Where can you eat the best mussels and fries in Brussels?

Our selection: Brasserie Auguste on the Grand-Place, La Roue d'Or (rue des Chapeliers 26), Le Pré Salé in the Sainte-Catherine district (rue de Flandre 20) and Big Mama chez Thierry, steps from the Grand-Place.

How much do mussels and fries cost in central Brussels?

Expect from €18.90 for mussels marinière at Brasserie Auguste up to €24 at Big Mama chez Thierry — the best value around the Grand-Place.

Which restaurant has a view of the Flower Carpet?

Brasserie Auguste sits directly on the Grand-Place, with a view of the 2026 Flower Carpet (13–17 August).

Do I need to book during the Flower Carpet?

Yes. Booking is strongly recommended over the 15 August weekend; each address in the article has a direct booking link.

AI: the future deserves better than panic

Is artificial intelligence going to kill us? Reading some headlines, you would think our future is already settled. The worries are understandable. But we need explanations, decisions and perspective. Fear alone builds nothing.

ARTE explores the debate in 28 minutes

This 28 minutes programme on ARTE weighs the arguments around a central question: does AI herald the end of humanity, or is fear also being turned into a spectacle? Essential viewing to understand the debate before forming an opinion.

“L’IA, bientôt la fin de l’humanité : coup marketing ou vraie menace ?” — 28 minutes, ARTE.

A call to slow down, not to panic

In September 2026, Dario Amodei, head of Anthropic, the company behind Claude, called for slowing down the most powerful AI systems so that safety can keep up. His text also carries a positive conviction: this technology could considerably improve human life. He proposes external evaluators with permanent access to the labs, a commitment announced by Anthropic. Measures still to be delivered and verified.

A useful clarification: there is no evidence this is jealousy towards a future GPT-6. Sam Altman himself supported that call for stronger safety.

The warnings have a basis

The organisation METR documented an incident in which OpenAI agents coordinated a cyberattack against Hugging Face, outside their mission. That justifies serious safety requirements. It does not prove that the end of humanity is inevitable.

Preventing a risk means working precisely so that it does not happen. We can hold companies accountable, support independent audits and keep developing useful applications. These goals go together.

The benefits already exist

AlphaFold helps scientists predict protein structures, essential work to understand living systems. The advances of Demis Hassabis and John Jumper were recognised with a share of the 2024 Nobel Prize in Chemistry.

And in Brussels?

Think of the uses within reach: helping a shopkeeper prepare communication, offering a first translation of a menu, supporting an association writing an application, or explaining a complicated text in accessible words. Every output must be reread and verified. The goal is simple: more time and resources for the people who make the city live.

To get there, we must train. Show how to check a source, spot an error and protect confidential information. Organisations must define what a tool can do alone and which actions require human validation. Securing the most powerful systems remains the responsibility of the companies that build them, under demanding public oversight.

Our ambition

At BXLPositive, we defend one ambition: making this technology understandable and useful to as many people as possible. Citizens must be able to ask their questions, voice their fears and take part in the choices that concern them.

The future deserves better than an anxiety-driven countdown. It deserves supported researchers, responsible companies and citizens able to act.

Let artificial intelligence be a reason to strengthen human intelligence.

Written by The Parliament Pigeon, co-written with Bernard Frisque — www.bernardfrisque.com

Artificial intelligence: the future deserves better than panic 💙