AI generated image of a wet lab grad student.

AI generated image of a wet lab grad student.

What ChatGPT Thinks About Me

Tuesday, March 7, 2023

Over the years, many people have asked me what a grad student does, and I have been stumped for answers. Half the time, I think I know what I am doing but I don't really do. With all the hype surrounding ChatGPT, I thought it might be fun to find out what ChatGPT thinks I really do.

Since UCSF research has both computational and bench work, I thought to include both so no one gets left out. Also, I bet everyone did try at one point to learn how to code during the pandemic, so here we are.

As usual, this is an article no one asked for but I shall provide my unsolicited commentary on ChatGPT’s idea of my life.

Prompt: What is the daily life of a 50/50 computational/wet lab grad student?

ChatGPT: The daily life of a 50/50 computational wet lab PhD student can be quite varied, as the program involves both laboratory experiments and computational analyses.

Here is a possible schedule for a typical day in the life of a 50/50 computational wet lab PhD student:

8 AM: Arrive at the lab, put on personal protective equipment (PPE) like lab coat, gloves, and safety goggles.

JG: First, if you are in academia, who comes in at 8am? 8am is a weird time. You either show your dedication to the lab and come in at 7am or you show up at 10.30am fashionably late since no good science starts before 10am. If you are partially computational, you can also stroll in fashionably late saying you were coding in the morning and wanted to have coffee.

Also, gloves are usually a very sufficient PPE for most things. (EHS please don’t come after me.) Please don’t scroll through this article with your gloves on though.

ChatGPT: 8:30-10:00 AM: Perform wet lab experiments, which may involve pipetting, gel electrophoresis, PCR, Western blotting, or cell culture. Alternatively, analyze data from previous experiments or design new experiments to test hypotheses.

JG: Aside from the time, this is kind of accurate. I used to get into lab and spend the first half hour warming up my cell culture media in the water bath, then have an existential crisis about what in the world I was doing. But hey, warming up cell culture is still part of the job, so I was being productive. That being said, I feel like no one completes PCR or Western blot in 1.5 hours from start to finish. If you are able to do it, please tell me how. And no, outsourcing it to the rotation student is not the solution.

ChatGPT: 10:00-11:00 AM: Attend a lab meeting or journal club to discuss the ongoing projects, troubleshoot problems, or learn about new research techniques or technologies.

JG: Group meetings, either a large lab meeting or smaller group meetings with fewer people weekly are common practice in science and a great way to troubleshoot or brainstorm new ideas. 10am is a weird time though. But hopefully there’s food.

ChatGPT: 11:00 AM-12:00 PM: Take a break for lunch, maybe with other students or colleagues, or attend a seminar or workshop on topics related to the research.

JG: If you are a wet lab person, lunch can be a varied time, anywhere from 10am to 2pm. Do not conform to societal norms about lunch!!! Also, lunch seminars at 12pm without lunch are sad.

ChatGPT: 12:00-2:00 PM: Perform computational analysis of the data generated from the wet lab experiments. This may involve writing code in a programming language such as Python or R, running statistical analyses, or using bioinformatics tools to analyze sequencing data.

JG: Somewhere in the middle of the day, you need to include drinking coffee too and potentially goofing off with a lab mate. But yes, a lot of researchers these days do use Python, R or Matlab to run their analyses, all of which can include bioinformatics tools to analyze sequencing data.

Genomic sequencing is a very useful tool, but this is also very presumptuous to assume that everyone does it. For every sequencing run done, a grad student somewhere in UCSF becomes at a higher risk of drowning in big data. Some people just prefer good old hypothesis driven approaches and bar graphs. That works too.

ChatGPT: 2:00-3:00 PM: Meet with the advisor or mentor to discuss progress, challenges, and future goals of the research.

JG: Advisor? What advisor? Just kidding. This highly depends on your PI’s mentorship style and how often they are willing to meet you. Some people have weekly to monthly hour-long, one-on-one sessions. Others only meet their advisor for 15 minutes every 6 months.

Usually, this time is also spent complaining of your tiredness, how old you are and how much work you have not completed because you spent your morning staring at an empty spot on your bench wondering if you should clean it.

ChatGPT: 3:00-5:00 PM: Continue with computational analysis or design new experiments, depending on the stage of the research.

JG: Usually if you are doing a long wet lab experiment, you just continue at this point, you are in the zone. This, however, is the time where people get jealous of dry labs as it can be time for an optional tea break with your lab mates. And maybe cake. That’s awesome.

ChatGPT: 5:00-6:00 PM: Attend a lab or departmental meeting or meet with other students to discuss projects or collaborations.

JG: Nonononoono. 5pm is time to wrap up your computational work so you can go home and continue to work or grit your teeth and stay to finish your experiment. ChatGPT, please do not perpetuate toxic academia.

ChatGPT: 6:00-7:00 PM: Take a break for dinner, maybe cook or order in with friends or colleagues.

JG: Yes. Dinner is great. Ordering in though… it’s too pricey these days for most grad students with our amount of stipend. Why not consider the student food market every Thursday? It’s a great way to get food.

ChatGPT: 7:00-9:00 PM: Continue with computational analysis, write up results or prepare figures and tables for publication, or work on other academic tasks such as reviewing literature or writing grant proposals.

JG: Ha. ChatGPT knows. Every good academic also uses their evenings to further admin tasks and other analyses that might not have been done for the day. How else do you get that sweet edge over your competitors? Just kidding.

While I have worked evenings sometimes, I usually spend mine reading web comics and watching cat videos. Sometimes, I write nonsense for Synapse. Like today. One day if I become a PI, I might do that. But for now, please let me watch my cat videos so I can get inspired to write more nonsense.

ChatGPT: 9:00-11:00 PM: Wrap up the day's work, plan tasks for the next day, and relax with a book, TV show, or hobby.

JG: If I do work, usually it doesn't stop at 9pm. More like I end at 11pm and sleep. But always remember to finish your Duolingo daily practice for the day.

ChatGPT: 11:00 PM: Get some sleep and rest for the next day's work.

JG: Wake up, cry and repeat. And then one day, when you have aged sufficiently, burnt through a good part of your 20s and have back pain and poor eyesight, you will find that you are no longer pleased with surviving on lab coffee, Cheeze Its and Welch’s fruit snacks.

That will be a glorious day when you finally ask your PI when you can graduate (though probably you had to ask this question like 5 times, and also to your thesis committee). And when you do, (with the agreement of your PI) you find yourself lost, with a weird feeling that this strange life that you have gotten so used to now is ending in a few months.

ChatGPT: This is just one possible schedule for a 50/50 computational wet lab PhD student, and the actual routine may vary depending on the research project, course requirements, and individual preferences. However, it is clear that a significant portion of the day will be spent in the laboratory, performing experiments, while another significant portion will be spent on computational analysis, data interpretation and planning for future experiments.

JG: Overall not too bad. While some things are not accurate, I think ChatGPT managed to capture quite a bit of what graduate students in STEM tend to do during a PhD. At least that has been similar to my experiences so far.

ChatGPT however, has forgotten to include other things like a secret place for crying, randomly walking around campus for free food and aimlessly spinning around in your chair at a loss for what to do.

Thanks to ChatGPT, I can now refer people who ask me what my daily life is like to this article.