|
Getting your Trinity Audio player ready...
|
“I am a professor at Foothill College. Until this summer, I was also a lecturer at Santa Clara University. Currently I am also an assistant professor at the University of California, San Francisco. I trained as a developmental biologist and also a molecular biologist.
When (ChatGPT) came out (around) 2022, I was already very much enmeshed in my role as an educator. I am a technologist. I am someone who likes technology. I realized that this will have an impact on education.
I’ve been using it in many different ways. At first, as a chatbot – to bounce off ideas. Last year was a dramatic change with agentic AI. Now I feel the sky is the limit because I can not just ask questions, but I can instruct it as a coworker, and then it will do a thing for me.
I use AI not to grade, but to categorize. So, (it) read(s) the output and then put(s) it into different bins. Then I go into the bin and check that it actually matches. For example, there might be a bin where everybody got everything correct. So that’s a quick check. Notice that I’m still in charge of grading. I make the final decision, but the agentic AI facilitates it. Instead of taking me two or three hours, it takes me 10 minutes. I was scared of grading before. Now grading has become so easy and quick that I can offer students something that I couldn’t before: the ability to revise.
Students use AI extensively. It’s completely widespread – I would say at least 90% if not more. A lot of my colleagues have taken that as a threat and a process that is weakening the institution and the concept of education. I disagree with that because I think it’s an evolutionary step. As a biologist, I think of evolution all the time.
But this is a cultural evolution because we, as educators, have developed a practice of using proxies. For example, the essay. We use it as a proxy to see how much the student understood and can explain it. Well, the proxy doesn’t work anymore because AI can write it for the student. So there is a disconnect between what we originally meant to test and what we are getting now that AI is available.
I think that really means education (needs) a vast rethinking. It’s not AI’s fault. I guess it’s nobody’s fault.
It’s clear to me that I cannot alone – without a computer – be as productive as together with AI. Here we go back to the role of education. (It’s) a wonderful opportunity for educators to teach students to use AI properly.
It’s clear to me that I cannot alone – without a computer – be as productive as together with AI.
giorgio lagna
In a general lab, there are a number of AI connected tools that can be used to analyze gene or protein expression. [For one research project], the idea was to use a technology that is already quite established, CRISPR, which is the ability to modify genes in cells or even in vivo.
Together with my first cohort of students, we… focused on one disease called alkaptonuria, which is a disease that leads to discoloration and arthritis, and the joints start to fall apart. There is essentially no cure.
In alkaptonuria, there are more than 300 mutations described in the same gene in different people. So, if we want to cure alkaptonuria, we will need to target all 300-plus mutations. The goal ultimately is to find a solution for all the possible mutations, whether [from] CRISPR or transgenic or some (other) approach.
In the past, this would have taken years for a team of Ph.Ds. Now it can be done with students who have a basic knowledge of biology, for whom I built a Gemini Notebook. These are built on Google’s Gemini, where I put all the papers that I want them to read, all the YouTube videos. Then they can go on and learn to use Claude Cowork….to crunch data, to analyze data.
I can build packets for them, always using AI, in which I break down an extremely complicated project and problem into little chunks that can be done within a week or maybe five or six hours of work by each student. Notice that the packet involves using Claude Cowork and validating the output. So it’s not just, ‘This is what a computer said.’ The students learn the critique aspect. They take the response and check, ‘Does this actually work?’
We just started a summer project, and lots of students are becoming interested in this because they see the utility of this. This is not just me playing with the computer. This is one day possibly curing somebody.”
Giorgio Lagna teaches biology at Foothill College and is an assistant professor at the University of California, San Francisco. He lives in Palo Alto.
This profile is one of nineteen featured in the “Humans in an age of AI” portrait series. See more profiles here.
