The Mistral AI fundraising might look like a green light, but we should always consider the downside for other players, not just the big names. A single company's success doesn't guarantee the AI Act is truly viable for everyone, especially if most of their funds are for compute infrastructure and not necessarily for compliance costs. What if smaller Belgian start-ups can't attract such capital and face huge drawbacks with new regulations? We need to know the threshold for success for average companies, not just the unicorns, otherwise, the whole system might fail for most.