When the world around us changes, so does truth and reality.
Jason Lodge has redefined the shame and social pariah status that often arises from the retraction of a paper, instead demonstrating that all responsible thought leaders need to recognise when their past publications are no longer correct.
The University of Queensland Professor, recognised as a national leader in application of technology and in particular AI to learning, has publicly announced that he asked the Tertiary Education Quality and Standards Agency (TEQSA) to retract a paper on options to preserve academic integrity, in recognition that world had changed.
While there are scientific writings over millennia which have been subsequently proven wrong by changes in environment or expansions of knowledge, most of the authors have been long gone before having to concede any error.
However, a new era of intellectual humility is required for AI experts, as the unprecedented pace of knowledge advancement and change can render comparatively recent papers wrong or redundant.
“Things have moved quickly over the last four years,” Professor Lodge writes on LinkedIn
“I'm sure I am not the only one who has changed my mind about some ideas and approaches. Continuing to endorse ideas or approaches well beyond when the evidence suggests that they are no longer viable, in my mind, is a form of malpractice.
“In that vein, the resource "The evolving risk to academic integrity posed by generative artificial intelligence: Options for immediate action" has been removed from the TEQSA website on my advice.
“Why?
“The resource was created to respond to a particular moment in time. That time has passed. While some of the ideas in that resource have proven well-founded, others no longer hold. I was wrong.
“In particular, I am convinced that it makes no sense to use the AI Assessment Scale to approach assessment design or (in particular) to address academic integrity issues. Keeping a resource published by the regulator available that endorses approaches that do not align with the evidence would be a form of malpractice. It needed to come down.”
Professor Lodge’s step was has been widely recognised as academic leadership in action and he told Future Campus he was concerned that oversimplified responses to an enormously complex problem were not helping.
“Over the longer term, the idea that we can codify AI use into a simple classification system or detect/limit/ban AI use when some people (including students) are using swarms of hundreds of agents in semi- or fully autonomous clusters to do work seems unrealistic to me.
“If we are ever going to adapt to the age of AI, we need much more sophisticated, nuanced approaches to handle the numerous ongoing challenges we face. If there were a solution to all of this, everyone in global education systems would be talking about it. There isn’t.
“In the meantime, I think the least I can do, as someone deeply invested in all this, is admit that I am still trying to figure it out. Often that means being willing to say I’m wrong, and I do so gladly. After all, it would be hypocritical of me not to, given the research on how powerful learning from errors can be.”