AI programs like ChatGPT and Perplexity can produce text that sounds reasonable based on language models. The possibilities seem endless. Such artificial intelligence would likely have all kinds of consequences for society.
But NWO staff are not permitted to use such software when evaluating research proposals, which is what the interim position now states. All documents received by the referees, panelists and jury members are confidential. Uploading them to an AI program may result in a breach of confidentiality.
The position will be on the NWO website on Tuesday Developed. Scientists are permitted to use generative AI in their research work, “given the capabilities and potential for development.” As long as it operates transparently and can be verified.
NWO Board Member Antal van den Bosch, professor at Utrecht University, is an expert in the field of language and artificial intelligence. “Here we draw a line” He says “But at the same time we see great opportunities now and in the long term for the field of AI itself and for the broad application of AI in science,” he said on the NWO website.
At the same time, there are “problematic aspects” of the technology. He also calls this a “scientific challenge.”
The interim guideline may change or be given more nuance. The NWO has established a working group for new AI policy. This policy should be ready in the second half of 2024.