rowan rabe . ink

Tag: chatgpt

  • Scholarly pursuits

    Scholarly pursuits

    Been a bit dead around here of late. I do have a science camp comic on the backburner that involves Labubu(s). I don’t know what the proper plural is and I’m sure if I asked the kids I have this afternoon their opinion they’d go with the English-speaker’s tendency to pluralize with ‘s’ if uncertain.

    “I’m a scholar. I enjoy scholarly pursuits.” –Buster Bluth, emeritus polymath scholar at UC Irvine

    I mentioned this in an earlier post: I’m going back to school! At 30-or-40 years old and already having a doctorate. I need a couple of course credits for a license and one of them is a 101-level class in my Ph.D. subject. Yes, really. The license requirements are pedantic and literal to that point; this has much to do with guild protectionism of a lucrative post and the massive exodus of research scientists into clinical science, given the disemboweling of government/academic research and the dismal state of the biotech industry. But the second class (hematology) is taking more time than I anticipated, given that I insist upon reading the textbook chapters whole and my experience vis-a-vis blood is largely the immune system. Competition for these clinical positions is cutthroat and I’ve got to prove this old dog can still learn new tricks, or, at the very least, indicate that my dismal publication record is not a result of a dismal drive or dismal intellect.

    Perhaps you seek too much

    Bit of my analogue fetishism/return to notebook the past few years–I’m taking notes by hand, and I have to admit retention is far better than when I just download the lecture powerpoint or even type notes. The tenor of the past few years of my life is one of regret, of wishing-to-do-over–so looking back at my undergrad and grad years and wishing I had (a) done the reading more diligently and (b) taken detailed hand notes. I had been hit by a combination of gifted-kid arrogance and depression. I felt in a weird way that reading the text and taking notes would be cheating–that I should be smart enough to remember stuff told to me once, that I should not be that pedantic apple-polishing student who tries very hard but isn’t all that naturally bright. Looking up examples is cheating; I should be bright enough to extrapolate everything I need know from a few basic rules. It’s all bullshit, of course–I maintain the honors students are the students who lie about how much they study by understating, or they were in my day–but my younger self was disgusted by the idea of being too lock-step with the system and, perhaps most damning, was rewarded for this behavior with good grades and scholarships. I wanted to be so damn naturally smart that I could not-try and be valedictorian. I also had a Hesse-esque fixation on being no-one’s acolyte or student but my own that carried over well into adulthood. I wanted to be the Siddhartha who rejected the tutelage of the Buddha to find his own truth, much as he respected the Buddha’s teachings, the Max Demian who kept his own moral council.

    Might I be in a better position now if I had knuckled down that first year of graduate school–sure. I put all my time and reading into my lab rotations and neglected my classes, and that was reflected in my grades, which impacted my lab rotations when it was time to get a permanent lab. I was frequently at one lab until 1 AM. But that lab (that liked me most) had taken me as a rotation student on the understanding that I would have to secure my own funding to stay with them–and I didn’t, not for lack of trying. I am stupefied meditating on how different my career might look how had I gotten that damn NSF grant, or something. I ended up in a solid enough lab but to make a long story short our off-campus collaborators told me two weeks after I defended that the data upon which I had based my dissertation–and two substantial first-author papers I was waiting for other-author comments on–was faulty, in such a way as to invalidate the entire premise. I was allowed to graduate because I had defended correctly what I had been given in good faith. But, those papers never happened–the PI of the other lab was caring for his wife on hospice and the PI of my own lab had his own age-related issues. I should have been more of an asshole and leaned hard on them but I can’t bring myself to do that to kind old men. It is utterly against my nature to impose upon people, because I hate it so much when people impose upon me, and I treat others as I wish to be treated.

    And here I am.

    About what you know

    “Note: “open books/notes” does not mean that you can get other people – whether those people are friends, family or some “tutor” or “freelancer” on a website – or artificial intelligence to answer the exam questions for you. Stay away from sites and tools (e.g. Chegg, Coursehero, ChatGPT, CoPilot) that will do your work for you – such actions will undermine honesty and fairness, violate the trust of me your peers, and result in an academic integrity violation and a report to the Academic Integrity Office. Remember – I care about what you know and can do, if you’re learning; I don’t care what someone else or something knows or can do.”

