Does AI Create Equality or Inequality in the Workplace? Five AI Experts Consider This Question. Employees Will Have to Learn to Deal with New Technology, but Also with New Expectations.
The experts at our round table consider it a certainty that AI will have an impact on office work. Lander ‘T Kindt, co-founder of the Belgian AI startup Donna, sees evidence in his personal sphere. “My mother has no technical training at all, but she was also an early adopter of ChatGPT. Today’s tools are developed so intuitively that it’s easy for non-technical people to embrace them.”
ITdaily brings five experts from the Belgian AI sector to the table to discuss AI in all its facets. ‘T Kindt is joined by Gianni Cooreman, Presales Director at Salesforce Benelux, Joachim Ganseman, Research Consultant at Smals, Maarten Callaert, co-founder and COO of Paperbox, and Christophe Robyns, Managing Partner at Agilytic. In addition to a semantic discussion about the term AI and the dangers of do-it-yourself projects, the impact on our work and life is also addressed: a very hot discussion.
My mother has no technical training, but she too was an early adopter of ChatGPT.
Lander ’T Kindt, medeoprichter Donna
In the Background
As an office worker today, it’s almost impossible to escape AI. Companies like Google and Microsoft are practically forcing it down your throat. Yet paradoxically, AI can gain more approval when it keeps a low profile, Robyns notes. “A combination of RPA with LLM technology for text applications works well in our experience. AI has the most effect on fixed processes in the background. Less interaction convinces people to use it more quickly.”
For Callaert, the key to success lies in the use case. “A lot is coming our way. If you can link AI to a good use case with a simple interface, it will offer success on both business and personal levels. The areas of application are legion, but the challenge lies in determining the use case.”
‘Tasks that people don’t like to do are precisely the best ones to automate,’ ‘T Kindt agrees.
Natural Advantage
Should everyone then train to become an AI expert? Ganseman adds nuance. “Study what you like, but those who are handy with AI will have an advantage. I’m not worried about young people in this regard, but rather the older workforce. When people who are used to working with AI tools from a young age enter the job market, they have a natural advantage over older generations if they don’t keep up. I advise everyone to keep experimenting and look beyond ChatGPT.”
read also
We’re Back to Having Many Bob the Builders: The Risks of DIY in AI
Artificial intelligence becomes more powerful when used by someone with experience in their field, Robyns believes. “Without knowledge and experience, you run into the limitations of the tools more quickly. Look at developers, for example: many use Copilot or other AI tools. That works well if the user knows what they’re doing. Without knowledge, you harm the basic principles of software development. We need to maintain the knowledge; that’s more important to me than which jobs will or won’t be influenced by AI.”
‘A discussion has arisen about whether an IT education is still relevant,’ Cooreman adds. ‘You need some knowledge to know if the output of an AI system is well-constructed. A layperson just presses a button and has no idea what they’re doing. The technology is accessible, but if you’re not aware of the limits, it becomes worrying.’
Cooreman believes that people should also become more aware of the ecological impact of AI. ‘You can ask LLMs for information about everything instead of looking it up yourself. People just do whatever and start using LLMs for things that are totally unnecessary, without considering the sustainability aspect.’ There’ doesn’t seem to be any ‘AI shame’ today.
If you’re not aware of the limits of AI models, it becomes worrying. A layperson presses a button without knowing what they’re doing.
Gianni Cooreman, Presales Director Salesforce Benelux
Learning by Doing
“AI models are actually pleasers that subtly feed your own convictions. You need to remain critical and curious and verify things to gain experience,” says Callaert. “Otherwise, you risk being led down the wrong path,” Ganseman chimes in. “Then you keep going until you eventually get stuck and lose more time than you gain. That’s why you need to build experience to know when you’re being misled, so you don’t get carried away.”
Learning to work with AI is done by doing, is the general conclusion during the roundtable. ‘T Kindt: “There’s no book yet on how to work perfectly with AI. You only learn it by engaging with it yourself, discovering mistakes, and guiding models. That’s why we develop a trajectory with our clients to provide them with some best practices.”
“Some users approach it more intuitively while others need more guidelines. Onboarding and setting the right expectations remains a challenge to provide the user with the best possible experience. As vendors, we certainly share a responsibility in this, but ultimately the user must discover for themselves how AI can or cannot work for them,” concludes ‘T Kindt.
“It’s about all aspects. The triangle of human-technology-process must come together,” adds Callaert.
Daens-like conditions
At the end of the conversation, the moderator puts on his brave face. A year ago, during our previous AI roundtable, it was still said with full conviction that AI would not threaten jobs. Meanwhile, there are companies using AI as an argument to lay off people. Is there a risk of a massive AI layoff round after all?
read also
From Agents to Your Thermostat: AI Cannot Be Put Into One Box
The panel doesn’t see it happening. “Just look at the government: we’ve been automating since the 1950s, and that has never led to fewer civil servants,” Ganseman says with a wink. “There will be a shift, and possibly fewer people will be needed for certain tasks. But that frees up space for other things.”
Ganseman illustrates with another example. “Thanks to apps like Google Translate, today everyone can translate with their smartphone. Does this mean interpreters are no longer needed? Their focus has shifted to verifying and improving automatic translations. Tasks that were initially seen as less of a priority can later become important. In my opinion, the impact on jobs will be limited, but the content changes. I don’t see us returning to Daens-like conditions.”
“Some tasks will become obsolete, so people will have to give meaning to the time that ‘becomes available’. The pressure will certainly not decrease. Everyone has the means to work more efficiently and intelligently. That brings new expectations from employers. We’re all guilty of that, myself included,” concludes Callaert.
“We’re not going to be able to retire yet, as nice as that might sound,” jokes Cooreman. “The situation is different for each company: in some sectors you have cyclical capacity, such as retail and call centers. Healthcare personnel struggle with stress and burnouts due to high workload, meaning patients don’t receive individual care. If you can shorten the onboarding time with AI and get the maximum out of available resources, it benefits customer service. Gaining efficiency in IT enables gains elsewhere.”
The last word goes to Ganseman. “Certain profiles will find it harder to keep up than others. That’s why we need to sit productively around the table to find possibilities for everyone. Let’s not make it a competition.”
We are not going to return to Daens-like conditions, but the content of jobs will change.
Joachim Ganseman, Research Consultant Smals
This is the third and final editorial article in a series on the theme of AI in practice. Click on our theme page to see all articles from the round table, the video, and our partners.