Everyone is talking about artificial intelligence (AI) and Chat GPT these days, and there is much speculation (and hand-wringing) about how AI will affect business, education, and the job market worldwide. AI will – and already is – affecting the evaluation field, but it’s too early to fully assess these effects and the best ways to use AI in evaluation.
As discussed in our recent blog post about data collection design, AI is far from replacing human evaluators. “The job of an evaluator is to make value judgements, and AI doesn’t have values,” the post reads. “The culture and context of the people from whom you are collecting the data is vital to analyze the data accurately.” In Khulisa’s recent gLOCAL seminar, evaluator Margie Roper also points out that AI is still mostly dependent on data that’s already been collected, making the data collection phase of an evaluation more important than ever.
But there are already some great AI tools available to help evaluators do their jobs efficiently and well. Below, we’ve compiled a list of AI tools that Khulisa evaluators find useful for various purposes:
Finding and synthesising literature:
- Semantic Scholar – access to 200 million research papers, links between topics, recommendations based on recent searchers.
- Scholarcy – article summariser.
- Elink.io – saves content from web articles, videos, cloud files, social media.
- Elicit – find relevant papers.
- Scite – where articles have been cited and finds supporting and contrasting evidence.
- SciSpace Copilot – multi-lingual AI tool including comprehending maths/tables, easy summaries.
Transcription:
Data analysis:
Academic writing assistants (they won’t write for you!):
Tune in to the Khulisa website and social media channels for future updates about AI and evaluation.