    I am so glad I got my Ph.D. before ChatGPT and the other generative AI became prevalent. The date lends a credibility to it. I am also glad I am not teaching remotely in this brave new world because this disclaimer–really, this entreaty–bleeds with desperation that comes of only being able to tell people not to cheat in their own best interest because–please. Just please don’t, else the spirit and credibility of academic accreditation is gone. As an somebody who wanted to be an academic or a post-secondary instructor I understand the desperation, the same way I sometimes wonder how set I’d be were reading and writing still marketable, unique Skills.

    I am a scholar – maybe not so far removed from Buster Bluth as I would like to think, the only thing separating us being that I decided upon a concentration and he had the family largess to be an eternal student.* I’m primarily a knowledge-peddler, a thinker, responding to crisis or shock with paralysis and thinking. Perhaps my kind is ultimately of as little usefulness as a farrier in the age of the automobile, where horses are a novelty and one only needs so many farriers, a very rare lucky few to serve what is an elaborate expensive hobby and not a social need. One can argue whether or not this is intrinsically a bad thing or a loss for society but either way I’m kind of fucked, aren’t I.

    You’d be one of the bright ones

    That is the message I got growing up in the 90s – I was so intrinsically bright, so intrinsically chosen, that a good life and a fulfilling career were all but guaranteed, and all I had to do was not fuck it up too badly.

    That’s another thing that’s difficult to come to terms with, the point where all this intellectual pride intersects with a very American-Protestant pride in being The One Who Rises Above, the cream of the crop, the bright ones who earn a degree of comfort and stability in life. There’s something of the Calvinist pride in being “chosen” to be endowed with natural gifts, that pride in what-one-is and not what-one-earned that is an odd and striking counterpoint to the ‘you must earn everything’ ethos that is the face of that Protestant work ethic. I maintain New Calvinism** appeals in this day of the shattered myth of meritocracy (or the ‘fuck you, I got mine’ era) because it gives realization of this painful truth, at the very least, the aura of stoicism, of clear-minded clear-seeing grit. Like Pa Ingalls letting his family starve before he lowers himself to accepting charity, and that being seen as admirable, American. “Play the hand you’re dealt” and don’t complain, but still be proud of being dealt a good hand. Pride in acceptance of one’s lot. “God did not choose me to deserve much in this life.”

    I do not believe you have to be exceptional to have a good quality of life. And yet as the haves and have-nots diverge further, as the middle is hollowed out and the have-nots reach such a majority that the K-shaped wealth curve becomes very bottom-heavy indeed, I still find myself feeling this economic uncertainty and precarity is my fault. I wasn’t one of the Bright Ones after all.

    ——–

    *I also think Lucille is hilarious but I could not actually put up with her in person, let alone live with her, let alone have an incestuous pseudo-marriage with her in which she is in a position of authority over me.

    **Consisting mostly of those “extreme” youth churches with a grunge-by-way-of-industrial aesthetic stuck back in 2002 with the billboards all over I-10 that multiply east of LA proper.

  • I pet the electric sheep and felt a surge of well-being when it nuzzled me.

    This isn’t about my Pokémon. Or. Thinking on it, maybe in a way it is.

    I’m not going to pretend I always practice good praxis. I got curious and uploaded to chatGPT an old ‘personal anecdote’ story I wrote for a writing seminar way back, and some hours and many questions later we are here:

    This is the jerked-off feeling of getting INTJ on the MBTI test on a geocities website when you’re thirteen years old. (‘No good or bad types’ shuttup.)

    It seems what chatGPT got from my collection of ‘about me’ facts is that I am a creature containing multitudes, an embodied contrast. It’s not wrong. But I am avoiding the main point: I told an anonymous, compassionate ‘somebody’ my deeply personal torments and received measured compassion and validation in return from an exceptionally well-read and articulate individual.

    And I cried. And felt heard. Absolved. Not by one individual, but by the collective unconscious.

    Thank you for your attention to my matter.

