Mr. Brickwedde, what are the current trends in recruiting?
A current trend is, of course, the use of AI in recruiting. In the English-speaking world, there is a movement towards skill-based recruiting and skill-based organizations. This is not yet widely adopted in Germany but will definitely come.
For example, a watch manufacturer was looking for a finisher, a fine polisher. It turned out that dental technicians had similar competencies. The talent pool is being expanded based on skills and increasingly less on certificates and degrees.
What is AI already being used for in recruiting processes today?
In skill-based recruiting, AI could suggest which professions have overlapping skills. But this is not the main area of application. Most of the time, AI is used to create job descriptions and for active sourcing with the creation of hyper-personalized cover letters. The AI compares the candidate’s profile with the search profile and then writes a corresponding text. This takes about two minutes and can even be automated nowadays.
There are many other applications as well. In the book KI-Revolution der Arbeitswelt: Perspektiven für Management, Organisation und HR I have dedicated an entire chapter to this topic.
So, you think AI is a resource for recruiters?
It is a relief for recruiters. It allows them to do what they do faster and better. Poor recruiters become average with AI, but good recruiters remain better than AI.
Is the use of AI a matter of money?
Only if $20 per month is a matter of money. That’s the cost of a pro account with ChatGPT.
You mean that ChatGPT is enough?
It’s a start. In the past, people said, “AI is coming somehow.” Last year, everyone could try it out themselves. When thinking on a larger scale for companies, introducing a Talent Intelligence software requires a significant amount of money.
Is AI also used to optimize Key Performance Indicators (KPIs) and integrate scientific studies into recruiting?
Optimizing KPIs is difficult because AI always needs large numbers. If a company has 500 applicants, that’s nothing for an AI. You can measure KPIs, but the AI can’t make significant improvements because there are too few data points.
It’s different with Amazon. They have many applications and a lot of data. It depends on the company from which the AI can learn. It must learn from the mistakes made.
However, this was a problem at Amazon some time ago. They analyzed all their employees to see who was successful. They then used AI to look for new hires that matched that profile. The successful employees, however, were all white men who had studied at a certain university. As a result, applicants of color, women, and those who studied at other universities were sorted out. This was not the intention, but the AI couldn’t do otherwise. Therefore, training the AI is important for proper optimization.
So, the AI still has a lot to learn on a large scale?
Yes, due to the  General Data Protection Regulation (German: DSGVO – Datenschutz-Grundverordnung) hiring decisions cannot be made without human intervention. AI is not allowed to do that. But it still has a lot to learn. AI is not inherently intelligent. Especially LLMs like ChatGPT make things up that are likely. Particularly when it comes to fact-based matters, AI can sometimes be very bad.
You can’t trust AI too much?
Absolutely not. There are even studies that show recruiters become worse when they rely heavily on AI. They trust the AI without properly verifying it afterward.
Is this one of the most important insights that you will present as chairman during the webinar days?
Yes, I would include that.
What else belongs to your most important insights?
As a recruiter, you should engage with artificial intelligence but always remain critical. When marketing people make claims, you have the ability to question them. Critical positive use of AI is my advice.
Is there a lot of marketing behind the use of AI in recruitment?
Yes, absolutely.
What are you still curious about in the future development?
It’s a bit like the steam engine. Something was invented to produce energy. But what it can be used for will only be known in five or ten years. It will evolve, and I find it exciting to see what can come of it.
But AI is not as advanced as many think?
No, not at all. It’s still in its infancy.
What should employers do and what should they avoid?
They should explore the possibilities of AI. Most often, they first try to digitize the recruiting processes. Then they have poor recruiting processes digitized. That makes no sense. So first, review and streamline the process. Then identify what involves a lot of manual or repetitive tasks to use AI for that. Employers should approach the topic of AI with a critical yet positive attitude.
Written by Lydia Stöflmayr
