 2024, Max Welling and Chad Edwards, founders of CuspAI, presented their innovative company during the IvI coffee & cake gathering.

Advanced AI for Surveillance Robots: A Collaborative Project by Dutch Institutions >

A consortium of leading Dutch research institutions and government agencies, including TU Delft, the University of Amsterdam, TNO, and the Royal Netherlands Marechaussee, has launched an ambitious project aimed at developing advanced artificial intelligence (AI) for surveillance robots. Officially initiated on September 4, the OpenBots consortium focuses on creating AI systems designed to assist human security officers in various security settings.

NeuroAI: Charlotte Frenkel explores the future of AI inspired by the human brain >

With the award of an AiNed Fellowship grant, Dr. Charlotte Frenkel from TU Delft delves into neuromorphic computing, pioneering research aimed at creating energy-efficient and powerful AI systems inspired by the human brain. This research bridges AI and neuroscience to develop faster, more energy-efficient, and smarter computing systems.