AI in the Workplace: 5 Learnings for the Cultural Sector

1. AI is not a search engine: understand what your tool does

Many people use ChatGPT as if it were Google. But ChatGPT does not search for facts: it predicts words based on language patterns. This means the output is not necessarily accurate.

Geert Wissink emphasizes:
“When ChatGPT emerged, many people saw it as a smart search engine. But it is a language model, a word predictor. If you don’t understand this, you use it incorrectly: you think you're getting facts, but you're reading a probable answer. It’s like using a hammer as pliers.”

What does this mean for cultural organizations?

• Start small and practical: use AI to summarize meeting notes, structure audience feedback, or draft an initial concept text.
• Ensure human oversight: never blindly trust AI output.
• Invest in AI literacy: understand what a model can and cannot do.

2. Creativity remains human, AI helps you get started

AI does not replace creativity but can serve as a springboard. It helps generate ideas faster or unblock stalled thinking processes.

Anne Doeser compares it to a familiar frustration:
“Do you know that roll of tape where you can’t find the start? That’s AI for us. It helps us get started. We never take the output literally, but it gets us moving.”

Concrete example for cultural professionals:

• Ask an AI model to come up with ten diverse seasonal programs for different target groups.
• Use the suggestions as a source of inspiration, not as the final result.

AI is thus a valuable springboard for creativity, but you decide what is truly valuable.

3. Prioritize autonomy, ethics, and safety

AI also raises questions: who is responsible if something goes wrong? What happens to your data? And how do you stay in control?

Rodger Werkhoven warns:
“AI is not intelligent. It reproduces patterns but doesn’t know what it’s doing. Humans must always retain the curating role.”

Both Rabobank and the Amsterdam University of Applied Sciences have new AI applications reviewed by an ethics committee. For the cultural sector, these are the key questions:

1. Where is data stored?

• Avoid tools that are vague about data usage.
• Where possible, choose European storage or open-source models.

2. Who has access?

• Do not share sensitive information in tools where you don’t know the data flow.
• Train your team in the safe use of AI.

3. What is the business model?

• Paid tools are often safer than ‘free’ tools that use data to enrich their AI database.
• Avoid tools that automatically make all output ‘public domain.’

4. Does the human remain ultimately responsible?

• Ensure that AI never makes the final decision.
• Document when AI played a role so you can remain transparent.

4. Build knowledge and experiment step by step

AI often evokes high expectations and fear. But it pays off to start small, safe experiments. Think of:

• Using AI for administrative tasks (summaries, schedules).
• Enriching brainstorming sessions with alternative perspectives.
• Creating audience profiles faster using audience data.

This way, you gradually develop expertise and support within your organization.

5. AI requires collaboration and knowledge sharing

The cultural sector doesn’t have to do this alone. Large institutions are already experimenting with safe AI environments, but smaller organizations can also learn from each other. By sharing experiences – successes and failures – we collectively build a responsible application of AI. Also, consider intensive collaborations with tech organizations and other sectors. They are often more open to this than is commonly thought.

Conclusion

AI is not a magic solution, but it can help the cultural sector move forward. From working more efficiently to creative inspiration: the possibilities are vast, as long as we remain critical, conscious, and human-centered.

This article is a summary of the original article by Hub-C (opens in new tab)