Sabbatical Diary #1

Since my decision to eschew keeping a timelog while on sabbatical, I’m wondering how I can be reflective about my “work” in a qualitative (rather than quantitative) way. One possibility is by writing a periodic blog post; I’m unimaginatively naming the series “Sabbatical Diary”. So what did I do last week related to my role in academia as a teacher-scholar?

 

I’ve started to learn Category Theory. When I stumbled across theoretical biologist Robert Rosen’s work about how to abstract what makes a system “alive”, I had significant trouble following the math. I deemed it to be Set theory because I recognized the symbols used when I learned rudiments of set theory way back in secondary school. I now recognize that Sets are simply one type of entity in the broader Category Theory. I’m returning to this because a mathematician friend has come up with a potential way to quantify emergent systems and I want to figure out if this abstract math can be operationalized to study chemical systems of interest to which I can provide microscopic data. Being on sabbatical, I have both the time and motivation to tackle this for now; I might give up if it does not look promising. One has to know when to cut one’s losses and move on.

 

In the mornings, I’ve been working my way through a textbook on Category Theory (by Spivak). I have to read very slowly soaking in the definitions and translating in my mind what the abstract symbols mean. These were relatively straightforward in the beginning, but by Friday morning, I was starting to run into trouble when I encountered homomorphism sets. I had a vague notion of what these were, but kept moving forward. Then I encountered pullbacks (fibre products) and was stymied. I caved and called up a genAI chatbot to give me concrete examples and some intuitive notions of how to think about these abstractions.

 

My use of genAI chatbots thus far has mostly been in chemistry, my field of expertise. I’ve been exploring its capabilities (which are improving as they evolve) as an aid in teaching and research. Most of my “experimenting” is poking around how it might help student learn chemistry – how it might be useful and how it might mislead. My working hypothesis is that if you have an expertise in an area, you can effectively leverage genAI’s capabilities to automate or synthesize tasks at hand; you also have the ability to quickly extract genAI’s ideation capabilities pruning out the useful nuggets and discarding the chaff. Conversely, the novice superficially interacts with genAI to get “surface” knowledge and can’t tell when or if they are being bamboozled. Worse, I suspect using genAI gives my students the illusion of knowledge – which is shattered when some of them take an in-class closed-book exam.

 

This morning I needed more genAI help as I started trying to understand monoids. Asking genAI to give me practical examples in chemistry is helping. That being said, I can see that part of my challenge is that many of the definitions in Set theory that I worked through last week haven’t really sunk into my long-term memory. I know these only superficially rather than deeply. It’s also why I’m having trouble moving forward. I laughed at myself as I recall the many times I’ve told students they need to memorize definitions because these are the building blocks of the terminology we will use in chemistry. If you don’t have these at your fingertips, you will stumble a lot like a blind person in the dark as we advance into subsequent concepts that build on these definitions. We’ll see if I can practice what I preach. I’m impatient and want to quickly get to the point to see whether this background will actually be useful in my research. (I am skimming over some parts in my reading and not working through the harder exercises. Just like some of my students.)

 

Two hours per day is about what I can handle on Category Theory before my brain feels fried. I also need time to digest some of the information I’m learning and let my brain do some consolidation when I’m asleep. The rest of my mornings last week were spent reading lighter material, usually but not exclusively education-related. I’ve also resumed writing more blog posts this month; I think I’ve gotten to the point where I couldn’t care less that armies of bots are scraping my writing to train commercial genAI models. I think writing regularly (without genAI help) is useful to me as a skill and in clarifying my thoughts.

 

My afternoons last week were spent on trying to tie up the many loose ends on a research project. I want to write up a paper that incorporates the work of one of my research students who will be applying to grad school next year, so it will be nice if she can have a peer-reviewed paper in the bag on her CV when she is applying. When attempting to consolidate the work into a coherent and publishable story, one always finds that certain threads need to be nailed down, in my case by doing further computational experiments. So I’ve been setting up jobs, running them, and doing data analysis. I feel I’m getting closer to a coherent story, and the question is when enough is enough. Research is inexhaustible. You pull on one thread and it leads to three more.

