2023-07-09 02:36:15
By Salvador Macip. Director of Health Sciences Studies at the UOC and professor of molecular medicine at the University of Leicester.
AI is neither good nor bad, as any technology depends on how it is used. We must find how to use it to move faster.
It is a conversation that cannot be avoided these days: how does the advent of artificial intelligence (AI) affect our work? I see this impact at different levels (research, teaching, medicine and literature) and, once the initial alarm has been overcome, my impression is that chatbots like the ChatGPT of OpenAI are a shock that was sorely needed in all these areas.
In research, we have been taking advantage of the tools provided by AI for a long time, but at the end of 2020 we experienced a revolution similar to the one we are seeing now in other areas, when the AlphaFold software demonstrated that it could predict the structure of any protein with an accuracy close to experiments that need months of work and machines that cost millions. Instead of seeing it as the loss of jobs for thousands of experts and the death of structural biology, a discipline with more than a hundred years of history, AlphaFold was quickly incorporated into protocols as a complement to studies that have yet to be done. in the traditional way, which has allowed us to accelerate key work for the discovery of new drugs. In this field, then, the debate on the threat of AI seems to be over.
Regarding the world of teaching, the abilities that ChatGPT has to answer questions that do not require analysis are spectacular, and this has exposed the weaknesses of the current educational system, often based on rote and uncritical data collection. When, a few days ago, I was designing the final project for the subject that I teach in the last year of biomedicine at the University of Leicester, and which is an important part of the note, I had to take a new step: enter the question in ChatGPT to see if it solved it. And he did, but with a fair pass. I tweaked it a bit and then it just didn’t make the cut. The trick is simple: ask the student for things that a machine cannot do.
Education has been tending towards a comfort zone, the result of lowering the level of the necessary requirements to progress. But, from primary school to university, school should be a journey that activates the ability to reason, connect and reflect, to integrate information to achieve an imaginative vision that proposes new solutions and interpretations, not a place where you simply teach to collect and regurgitate data. Now that we’ve seen that this is already done by software, we need to re-emphasize training brains to do what they do best: think.
In medicine, AI is already capable of proposing possible diagnoses based on a list of symptoms. This does not announce the disappearance of the figure of the doctor, on the contrary: with this initial work already done, he would have to have more time to invest in the human part of medicine, which we have lost as the public system has been cut back. resources and was overloaded. Treating a patient as a person and not as a series of figures should be the essential skill of the profession. The fact that an AI can get a good grade in a MIR exam has to force us to rethink how we define what makes someone a good doctor.
AI is neither good nor bad, as any technology depends on how it is used. Pretending that it does not exist or trying to restrict its use is as impossible as it is counterproductive. Just as we’ve integrated the internet into everything, to the point that now we wouldn’t know how to do without, we have to find ways to use AI to move faster. It is as if 50 years ago we had prohibited the use of calculators for fear that they would numb our neurons, instead of seeing them as an instrument to free ourselves from tasks that prevented us from investing time in more creative activities.
The AphaFold example should be transferred to other disciplines: instead of fearing AI, let’s jump headlong into the challenge. We have to celebrate the popularization of AI tools because they force us to redefine how we work and, above all, how we educate. It is a golden opportunity to stop seeing people as machines and to dehumanize students so that they become these machines. It is the push we need to re-use the areas of the brain that we have allowed to become locked up.
I’m done with literature. At the moment, AI cannot compete with human creativity running at maximum revolutions, it only produces rehashes of ideas that have been around for a long time. But this shouldn’t be much of an obstacle because, in fact, many of the best-selling books already seem to be written by an AI. Perhaps this is where we will notice less differences.
#artificial #intelligence #Health #Medicine