Throughout my career in science AI and automation have been things that were ever-improving tools in helping do science quicker and in bulk. AI and machine learning have slowly been improving the rate of data processing and even expanding the kind of data processing that’s possible. Automation has led to a more mechanised lab where something time-consuming like filling 100 well plates has been reduced to the press of a button.
And as tools that made research better, I have always been an early adopter of AI and Automation. There really was something pretty magical about being able to control my experiments from my phone or automate them to the point where I just collect the data at the end and get on with writing papers and designing new experiments.
But throughout my adoption, I very much considered AI and automation something that can augment me as a scientist not replace me. That is until I started reading about AI bots like ChatGPT.
ChatGPT is a chatbot made by a company called OpenAI. It is one of several out in the world that uses some very advanced reinforced learning to produce answers to questions. It burst into the science world when various institutions started realising that students were using ChatGPT to write whole research papers and the research papers were actually quite good.
While writing student papers is still a little way off ‘doing research’ it is certainly starting to get worryingly close. If it can write research papers can it write other research documentation? I mean lab-based research is about 90% following protocols and 10% creatively ignoring protocols. If ChatGPT can write protocols it is only 10% away from totally replacing me!
So the purpose of this article (and my future career in science) I thought I better check how the AI competition is shaping up and I asked ChatGPT for instructions to make Phosphate Buffered Saline (PBS). PBS is very common and there are a ton of simple recipes online ChatGPT could copy and this seemed like a simple one to ask.
In answer, ChatGPT spat out a 3 sentence explanation of what PBS is followed by 6 very clear steps for making saline then adding in the phosphates etc, and a short note at the end to ensure I use the correct purity of ingredients. I’m not going to lie I was impressed.
This wasn’t copied and pasted, this wasn’t just nicked from a forum or online calculator this was a well-written protocol that wouldn’t look out of place in Standard Operating Procedures or a lab book. But (luckily for me)… it was phosphate-buffered garbage.
First off, it uses distilled water, not deionised water. Distilled water still contains various random ions and impurities that will mess with experiments. I frown on things that mess with my experiments.
Secondly, the protocol missed out potassium chloride. Which seems like a biggy but I’ve seen lab techs miss this out too as it’s not ‘required’ for the buffer to err buffer. So semi-excusable but not really as I asked it for PBS, not lazy PBS.
Thirdly, you make too much. ChatGPT has you make a total buffer of 1000 ml and then added a few ml of pH balancing which dilutes all your carefully measured concentrations.
Fourth, the phosphate concentrations were wrong, about 4 times wrong. Call me a stickler but having the correct concentrations in a buffer seems fairly fundamental to me.
Now, these sound pretty obvious when written out but honestly, until I read the protocol very carefully at first glance I thought ChatGPT had generated a PBS protocol that looked pretty legit. And even more worryingly these are the kind of mistakes I would expect from a new trainee or student which is the equivalent of ChatGPTs beta status in a way.
For now, I think I’m reassured that my job is safe but not for long. If you go on ChatGPT and ask it for a PBS solution recipe I suspect it will get it right as in the process of writing this article I corrected its faults. ChatGPT is coming for us and fools like me writing this article are helping! Let’s hope slightly irreverent humour continues to elude ChatGPT so that I at least can keep this job.
ChatGPT give me a final sentence for a funny article about ChatGPT trying to write a PBS recipe.
“And that’s how ChatGPT learned that making a recipe for ‘PBS’ is a recipe for a chemistry experiment, not a tasty meal.”
I feel safe.
Sasha Egan · 2 February 2023 at 00:33
Hmmm…how does it handle buffer solutions outside normal temperature ranges? Let’s say for example, I wanted a pH 7 buffer solution of KCl/KI that was applicable for temperatures above 240C and at 50bar.