 

Okay, that’s it for Diary entry #1. I don’t know if there will be more. I do know that I have many more thoughts swimming in my head.

Joint Adventures

The cells in our bodies are constantly being replaced naturally. Like the Ship of Theseus, am I still me? I feel like it’s the same me but with inevitable deteriorating physical capabilities as I age. I hope I retain my mental acuity, but there may come a point where I don’t recognize others or even myself. We consider aging a natural course of events although we don’t know exactly why our bodies have a clock that winds down towards eventual death. But thanks to doctors, scientists, engineers and inventors, there are artificial replacements for the wear-and-tear.

 

Which body parts can be replaced and what can they be replaced with? This is the question prompting Mary Roach’s latest book, Replaceable You. I’ve enjoyed several other books by Roach, who deftly combines humor, fearlessness, all while she teaches you some very interesting facts about the limitations of being humans and possibly how to get around them. She also somehow gains entry into surprising places and manages to get people to divulge interesting information that is unexpected. With words, she also capably paints a picture of the sights, sounds, and smells of wherever she happens to be.

 

Today’s blog post is on Chapter 8, “Joint Ventures”, subtitled “woodworking without wood”. As a reader, I feel I’m transported by Mary’s Adventures. To give you a taste of her writing, here’s how the chapter begins: “The third hip replacement of the morning looks very much like the second and first. The patient and the whole operating table are covered with surgical drapes, resembling not so much a person having surgery as a small vehicle under a tarp. A surgeon stands alongside, holding a metal instrument in a hole in the patient’s side. The hole – the incision – is held open by a circular plastic retractor the size of an automobile gas cap. From where I stand, six feet back, this is all I can see. Hip replacement has the visual drama of a visit to a Chevron station.”

 

The next paragraph begins in an arresting way: “It’s the sounds that undo you. The whine of the bone saw as the surgeon cuts…” I’ll stop describing here, but you can bet that Roach makes plenty of comparisons to a woodworking shop. But there’s a major difference as Roach goes on to describe: “A (wood) cabinet has no immune system. It doesn’t throw up defenses against building materials it perceives as hostile invaders. It doesn’t die under siege from bacteria that gained a foothold on a piece of inlaid metal or plastic. In other words, the surgeon’s skill can take you only so far. It’s the material guys you’re depending on for a lasting, complication-free build.”

 

My aging mother has had both hips replaced in the last three years. I’m glad for the significant pain reduction that has resulted, and the efficiency of modern medicine. I don’t recall exactly what materials were used, but Roach takes me through the history of such joint replacement starting in 1938 when stainless steel was used for both stem and socket cup. Titanium and other alloys eventually replaced this, but metal-on-metal wear and tear can result in debris that leads to inflammation. Ceramics (metal-oxides) can reduce the wear but their underlying brittleness can be a problem in a high-impact situation. Teflon was used at some point, but there were problems; compact polyethylene has proved better. And so it goes in the evolution of materials. The challenge, as Roach notes, is that “you can’t know for certain how a material will perform or react until you put it into a patient and watch what happens for five, even ten, years. Yet the time required by the FDA to establish the safety of a new medical devices is often shorter.” Worse, there’s a “minor changes” loophole that can avoid clinical trials.

