[apologies for cross-posting]
The term 'intelligence' is often used in various fields including Artificial Intelligence, Cognitive Science, Natural Language Processing, and Machine Learning. But research papers rarely specify what is meant by that term. As a result, we as a research community often talk past each other, even about our basic premises and goals. Authors may even change how they write their papers, based on a wrong impression of what most other researchers believe.
We are a team of CogSci/NLP researchers from the IT University of Copenhagen, Denmark. We combed through the literature to collect the key definitional criteria for 'intelligence', and we designed this survey to find out the following:
*- to what degree do the researchers in different areas agree about what 'intelligence' even means?- what role, if any, does this term play in their research agenda?* *- what is their perception of the current systems based on large language models (e.g. Claude or ChatGPT)?*
We would much appreciate hearing your perspective! The survey takes 5-10 minutes and is available until December 18 at this link: https://www.survey-xact.dk/LinkCollector?key=4S7VSN4MS536
Your response is anonymous and will be used for research purposes only. Anonymous and aggregated responses will be publicly available under Creative Commons-Attribution-Non-Commercial license, and we will also publish our analysis.
Project team: Bertram Højer, Terne Sasha Thorn Jakobsen, Anna Rogers, Stefan Heinrich Project page: https://bertramhojer.github.io/projects/intelligence-survey/