Hype, Horror, or Hope?
Should scientists be excited about generative AI’s potential utility as a digital assistant – troubleshooting problems, curating research, and saving time? But what will the AI revolution mean for educators when ChatGPT can pass a chemistry exam?
| 4 min read | Technology
AI Meets AI
Upholding academic integrity (AI) in a world of artificial intelligence (AI)
By Alan Doucette
“I’ve always stood by the importance of academic integrity. I believe every student needs a fair and equitable opportunity to demonstrate their success. I also believe every student has the potential to succeed, and that hard work is the key to that success. With that in mind, it’s a silly exercise to ban AI tools from being used in education. AI is an invaluable educational tool, just like the internet before it, or books before that – did you know that Socrates felt writing would train the mind to forget?. What parameters do we establish to define how much “help” is acceptable with AI tools? How do we know if students have passed those boundaries? And how do we encourage our students to uphold these limitations? I don’t have these answers. I just know that the education system is changing at such a pace that no one person can keep up.”
Read the full article
Your Students Are Already Using ChatGPT
Can you even tell? If so, please let me know how…
Christopher R. Harrison
“This past fall, I was having a discussion with my fiancée about ChatGPT and the ramifications it could have on teaching. She’s the Chair of an English department, so the impact of the AI on writing assignments were obvious and concerning to her. Smugly, I claimed, “We in chemistry don’t have that problem.” After all, how could an AI that is good at spitting out summaries of classic novels or personal statements be of use to students trying to calculate the exact amount of acid to add to a solution to get it to buffer at the correct pH? Oh, how wrong I was!
“Later that evening, as I thought more about it, I decided to test ChatGPT with one of the simpler “Calculate the pH of a solution of X” problems. What didn’t surprise me was that the AI failed spectacularly at getting the correct answer. What did surprise me was how it failed. It solved the Henderson-Hasselbach for the pOH using the pKb for acid, despite my request for pH. But then it dawned on me. A group of students in my class had used that approach several times to solve similar homework problems, despite me never teaching it to them.”
Read the full article
Talking AI with a Big Tech Veteran
Why machine learning expert and former Metaverse and Apple AirPod engineer, Lalin Theverapperuma, thinks artificial intelligence-enabled automation will transform life in the analytical lab – allowing analytical scientists to eliminate mundane tasks and focus on more creative work
“A model like ChatGPT is trained with five petabytes of data over weeks – which costs millions. This is a very generic model for general purposes. In the analytical lab, you’re dealing with much smaller and specialized sets of data. In this world, you need precision. For ChatGPT, 90 percent accuracy might be acceptable, but for an analytical science AI tool, you want 99-plus percent accuracy. So that’s a big differentiator – and a challenge. Our idea was to create an AI architecture that allows analytical scientists to develop AI-powered methods tailored to their specific data environment. For example, we’ve helped customers make functional models with around 30–50 samples-worth of raw data; of course, if you have 1,000 samples, the accuracy will be much higher, but there’s a lot AI can do even with smaller data sizes.”
Read the full article
ChatGPT: the Chromatographer’s New Best Friend?
How does AI in the form of the large language model stack up as an educational resource, troubleshooting companion, and research curator?
With Tony Taylor
“In some areas, I have been really surprised by the usefulness of LLMs, especially in the areas of improving fundamental understanding and troubleshooting separations via the ability to upload problematic chromatograms and derive ranked pointers for possible solutions. I believe it is reasonably widely understood that the popular LLMs, such as ChatGPT-4, have training datasets that are inadequate to be fundamentally impactful in terms of domain expertise or insight. The models do not currently contain or interface with ‘expert’ computational tools; however, as I’ve described above, this may well be changing with initiatives such as ChemCrow.”
Read the full article