 

Ivory turns out to work surprisingly well. We know this from a surgeon in Burma who managed to persuade a local ivory carver to fashion “knob-topped stems to push inside people’s bones, where no one would even see them. It was like trying to hire Georgia O’Keeffe to paint the janitor’s closet.” No wonder most of the artisans turned him down. The surgeon performed hundreds of successful surgeries with only a two percent failure rate, which is an astoundingly successful rate, given this was in the mid-twentieth century. Part of why is the low infection rate with ivory even without antibiotics. There’s a tricky balance at play: Ivory is very smooth providing fewer nooks and crevices for bacteria to invade, but modern materials are also intentionally made porous to encourage bone growth. You want the bone to grow in before bacteria can proliferate. I also learned that an exception comes from dental implants because saliva and possibly our gum tissue seem to have natural antibiotics because the mouth is literally a cesspool. The stringent “clean” practices in joint replacement surgery have evolved significantly to reduce infection rates, and Roach gives credit to those who painstakingly tested different protocols.

 

I haven’t named any of the people Roach discusses as she sets up her joint-replacement learning adventure. That’s because Roach has some very funny bits about the name coincidences; you can get your hands on her book and read them laughing out loud for your own enjoyment. Her acknowledgements section is also hilarious, where she thanks all these people who said “yes” to her invading their workspaces and pestering them with questions. You might also want to know about all those other body parts you might want to replace and where we are with the technology. Let’s just say I was surprised at the very wide range of stuff discussed in Replaceable You. You might be surprised too!

AI TLDR

Today I read an article in the Chronicle of Higher Education (CHE) titled “My Students Can’t Read”, one of many I have read in recent years about the inability of many college students today to sustain the attention and focus to read academic material at length. Many reasons have been proffered for this alarming state of affairs including smartphones and COVID lockdowns; some hark back to the rise of television; but the most recent culprit is Generative AI.

 

Unfortunately, CHE has a paywall, so instead I’ll point you to Scriptorium Philosophia’s most popular post titled “the average college student today” which I read a year ago. It treads on similar ground as the CHE article, but it’s shorter to read, broader in scope, and funnier. Or if that’s still TLDR*, here are the three quoted highlights from the CHE article.

·      Every generation of professors has complained that their students cannot read. The lament is usually overblown, but data have caught up to anecdote.

·      The neural pathways that support sustained attention are built by use, and they atrophy without it. Your body is a use-it-or-lose-it system, and the brain is no exception.

·      Offloading tasks to a chatbot does not “free students up for higher order work”. It deprives them of building up strength to do any substantial cognitive work at all.

 

In the original article, these three quotes are blown up in larger font in separate boxes to catch your attention. Sort of like an abstract or summary. While I suspect a human editor picked them out to highlight, maybe AI could do this for you. (No, I didn’t copy-paste the article into ChatGPT and ask it to highlight the main three points to verify.) If you’re a busy person, and college students seem busier than a generation ago, then you might use AI-mediated TLDR – read the AI generated summary. I have so far resisted the temptation to use AI to skim or summarize research papers, but maybe that’s just the old-school stuck-in-my-ways curmudgeon in me. I don’t quite trust AI to give me the nuggets I might miss, but then again, I do wade through lots of irrelevant stuff. I suppose it’s mildly comforting to know that I have built up the ability to quickly pick out what I need to know and skim the rest.

 

Two weeks ago, a rep from the publisher of our G-Chem textbook emailed faculty to let them know of a new feature. You can now have AI summarize key points in the eTextbook. This just about guarantees students won’t read the actual textbook. Why bother when you can AI TLDR instead. We are inherently lazy and would rather save our ATP molecules to be expended on more interesting and stimulating pursuits than reading the boring textbook. I’m not sure anyone’s thinking about whether intellectual muscle is being atrophied. Or worse, not built up at all. Even though I now provide Study Guides for every G-Chem class meeting, I’m very sure that many of my students just use generative AI to run through them. They were happy to tell me how they use AI. Many even felt it helped. I think for many (but not all) of them, it gave them the illusion of learning without the substance. This is likely partly the cause for the growing number of D’s earned – dismal exam performance.

