Some examples of AI application
Health care: AI will provide quicker and more accurate diagnosis. Virtual health assistants will monitor and support patients outside the hospital. AI will enable healthcare workers to create a more personalized patient experience.
Transport: we are not completely there yet, but maybe in ten years, AI will make sure autonomous cars will drive us from home to work everyday
Education: virtual tutors will assist teachers and through facial analysis emotions of students are measured to help determine who finds it difficult or who is bored. With the goal of better tailoring the experience to the individual’s learning needs.
An interview with Dr. Mahdi Ebrahim, assistant professor at the Brightlands Institute for Smart Society.
What is Artificial Intelligence and what is its potential?
Artificial Intelligence (AI) makes it possible for machines to learn from experience, adjust to new inputs and perform human-like tasks. Well known examples are autonomous cars or for example the chess playing computer. Did you know artificial intelligence isn’t that new; already in the 1950s the first AI program has been developed.
The potential of AI is gigantic. It has the potential to make the world a better place, it can improve health, it can further education, and it can also increase wealth.
Artificial intelligence will affect product/service innovation and processes. For product/service innovation new tools are being developed to increase sales, cross-selling or up-selling. Process innovation will have a bigger effect on the organization and will increase automation. Mahdi is convinced it will help us more than it will harm us. It will enable us to use our creativity more, with additional time we have saved from delegating routine tasks to machines.
What is the first question you have to ask yourself as an organization when thinking about using AI in your business?
One thing for sure, it’s not about using data or processes. First question you need to ask yourself is what is the problem/challenge you want to solve? What is your customer’s problem? It needs to be in line with your organization’s strategy.
Other important considerations are:
Can AI be threatening?
It’s highly unlikely that we will get into the situation that many databases and systems become highly integrated, like in the sci-fi series Person of Interest. During this series, an ex-CIA agent and wealthy programmer save lives via a surveillance AI that sends them the identities of civilians involved in impending crimes.
What is potentially problematic about AI is that the quality of the output is based on the quality of the input of the data. AI is trained on human data and therefore is prone to re-produce human biases. If you train AI in a bad way, it will behave bad. For example, if you train an investment advisor robot based on data from an investment company with a high level of male investment managers it might regenerate a diversity bias. Since supposedly men take on average more risk than women, there is a bias present that will be included in the trained AI.
Another important issue is data privacy. If medical records of individuals are available and AI tools will be developed for insurance companies, it can harm the privacy of individuals. You already see this happening in the US. There is also the ethical aspect, where should we draw the line?
The GDPR (general data protection regulations) is a first step in protecting individual data, but will this be enough? On the ethical aspect, the EU for example is working on developing ethical standards around data when working on an EU funded project. And of course, also society demands from organizations that they act more responsible and ethical, especially the younger generation.
Why should you learn about AI? Should we learn how to apply it or do we all have to learn code?
Not everyone needs to code, that’s just a myth. Let’s leave coding up to the specialists. What is important however is that people develop an artificial intelligence literacy. Know about the potential of AI, know how to use it in organizations, what resources are needed to utilize it best (time, money, expertise, data).
What is the latest AI research paper you wrote or read and what were your conclusions?
Earlier this week I was reading a paper about the role of strategy in digital transformation stating that strategy should be the driver behind the whole transformation. If you don’t know where you are heading, you will be lost. Last year I was working with a large company that experimented with a lot of different AI tools. Experimenting technological innovations in itself could be a good thing, but AI was not integrated into their strategy. What happened is that the company had implementation problems, they couldn’t make it happen because it didn’t contribute to the companies’ strategic priorities.
What are the 5 most common AI terms you should know?
Machine learning is the science of getting computers to learn and improve their learning over time in autonomous fashion, by feeding them data and information in the form of observations and real-world interactions.
Supervised learning is a way of learning in which we teach or train the machine using data which is labelled (classified). This implies that some data is already tagged with the correct answer.
Unsupervised learning is approach of training a machine using information that is neither classified nor labelled, which allows the algorithm to act on that information without predefined guidance. The machine groups unsorted information according to similarities, patterns and differences without any prior training of data.
Artificial neural networks refer to artificial intelligence systems that simulate connected neural units, modeling the way neurons interact in human brain.
Deep learning uses neural networks consisting of several layers of artificial neurons. Deep learning is perfect for projects involving huge, complex datasets.
Dr. Mahdi Ebrahim is an assistant professor of digital innovation and strategy at the School of Business and Economics in Maastricht University. He is also an academic affiliate at Brightlands Institute for Smart Society (BISS). Mahdi is interested in digital innovations (such as Artificial Intelligence), how organizations adapt to these innovations and how they organize to develop digital innovations. He frequently teaches managers and executives about AI, its use-cases and value contributions to the firms and its resource requirements. He has applied and extended his knowledge of digital innovation management in projects with large organizations such as APG, KPN, Volksbank, and CBS, as well as with Small and Medium size Enterprises (SMEs). He has a PhD in technology and innovation management from Bocconi university in Milan. Before, he got a BSc degree in mechanical engineering and MSc degree in entrepreneurship management, both in Tehran.