Artificial intelligence

Advances in artificial intelligence (AI) have never been faster. The technology seems to hold almost unlimited potential, and it is hard to imagine where it will all end.

Just a few years ago, few would have imagined a solution like Open AI's ChatGPT, which the whole world now knows and uses. It is a solution based on AI.

AI is playing an increasingly important role, both in our private lives and in key functions in society. In the healthcare sector, for example, AI is increasingly supporting doctors and other healthcare professionals in diagnosing and treating patients. In the energy sector, AI is used to save energy and make the best possible use of renewable energy sources.

AI employs mathematics and logic. We know how AI works, but we cannot always explain how it arrives at a specific solution or assessment. This underscores our obligation - as researchers and as members of society - to set the right standards for use of the technology, both in legislation and morally.

At DTU, we have a strong focus on the ethics of AI and on ensuring that future AI solutions do not contain biases or favour certain groups of people. 

FAQ about artificial intelligence (AI)

See the answers to the most common questions about AI.

Artificial intelligence (AI) leverages computer programs and machines to mimic one or more aspects of human intelligence such as abstract thinking, analysis, problem-solving, pattern recognition, language mastery and comprehension, planning, etc.

Computer programs that play chess, diagnose patients, plan routes, or engage in conversation are examples of AI. 

In principle, there are no limits to what artificial intelligence can be used for, because it's all about trying to mimic human thinking. Almost anything we can do with our brains, we can try to mimic with artificial intelligence.

Of course, some things are easier than others. It's easy to make a computer that can play chess or find a route from A to B. It's quite difficult to make a robot that can help with cooking and cleaning and things like that.

Some of the things artificial intelligence is being used for today are things like chatbots in smartphones and on websites, medical diagnostics - such as diagnosing from X-ray images, controlling robots in factories and warehouses, editing content - for example, what we see (and don't see) on Facebook and what we get recommended on Netflix.

AI affects society in many ways. Among other things, it is used by both private and public companies to boost productivity and automate routine tasks. It is also used to improve and support decision-making and assessments. In the health system, AI assists in diagnosing and treating diseases based on analyses of images from X-rays and scans, for example.

The ever more widespread use of AI also raises concerns, including fears that certain types of jobs and functions will be replaced or greatly affected by AI. But also concerns about the possible bias of the algorithms and data used to train AI engines, resulting for example in sexist or racist solutions. Finally, privacy and security concerns have been raised about the unintentional spreading of sensitive personal information through AI-based solutions.

The use of AI comes with a number of ethical challenges. Among other things, there are concerns that the complexity of the technology makes it hard to fully understand how it arrives at its solutions. Another concern is to do with the possible bias in terms of gender or ethnicity of the data used to train AI engines, and which will therefore also have a bearing on the solutions arrived at and subsequently applied. And privacy and security issues may surround the unintentional spreading of sensitive personal information through AI-based solutions. 

A great many forums - including forums of researchers - are addressing how to establish ethical guidelines and standards for the responsible development of AI and its use for the benefit of society.

AI is expected to play an ever greater role in the future, not least due to the rapid evolution of the technology. The algorithms are getting stronger, we are collecting more and more data, and the data-processing capability of computers is increasing. This means that AI can be integrated into an increasing number of products and services.

The EU is currently passing legislation on the responsible use of AI. This is done, among other things, with the EU AI Act and the creation of testing and experimentation facilities (TEFs) for AI in four main sectors: agri-food, healthcare, manufacturing, and smart cities and communities. Here, it will be possible for companies to test new products based on AI.

The future proliferation of AI will depend on technological developments, political decisions on limitations in the use of AI, and society’s acceptance of this. 

Machine learning is a digital technology that enables computers to identify patterns and learn from data and on this basis suggest how a task may be solved. At the same time, the computer can learn from its experience and thus automatically improve its performance over time.

A large language model (LLM) is a type of AI trained using large volumes of data in the form of articles, books, websites, etc. From the data, the model acquires an understanding of human language, enabling it to calculate the probability of an answer or result. ChatGPT from OpenAI is powered by large language models.
A neural network is a computer model inspired by the brain and nervous system of humans and animals. Just like the brain, a neural network is made up of (artificial) neurons that are connected and can transmit signals to each other. Each neuron receives input from other neurons and then calculates an output that is passed on to other neurons.

Neural networks can learn to solve tasks by training on large amounts of data. Neural networks rely on training data to learn and improve their accuracy over time. But once these learning algorithms are fine-tuned, they are powerful AI tools that allow us to classify and group data at high speed.  

A transformer model is a type of neural network architecture that transforms one type of input into another type. The term was introduced in connection with Google’s translations from one language to another, but has since been used in many other contexts. Among other things, transformer models are behind services such as Dall-E and Midjourney that generate images based on natural language descriptions.

AI covers multiple techniques and approaches. A rough distinction can be made between symbolic AI on the one hand, and machine learning or statistical learning on the other.

We use symbolic AI techniques to try to get computers to make plans or think logically in the style of humans. Symbolic AI is often used to create computer programs that can search for a solution to a problem, e.g. search for the best move in chess or search for the best route from A to B. Symbolic techniques are used to control transport robots, for example.

Deep learning is a special modern approach to large (deep) neural networks, which has significantly boosted the ability of neural networks to understand and assess images and text. For example, it has been possible to get neural networks to accurately describe the content of images.

Contact our experts in AI

Thomas Bolander is a professor and researcher in logic and artificial intelligence. He works primarily with social aspects of AI, enabling computers and robots to interact socially competently and flexibly with humans and other robots.
Aasa Feragen is a professor and expert in medical image analysis and artificial intelligence. She is particularly interested in responsible AI such as transparency, fairness, and ethical aspects in the use of AI. Aasa Feragen's work ranges from mathematical modelling of algorithmic fairness in healthcare to the development of explainable AI algorithms for clinical use.
Brit Ross Winthereik

Brit Ross Winthereik Professor, Head of Division

Brit Ross Winthereik is a professor specialising in people's interactions with digital technologies. Originally trained as an anthropologist, she has spent more than two decades researching the role of technology in society, with a particular interest in the impact of public sector digitalisation.

News about AI

More news about AI