Jelle Demanet

The through-line

Everything Jelle Demanet says about technology runs through one question: what does this do to attention, and how fast does the brain absorb it. He reviews hardware as a cognitive psychologist rather than as a gadget buyer, and the verdict on the Meta Ray-Ban Display is double edged. It is the moment the category arrives, and it is also "eigenlijk uw smartphone op speed hè. Dus je krijgt alles constant zonder filter" [4][5]. The second half of his position follows from the first: if these devices are going to be worn all day, someone has to design the interface around how human attention actually works, and he does not think that someone has to be American.

That leads to the other recurring theme, an insistence that Europe and specifically Flanders build rather than merely observe or legislate. The dismissal of the defeatist line is flat: "Zijn veel mensen die zeggen dat we te laat zijn, maar dat is geen waar" [4][5]. His day-to-day work sits in the same place, connecting research, start-ups and established players so that innovative ideas become economic and societal impact [3].

On why this is the breakthrough moment for smart glasses

The Meta Ray-Ban Display is, in his reading, the point where the category becomes real, because the form factor and a six to seven hour battery finally make all-day wear acceptable, and because it combines a visual and a verbal interface to AI in one object [4][5]. He frames it against a long wait: "Mensen die ermee bezig zijn, dachten al 10 jaar: dit is het moment dat het allemaal gaat samenkomen. En dit is voor mij nu echt het moment" [5]. That is not the same as saying it is finished. "Voor mij is dat we zijn er nog niet helemaal, maar het toont wel al aan: het gaat niet meer weggaan" [4][5].

He is precise about what the device is not. It captures an image only on explicit request rather than continuously, unlike Google's Android XR demo, which limits the privacy exposure but also caps what it can do: real mixed reality needs depth sensors and a constant understanding of the world around the wearer [5]. Live translation runs into a hard floor of roughly two seconds of latency, and not for want of processing power, because in some languages the verb arrives at the end and the system has to wait for the sentence to finish [4][5]. The buying advice that follows is sober. For an ordinary consumer the €300 to €400 premium over the standard Ray-Ban Meta is not worth it yet, and the Display is for people who want to experiment with where the interface is going [4][5].

On attention, multitasking and the danger of the notification

The cognitive objection is the sharpest thing he says about the product. "Wij kunnen niet multitasken hè. Onderzoek heeft dat aangetoond. Wij kunnen switchen tussen taken, maar wij kunnen onze aandacht maar op één iets houden" [4][5]. A notification that appears inside the visual field while someone is driving or holding a conversation is therefore a genuine cognitive hazard, not an annoyance, and the consequence is a design obligation: the timing and the placement of messages have to be rethought for this medium rather than inherited from the phone [4][5]. This is the direct counterweight to the unfiltered stream he describes when he calls the glasses a smartphone on speed [4][5].

On brain plasticity and interfaces that become intuitive

The flip side of the attention risk is how quickly adaptation happens, and he offers himself as the evidence. "Ik had zelfs na een paar dagen als ik een berichtje voelde binnenkomen op mijn GSM dat ik al rechtsonder begon te kijken zelfs als ik mijn bril niet aan had" [4]. He reads that as a demonstration of how flexible the brain is, and as the reason these devices could end up more intuitive than smartphones rather than merely faster [4][5]. It also underlines the open problem: there are no UI and UX standards for smart glasses yet, so the conventions being laid down now are the ones brains will wire themselves around [5].

On regulating after the experiment, not before it

His line on European rulemaking is unambiguous: "Vanuit de wetgeving gaat er nooit geen innovatie komen. Dus je moet wel de grenzen verleggen in use cases" [4][5]. The sequence he wants is experimentation first, then critical thought about the dangers, with the European AI debate as the cautionary example of legislating first and choking innovation in the process [5]. That position sits alongside, rather than against, his concern for protection of creators: shielding the intellectual property in your creations, insights and inventions is, he argues, a very important theme in the age of AI [2].

On building sovereign European hardware

Rather than trying to move first, he proposes a deliberate second-mover strategy. Let the Americans launch and iterate, watch which use cases actually work, and then "We kunnen stelen met onze ogen en iets beters maken die beter past bij wat dat wij echt willen dat die devices kunnen" [4][5], producing a compliant, sovereign version [5]. The industrial argument underneath it is that the foundations are already here: "Als je beseft dat de meeste grote techbedrijven eigenlijk draaien op basiskennis die in Leuven door IMEC ontwikkeld is, dan zitten we toch in een ideale positie om ook hier in Europa hardware te beginnen ontwikkelen" [4][5]. What is missing is capital appetite, and he says so bluntly: "In plaats van allemaal in vastgoed investeren en aan een appartement, misschien moeten wij eens meer overwegen om ook op hardware start-ups te durven investeren" [5].

On Flanders as an experience innovation cluster

He makes a specific regional claim rather than a generic one. Flanders can be a world leader in experience and entertainment innovation because of the density of the weave: roughly 16 research centres, a start-up network, and globally known names including IMEC, Barco, Kinepolis, Stageco and Tomorrowland all sitting close together [4][5]. His enthusiasm for the work of connecting them is undisguised, describing it as a top job to work with the cream of innovative, entertaining Flanders to turn innovative ideas into economic and societal impact, with a top team [3]. The same instinct shows in the settings he champions, such as bringing an intellectual property programme with major speakers to the Lab of Tomorrow buildings a stone's throw from De Schorre in Boom [2].

On mapping technical talent

Alongside the hardware argument sits a quieter interest in people and skills. He points maatwerkbedrijven and regular companies alike toward the skillscan project run with Spatial Dynamics as a new way to map technical talent, explicitly covering both lower-skilled and higher-skilled workers [1].

Takeaways

• Judge smart glasses by attention, not features: humans cannot multitask, only switch, so notification timing and placement in the visual field must be redesigned or they become a real danger while driving or talking [4][5].
• The Meta Ray-Ban Display is the arrival of the category rather than its finished form: six to seven hours of battery and an acceptable form factor mean "het gaat niet meer weggaan" [4][5].
• Skip the Display as an ordinary consumer; the €300 to €400 premium over the standard Ray-Ban Meta is only worth it if you want to experiment with the future interface [4][5].
• Two seconds is the practical floor for live translation latency, because languages that put the verb at the end force the system to wait for the sentence to end [4][5].
• Europe should be a deliberate second mover: let the Americans launch, watch the use cases, then "stelen met onze ogen" and build a sovereign, compliant version on IMEC's chip foundations [4][5].
• Regulation follows experimentation: "Vanuit de wetgeving gaat er nooit geen innovatie komen" [4][5], while intellectual property protection becomes more important, not less, in the age of AI [2].
• Redirect investment appetite from property to hardware start-ups if Europe wants its own devices [5].
• Flanders' cluster of about 16 research centres plus IMEC, Barco, Kinepolis, Stageco and Tomorrowland is the concrete basis for global leadership in experience innovation [4][5].