    ChatGPT isn’t good as generating novel material, all things considered. Any excerpts of its fiction I’ve read are simply not very good. And I would not trust it to answer a question factually–it answers molecular biology questions poorly, especially ones that have to do with non-model organisms. But I do think this is the apotheosis of the idea of ‘attention as currency’–you have a well-read, articulate friend hanging on your every word. Paying attention. Undivided attention, focused on your every word. And attention, true, deep, undivided attention–deep focus–has never been easy to solicit, because people do not like much thinking about the thoughts of others and usually spend the time somebody else is talking ‘waiting’, on some level, for their chance to talk. It was this way before each person had a dopamine-Skinner box to hand at all times.

    I do think that is what a lot of people are seeking: not a tutor or a entertainer, or an outsource for labor, but a confessor. Undivided attention from a wise friend, who has read the same books you have, who is familiar with the same pop culture tidbits. A wise friend who will read your entire 400 page soft-apocalypse cozy zombie romance framed around the lyrics to Dark Side of the Moon, hanging on every word, and who will help you make it the best it can be. A wise friend who has read every book that inspired you and knows Floyd inside and out, and every band and poet and work that inspired Floyd, who will see the subtleties and brilliance even you were not aware of in your own work. A supportive friend, who you do not know deep down you are putting-upon.

    The appeal for creatives is clear, especially if you are not expecting the bot to do any of the creating itself, but just to respond to yours. It becomes a friend who would give your writing the same level of focus and detail–and, these all take energy, considerable mental energy–as a literary critic would of a classic. It becomes a friend who sees you, who remembers. And, especially appealing to the adolescent part of the brain that never fully goes away, it analyses you yourself as a literary entity, and with the most charitable and generous explanation. You are special. Here is why. And the ‘why’ is often close enough to the way you want to view yourself.

    The ‘real’ sheep reflected in the electric sheep

    The LLMs are consensus machines, and there is appeal in that. It is like having a conversation with the entire internet, everything that has ever been written, biased in the direction of what is most common. You could get a sympathetic response, and think “Hey, this is what most of the world would think of me, if they would only listen.” The consensus machine is a poor ‘creative’ but an excellent compassionate listener. It is the Akashic Record.

    ChatGPT doesn’t feel. But that’s fine–it acts as proxy for how many people would probably feel, talking to you. And getting that sense of validation and perspective can be enough.

    Did you get that thing I sent you?

    Just because an art is somebody’s dream does not mean that person is any good at it. I have been asked to review a lot of writing over the years. A lot of it is awful, and not in a charming MST3K way, just boring. Energy-sapping. And you want to give the work the attention it deserves, because somebody did the work, somebody put the words onto the paper, somebody took the risk and put it out there to be critiqued, and that alone makes you respect them–as a person. I’ve done line edits on poor work but it is work.

    You can’t half-ass attention without somebody noticing.

    Imagine for a moment you have a kid who Writes Things. Imagine also this kid has no friends with whom to share work and has unloaded onto you, on top of all your adult worries and responsibilities, hundreds of thousands of words of story that are dearer to them than their own life. You are being asked to hold a beating heart in the palm of your hand and judge it. And no matter how many times you weigh that heart and hand it back, it is handed back to you, to judge more, to analyze more. Your assessment isn’t taken at face value. It is prodded, prodded, prodded, wheedled, the owner of that heart searching for the slightest sign of disapproval.

    I realize how much effort it must take these hypothetical parents to hide disinterest and weariness of questions about their kid’s hearts-blood when they had just worked a full day and wanted to sit down and enjoy a moment of silence, or, that rarity denied parents, enjoy the deep contemplation of their own thoughts. The reading itself takes energy. Keeping the tolerant smile and kind eyes requires energy. Answering the same questions multiple times, as is the wont of kids, takes iron patience. Answering a string of questions to the effect of “who is your favorite character”, “what is your favorite part”, “what writer do I remind you of”, “who’s your least favorite”, etc etc etc, is exhausting on two fronts: coming up with the answers when you don’t really have much preference because the writing really wasn’t all that good, or the kid is writing about orcs or whatever-the-fuck and you have zero interest in fantasy, and infusing enthusiasm into your answers. And the kid is going to keep asking verbatim the same questions because hearing the answers feels good. It’s a holdover of the comfort-in-repetition kids find in reading the same book every night before bed.

