To launch Silicon’s latest issue all about the world of AI, representatives from both private and public sectors gathered at SilverSquare on Tuesday to take the pulse of The State of AI.
With an opening presentation by Daniel Stoica from RMT Labs and a panel discussion and Q&A with Fateh Amroune (AI Factory), Emilia Tantar (Luxembourg House of Cybersecurity), Monique Bachner (Independent Director & Government Advisor) and Daoud Chami (CBTW), the evening presented an in-depth look into conversations, concerns, and considerations from among Luxembourg’s AI experts.
Here are some key takeaways:
AI-powered Employment Opportunities
In his opening speech, Daniel Stoica from AI-powered recruitment company RMT Labs addressed a prevalent issue in Luxembourg’s job market: a hiring crisis. While the conversation often revolves around AI taking human jobs and replacing workers, Stoica demonstrated how when used right, AI can simplify job searches for both employers and job seekers.
“With AI, what I’ve seen is that it’s not enough to build a chatbot. We have more than 100 AI agents, and we deeply integrate them into the platform and into everything that you do in the recruitment process,” he said. “It has to be suitable from a technical perspective, but also from a personal perspective.”
Luxembourg’s Ecosystem
High praise was sung by several panelists when it came to Luxembourg’s role in AI development, and the support of bodies such as LuxInnovation for advancing AI startups and initiatives.
“There’s a whole ecosystem,” said Monique Bachner, Independent Director & Government Advisor. “Advisory panels, startup hubs, people you can sit down with and pick their brains. The community is so interconnected here.”
Regulation & Understanding Laws
Much of the discussion at The State of AI focused on regulation – how it supports AI, how it poses necessary challenges, and ultimately, how regulation can ensure a thriving environment in which AI can safely develop.
As of 2024, the regulatory European AI Act has been in place, with options for engagement with stakeholders, including through public inquiry to gather feedback.
Emilia Tantar, Chief Artificial Intelligence Officer from the Luxembourg House of Cybersecurity, said that understanding the standards of the EU market is essential for those aiming to introduce products or technologies. And while AI may seem a relatively new adoption for many, the existence of industry regulation isn’t.
“Machines which are currently on the market don’t enter the market unless they are tested,” she said. “The same will happen with artificial intelligence.”
Tantar clarified that established authorities such as CNPD (National Commission for Data Protection) are also responsible for expanding their reach to cover AI. The Luxembourg House of Cybersecurity also contributes to this process.
“In order to apply a law, you need to understand it,” she said. “And in order to understand it and to be able to test it, we are providing standards. We have issued a draft for quality management standards.”
Challenges & Opportunities
The panel agreed that while regulation can seem like a daunting start, there are benefits that come later.
“Just like any government module, it can be a bit painful at the beginning,” said Bachner, “but you will reap the rewards afterwards when you don’t have to redo your entire model.”
Daoud Chami from consulting firm CBTW shared how clients are often afraid of regulation, which makes it important to ensure they are accessible and understandable.
“Three green lights are important for any step into AI: If you’re not afraid of [regulation], if technology-wise you know where you’re going, and if, from a business point of view, it makes sense,” he said, adding that AI doesn’t need to be everywhere. “There are a lot of things that humans can do very well without AI, things that are enjoyable to do!”
Several speakers agreed that in many ways, amidst real challenges like AI hallucinations and overreliance on unchallenged information, regulation can be leveraged to build trust and confidence in AI solutions.
“Take cars, for example. At first, we were asking ourselves a lot of questions: horse pneumatic, electricity, diesel,” said Chami. “Now we have a formula that works, and we’re improving. The same goes with AI: with documentation, regulation, and testing, there will be a day where AI will be treated as any other form of technology.”
