Use of Artificial Intelligence (AI) has not always yielded the desired results or has even spiraled out of control. A well-known example is the recruitment and selection algorithm that was self-learning and eventually only hired men. But also algorithms that draw discriminatory conclusions based on colored data. To prevent problems, I am hearing more and more organizations banning the use of AI. That sounds logical, but it is not. But let's first establish what AI is.
What is artificial intelligence?
Simply put, artificial intelligence (AI) refers to systems or software that emulate our own intelligence to perform tasks and can improve themselves during that process based on the gathered information. The same organizations that still believe they can stop the use of AI have probably been working with it for a long time.
AI is already widely used in photo and image recognition software, which independently draw conclusions from the results and are often self-learning. But also in chat and voice bots, and in recommendation software. The use of this type of AI therefore takes place in many organizations without their awareness and without assessment, supervision, or knowledge in numerous processes. Using data whose reliability is unknown. Here too, the principle applies: garbage in, garbage out.
With the conclusion that AI has long been 'among us', we are long past the point of banning its use. It is therefore better to try to gain control over it. No organization of any size will be able to avoid using AI, either directly or through purchased systems or solutions. Organizations that stick to a ban on AI will ultimately become extinct. That is evolution.
So what should organizations with a fresh fear of AI do?
First, you can start by asking whether AI is already being applied. This means using a good definition (there are many and not everything is AI) and taking stock of where AI is already being used in the organization and for what purpose. Based on the results, policies regarding AI, legislation (the EU is working on guidelines), or an AI assessment framework can be pre-sorted.
Next, starting a proof of concept is important. Outcomes are not guaranteed, but gaining knowledge about what AI can and does is important in this phase. 'The proof of the pudding is in the eating,' because it is new material. So put AI on the list for 2022!
Do you want to start using AI in your organization? Learn more about our inspiration session: starting with AI and contact us for more information about when you are interested in our session!