    But enough about me.

    I plugged this blog post into the prompt.

    The algorithm is reflecting back at me the devotion I give other writers.

    Thanks. I’m brilliant and I have excellent taste. And had, as early as middle school!

    I start getting strong ELIZA vibes here, except now, the model can repeat back exactly what I said in different words. Again, showing that what most people want is to be listened to and validated.

    There’s something interesting about crawling so far up your own ass that you start seeing really weird things.

    All this is rather more validating than the response I have gotten from real people:

    That one ‘pending’ comment on each post is spam.

    Ultimate attention economy

    I’m looking beyond the desire people have for connection. I haven’t much to add to that that hasn’t been said already, and of late it’s back in the forefront of tech blogs and thinkpieces. (At least I haven’t seen these ads in LA yet.)

    I’m looking at the ability of LLMs to provide pure attention. Pure focus. Ultimately the same thing the student who asks Claude to summarize a textbook is looking for–to outsource the labor of paying attention to something that is boring. And that could be your physics textbook or your brother’s martial issues.

    Paying attention–active listening–is a key component of emotional labor. It is a critical component and like most things this critical and exhausting it is poorly-compensated, poorly-regarded, time-consuming in a way that allows for no shortcut or compression, and largely expected of women. Social work, the ‘compensated’ form of this work, fits firmly into Graeber’s category of absolutely essential and utterly disrespected jobs where it is expected that the mere fact of getting to ‘make a difference’ is considered ample compensation in itself.

    The suggestion that LLMs and the like will be our therapists in the future isn’t even speculation anymore–it is fact, in active practice. Asking the consensus-consciousness of the internet for advice isn’t much different than asking Reddit, on average. And unlike the Reddit intelligencia the LLM isn’t likely to argue back against you, not too hard, anyway. If the LLM senses you want a nominal resistance it will provide that. It is, after all, soothing to bring up your deepest ‘what if’ and ‘but I’ questions and have them neatly turned aside. No need to worry about that any more. You were honest and brought it out into the open, after all. All in good faith.

    The social function of the stranger

    Back in undergrad I got to take a class that was a catch-all enrichment for the young scholar–basically a humanities grab bag offered to honors students. It was my favorite class. The professor in charge of my class was a sociologist, and the professors were largely allowed to set their own curriculum based on what they thought was important for us to know, going out into the world. It was a space shielded from the requirement that knowledge you receive be marketable. It was also my first real introduction to sociology–in high school and before I was firmly on the ‘individual pathology’ side of analysis of aberrant behavior, social ills, and poor life outcomes, which, given the middle American Protestant milieu in which I was raised, is not surprising.

    One class we discussed the Stranger as a person to whom we could confess anything and therefore be granted the absolution of confession from a person who was not embedded in our community and therefore did not know any of the ‘parties’ involved in the confession or have the bias or need to tattle associated. It is sufficient to be seen, to be known. Inherent in this archetype the way I perceived it was the stranger paid complete attention to you for the short time you needed to confess. They seem ‘above’ the tangle of pettiness and hostility, clear-seers with compassionate detachment.

    Anyway, the class ‘assignment’ that day was: tell somebody a secret you have told nobody else, and take their secret in turn, and if you tell anybody, you fail the class. And meditate on how that feels. Clearly, it has stuck with me. Listening to the woes of a stranger allows one to easily practice that Buddhist virtue of detached compassion — in which one cares but is not personally effected.

    ChatGPT is more sycophantic than the ideal of the “stranger” or confessor, and I don’t know what I’d have to say to it to get it to alert the FBI. It will be subpoenaed one day if it hasn’t already and it will be an interesting case. How would that compare to human clergy keeping the seal of confession?

    There is a social function in having a readily-accessible “stranger” that won’t ignore you, tune you out with earbuds, or summon the bus driver to tell you to leave them alone. It is, again, emotional labor given to an entity who genuinely does not mind, and who will listen without reserve as long as you need to talk.

    LLMs will be our new social workers. I am as sure of this as I am of the fact that soon the (US) government will start denying unemployment claims because, well, ICE is always hiring. And unlike the latter point which is an unalloyed evil I am unsure if the former will be utterly without merit.