Month: April 2023

For those who are familiar with behavioral psychology, the concept of sunk costs and the eponymous fallacy is not new information. Sunk costs cause people to value what they have, and more specifically what they have sacrificed for, more dearly than what they don’t have. This can create inefficiencies in markets but those are usually marginal enough that they don’t have a major impact on a person’s life. What can have an impact on a person’s life is allowing the sunk cost fallacy to keep them in a career they don’t like out of a desire not to have “wasted” the years spent obtaining the credentials necessary to pursue that career. Or staying in a bad relationship (though neither of those things is applicable to me at this point). Nothing that transpired in the past can be changed, but sunk costs keep us from moving forward anyway.

I’ve been plagued by sunk costs this year—I have lingered as a result of money spent on housing elsewhere. Sure that money was already spent and there has been no recourse available to recover even a single penny, but I’ve hesitated to spend yet more money on housing. Since money is fungible, this is irrational. I knew this all along, but that didn’t change my behavior: that is the sunk cost fallacy in action. At least now I am doing so out of agentic choice as I have trips planned for the next few weekends for which I am more-or-less centrally located. Then after that it’s time to get moving. I don’t have a fixed plan of where I’m going yet, but even contemplating the options has been an enjoyable exercise over the past few days. Whatever comes, there should be great variety in content over the next few months. Also, let this be a reminder that just because you have poured a lot into something doesn’t mean you have to keep doing that thing. Time only moves forward and so should we.

Back in January I wrote a post entitled “Double-Barrel Information Intake” about an experiment I was about to conduct in which I would simultaneously read and listen to a book. After about two months of going through the exercise, I can now report the results. I will not be adopting this approach to my nonfiction reading moving forward. In fact, I don’t think I’ll be using audio versions of nonfiction books at all in the future as I haven’t gotten the results I wanted to obtain. That isn’t to say that I won’t listen to a work of fiction ever again. It’s just that right now I don’t have a commute and making my commute better was the function of those books.

My conclusion about listening to nonfiction books was echoed by some findings that I took in (somewhat ironically) through Cal Newport’s podcast describing how there is magic in the time spent between words and sentences that allows for the absorption of material and forming synaptic connections in the brain. When I tried to click through the link in the episode description to the study he referenced, the site wouldn’t open in my browser, but here is a link anyway: LINK. There may be broader evolutionary implications of those findings given the shift away from prolonged and focused reading in favor of ever-shorter bursts of dopamine in short-form video applications but this post isn’t a polemic; it’s just me reporting my own experience and being honest about something that I tried and will not continue.

In the last couple of weeks I’ve been reading four different books, each at a more-or-less consistent time and for a different purpose. The early returns on this experiment are positive. Maybe I’ll write another post about this approach in a few months.

My birthday was this week. It was a subdued affair and a solo dinner out. I did at least I get myself an ice cream cone for dessert, vanilla lest anyone have doubts. I’ve never been a fan of big birthday parties anyway. As with many other milestones I took a little time to reflect on another circumnavigation of the sun, this time sitting on a bench overlooking the ocean. It’s now 10 years since I graduated from college and 5 during which the rate of change in my life has accelerated. I’ve had five more-or-less permanent addresses during that time. The only place I’ve woken up more than once on my birthday during that span is where I am now in what has become my in-between home. And every indication is that the trend will continue over the next twelve months. Where I am isn’t at all where 22 year old me thought I’d be when I left Wofford for a stint in Boston. That disconnect doesn’t look like it’s going to change in the year to come either and I’m glad that that is the case. I tried a version of the dreams 22 year old me dreamed and as it turned out my vision missed the mark. It may have been directionally good but many of the particulars were off. I doubt I’ve got all of the details sorted in my current vision of my future either, but I’m more directionally accurate than I was 5 or 10 years ago. At least I hope so.

I was having a conversation this week about ChatGPT, AI in general, and what the technology might mean for the future. I’ve continued to experiment with the ChatGPT interface and have found several resources that have helped me write prompts that get more out of the tool. In the short term, I’m now convinced that the people who learn how to interact with these new AI tools will see massive productivity increases and rapidly outcompete others, at least in the domains in which the AI tools function.

During the conversation, a thought came to be in the form of a connection to another idea I heard on a podcast a few weeks prior. I haven’t spent hours and hours turning over this thought, but that is one of the purposes of the act of writing. The thought is this—these AI technologies are going to do to the digital world what the discovery of the neutron did to the physical world. The discovery of the neutron opened up the possibility of nearly unlimited energy by tapping into potential of which we had known nothing previously. It also led to nuclear weapons. These versions of AI systems are surely already being weaponized (after all, the real breakthrough of ChatGPT was to make the outputs more human-like; there are already more advanced AI systems that have been created), but the effects will only compound as future iterations are created and released. Sure, nearly every new technology comes with positives and negatives and metaphors have their limits, but this one does have a Pandora’s box sort of feeling to it. Or maybe more of a 1984 feeling to it. And if that inkling is correct, then it is already too late to reclose this particular Pandora’s box. It probably isn’t too late to course correct, but my knowledge of how these systems work is too limited to allow even a half-educated guess as to how that might occur.