AI will transform our lives, but we must learn how to properly use it, says Guillaume Massol, a teacher of scientific and medical English, specialising in ethics, at the University of Toulouse.
“I think there have always been fears when there is new technology. I’m old enough to remember when Wikipedia arrived, Google, the mobile phone, and now we don’t even think about them – it’s like using electricity.
“AI will be the same soon, it will be everywhere. The transition is difficult but we must adapt.
“You need not to be mastered by technology but rather learn to master it.”
Guillaume Massol
'Task, not ask'
AI is already transforming the classroom and Mr Massol has embraced its use, allowing students to do the same.
“If you use ChatGPT the right way, it’s not cheating. If you can teach them to use the tools correctly, we can empower students with something great.”
Generative AI such as ChatGPT are “brilliant tools for educators and learners”, he said. They can help to ease admin for teachers and “democratise access to knowledge”.
However, proper training is essential. Mr Massol teaches his students to “task, not ask”.
“If I had an assistant and I asked them to make me a cup of coffee, they would go to the vending machine and probably bring back a horrible coffee.
Ad
“But if I tell my assistant: ‘Go to my office and in the cupboard you will find a special coffee I want to drink at 10:00. Brew my coffee in the machine and, while it is still hot, run fast to give it to me. That is the difference.
“If you give specific instructions and you know what you want in the end, you get the best results. That is tasking.”
Avoiding AI entirely is counterproductive, he said, as companies will increasingly look to hire graduates who can use the technology.
“I tell (my students) that if you can use generative AI correctly, you will get a job.”
AI in hospitals
Medicine is where AI will “probably have the biggest impact” said Mr Massol, who teaches in the university’s medical school.
Studies have shown that AI is incredibly accurate at detecting cancers (it showed 92% accuracy in diagnosing ovarian cancer in one study) and it has even been found to show more empathy than some doctors when breaking bad news.
One obstacle, however, is training. While countries such as Sweden and Denmark include structured AI tuition for healthcare professionals, many countries, including France, are lagging behind.
“It’s not about being scared of the tool, it's about learning how to use it correctly,” said Mr Massol.
However, he acknowledged concerns, such as how best to protect data and how to use AI, proven in the past to have biases, in areas such as insurance decisions.
“We need to make sure the human is not forgotten for the benefit of economy.”
Despite AI often being used to speed up processes, it is also important to spend time going over results thoroughly, Mr Massol added.
“Sit back, read them again, make sure you are not delegating your reason, your judgement to the machine. In the case of doctors, it is so important.”
'We have to adapt to AI'
As for the enduring question about what AI means for the workforce, Mr Massol has a measured response.
“I don’t think AI is going to take our jobs, but we need to make sure we can work alongside it,” he said.
“The definition of intelligence in biology is adaptation. You need to fit your environment and AI is now in the environment.
“In medicine, management, and other fields, if my colleague uses AI and I don’t, and they are more productive, faster, better, obviously at some point there will be an issue.
“The difficulty will be training, adapting and how we fit in this whole new environment.”
