Without the right use case, AI costs stay hard to control
In brief
According to Steven Latré, AI vice-president at the imec research centre, companies cannot keep AI integration costs under control if they do not pick the right use case. He said so in his opening speech at the Flanders AI Forum, stressing the need to watch token costs closely. Nicolas Deruytter, CEO of ML6, said his firm had seen that cost rise without immediate extra value.
What it changes for an SME
For an SME of 10 to 250 people, this is mainly a budget issue: without a suitable use case, AI integration can raise your costs without cutting workload. Token costs can climb before any gain appears. The risk is extra spend that does not yet bring value, as Nicolas Deruytter reported for ML6.
What you can do
- Name a specific use case before you add AI to an existing process.
- Track token costs from the first tests, not only after go-live.
- Check whether extra spend is matched by extra value before you scale.
Let's talk about the time you could save.
In 20 minutes, we find where AI could save you time or money. Free, no commitment.