 

Four weeks ago, I surveyed my G-Chem II class about textbook use since we had switched to a new textbook this past academic year. Once again, students were brutally honest about not using or hardly using the textbook, and they were the significant majority. So maybe the inbuilt AI would be an improvement? They might at least read the summaries? But the summaries they read are unlikely to stick, and will be forgotten very soon after. Long-form sustained attention while reading has a higher chance of some knowledge sticking, but only if the students have the staying power to stick with it for a while. All this might be the final death knell of the textbook. Publishers have already anticipated this; the online homework system is what you pay for (because of lazy professors?) and the eTextbook is along for the ride. This is unfortunate, because our present G-Chem textbook is quite good and, in my opinion, quite readable. But that’s coming from someone who grew up reading a lot without a TV at home and certainly pre-internet. (Unlike P-Chem where I ditched the textbook early.)

 

There is a potent larger danger to most of society no longer able to digest a longer, sustained, and more complex explanation or argument. Issues of import will devolve into memes and vibes turning democracy into idiocracy. I shudder at the thought that AI TLDR is taking over and mushing up solid intellectual food into baby paste, but worse, stripped of any underlying nutrition. How depressing.

 

*I first learned the acronym TLDR (Too Long Didn’t Read) over a dozen years ago from a short reply by a younger colleague to a long-form argument I had written in an email to a group working on some curriculum or administrative issue that I no longer remember. It was a rude awakening.

No Timelog Experiment

I arbitrarily chose today to mark the beginning of my sabbatical. In my mind, the most significant change I will be making is that I will NOT keep a timelog for the next twelve months. I had been keeping one since I started my tenure-track faculty position a long time ago; I even provided a ten-year snapshot with brief analysis in this blog.

 

Many faculty members maintain a flexible schedule through most of their careers. Sometimes work is done on evenings and weekends; sometimes one runs a non-work-related errand in the middle of a weekday. In that sense, their life is like that of a college student: Show up for classes and meetings, and do other work when you’re not in class or in a meeting, with more frenzied activity as a deadline approaches. Like some students, professors can also be procrastinators and constantly doing things at the last-minute. On top of that, there’s research. It never ends. There’s always something to read, something to think about, an experiment to run, and it constantly occupies one’s mind-space.

 

I’m an atypical professor with regard to time. From the beginning of my career, I made a clean division between work and non-work. I treated work as a 9-to-5 job (more like 7-to-4 in my actual case). When at work, I focused on work. When away from work, I did no work and did not check my work email on evenings and weekends. Studies suggest that most professors work significantly more than 40 hours per week; one that includes charts and breakdown of time usage claims 60 hours per week on average for tenure-line faculty although this is self-reported data. My average is 42 hours over more than two decades of data, and while my data is also self-reported, I meticulously logged my time use daily. But this also means, that when I’m at work, I have to be efficient to get things done. There’s no time for procrastination or distraction, and overall, this has made me a disciplined worker.

 

Keeping a timelog is very useful when you have a self-directed job with many different things you want to accomplish on top of teaching obligations and open-ended research projects. To make sure I was maintaining the balance I wanted, I regularly looked at how my time was being spent and make adjustments. This provides stricture and discipline, but I wonder if over the years it has reduced my flexibility and squelched my creativity. Perhaps I have even become risk-averse intellectually; honestly, I’m not sure. Hence, I am embarking on my no-timelog experiment. I will, over the next year, not place strictures on when I work and how I work, allow some of my work and non-work to mix, and observe what happens. My pitch to myself is that I am aiming for a wholistic sabbatical.

 

I did write a sabbatical proposal which was approved, so I do have some professional goals to accomplish. I have prior research I’d like to write up, new research projects I’d like to think about, classes I’m excited to overhaul, and new classes I’d like to design. Maybe I will learn a language; maybe I will do more traveling; maybe I will learn to use some new tech tools; and I will certainly do more reading! I would also like to be rested, refreshed and rejuvenated when I return to my regular daily job for the Fall 2027 semester. But one step at a time, and this morning’s blog post is just the first step on the first day of the next twelve months!