AI as a Superpower for Your Business

Since the 1960s, universities have been researching how to make predictions about the future based on collected data. It is the seed of artificial intelligence, because that's what it's about: programming a system that makes predictions, recognizes patterns, and learns based on data analysis. Today, AI is experiencing an exponential acceleration, with generative AI and deep learning. Maarten Verschuere, founder of Cleverland.io, tells us how AI will give us superpowers in the future.

How has AI shaped our professional lives in recent years?

Firstly, by data analysis and forecasting to support decision-making. Companies are increasingly basing strategic planning on data analysis: about production, sales, customers, and products. They take into account variables such as time, weather conditions, product layout, holidays, and so on. Too much data for a human, but AI has almost no limit on the amount of data it can process and quickly brings the desired insights and forecasts. Companies that do not yet have data benefit from AI models such as a smart accounting package that can read and classify documents thanks to AI-driven computer vision.

Secondly, it has greatly relieved us by automating routine tasks where you can recognize a pattern. Accounting is an example, but also robotics such as in production in a factory. In the past, it was mainly sorting, but today you can perfectly train an AI-driven machine to, for example, plant plants flawlessly at high speed. Customer service is also a good example. Chatbots then answer FAQs in customer service, like Kate from KBC. These questions often account for 80% of incoming queries. This allows employees to focus on work that requires insight, creativity, and judgment.

What is deep learning and what is its most well-known application?

Deep learning is a technology within AI where you train computers and systems to learn based on examples. These examples are fed to an artificial neural network with multiple layers, which imitates the functioning of the human brain. This allows them to analyze and recognize images and information better and better. The most well-known application is computer vision technology. Computers and systems learn to store examples from countless images and videos and based on that to perform actions or make recommendations – such as recognizing how to hold a plant in the correct position, where to touch it and where not, and how to plant it. Self-driving cars also capture the imagination.

What AI applications are being worked on in 2023? What is being invested in the most?

There is a lot of focus on Natural Language Processing (NLP). Thanks to NLP, a computer program can understand text or speech data and respond to it with text or speech – for example, for drafting texts and documents.

ChatGPT is very well known. On the one hand, it is language processing - pattern recognition in texts - and on the other hand, it is language generation - pattern generation. Based on a well-defined question (a so-called 'prompt'), the program can produce a fully written piece of text. ChatGPT's brain was trained by reading thousands of texts available on the web. This way, the system learned how texts are structured and can also write texts itself.

This can accelerate AI, because if everyone creates content faster, the amount of information that is spread and picked up will grow exponentially. Social media brought an explosion of images; now we will get an explosion of written content. But it will be little original work, it will mainly accelerate info sharing. Papers, business texts, purely informative texts, all of that can be automated.

Won't we then have an explosion of fake news? Doesn't quality and source control become extra important?

Absolutely. You can feed ChatGPT completely wrong statements – the earth is flat, there is no gravity, and people are unequal – and get very credible texts from ChatGPT. Fortunately, AI is also able to check when AI is used, among other things by cross-checking with other texts. If the story is not supported by a sufficiently large diversity of sources, it is identified as fiction. But we must of course continue to think critically. Many people are easily misled.

In the school environment, you can already write essays via AI, that can't be stopped. More important is to teach children better skills in critical thinking, fact-checking, and source research. Better train them to handle information. That will work, because they are in full learning mode. The current generation is more susceptible to deception. When you see how Cambridge Analytica was able to identify groups susceptible to fake news, to then influence them with information tailored to their personality to help Donald Trump come to power... That is data-supported manipulation. We must be mindful of that.

What lies ahead in the next 10 years? What is the next shift in the professional landscape?

We are on the eve of a very big leap, as with the industrial revolution. First, all the hard work had to be done manually, then machines in factories took over some of the work but hundreds of people still stood on the factory floor. Today, you see only a handful of people in those factories, who operate the machines.

The same thing is happening now with office worker jobs: AI tools take over repetitive tasks and office workers get a more responsible, creative job. In the past, the blue-collar worker lost their job due to automation; now it's the white-collar workers' turn: administrative staff, help desk staff, but also programmers and graphic designers. Their job will transform into more creativity and responsibility. Previously, programmers had to program everything themselves. With AI, they can describe what they want to code and the coding happens for them. Also, writers of purely informative texts will soon see those tasks automated. Text writers will become people who produce unique texts, based on original conversations and with input from human creativity.

So, this progress gives us space to develop new skills, and makes our work much more efficient. AI gives us superpowers and we go much further in our field.