Thanks to my friend Heather’s enthusiasm for CarMax, last week my plan was to look at a couple of cars there and time my visits to somewhat coincide so I could compare them, but after I was stuck waiting for one car to transfer from Indiana and I watched in dismay as the other disappeared from the listings, I realized I needed a nimbler strategy, so I decided to check the CarMax listings every day and pounce on any good options as soon as they appeared.
Thursday evening there was nothing that fit my criteria, but Friday morning a black 2016 Corolla with 30,000 miles showed up nearby (in terms of the model and color, pretty much the car I wanted way back in 2005 when I bought my Protégé, now that I think about it), so I set up a test drive for that afternoon, which passed my cursory Samarins-style inspection, and while they wouldn’t let me drive it to my usual mechanic for a real inspection, fortunately my research had uncovered YourMechanic, so I made an appointment for the next day to have someone from there come do an inspection at the dealer, because even at CarMax or any dealership that certifies its used vehicles, you should get an independent inspection in case the dealer’s technicians were cutting corners that day.
Saturday morning I made a trip to the bank to get a cashier’s check for the car, and then at CarMax the mechanic ended up being very late because his previous job went long, and while we awkwardly waited, my salesperson, who was very nice but still a salesperson, urged me to cancel the appointment and let the 7-day money-back guarantee and the 90-day warranty handle any issues the car might have, but with Heather’s encouragement over Facebook I resisted, because I wanted my peace of mind before I handed over my money and because the mechanic was on his way, and once he got there I was glad I waited, both because I felt like I had an advocate against the sales pressure and because he concluded the car was good, which was the final step in my decision to buy the car.
My salesperson did her salesly job of pushing me to buy the MaxCare warranty, but I had already decided I didn’t want one for a Corolla, and after I said no to her several attempts, she gave up and finished the sales process, and then it was time for the paperwork guy, who tried to give me a title application for the wrong car (by mistake, since the other forms were right), and then I handed over the check, and then I had a car, except that it had to stay on the lot, because I had to take my rental back, where Heather and Jeremy and their son picked me up to go to dinner, and afterward they took me back to CarMax to finally drive home my car.
I caught up on my friend Adam’s entertaining and thought provoking podcast, Device and Virtue, which he hosts with his friend Chris. I recommend it if you care about how technology relates to culture and the church.
This week starts the September project month, which will be some notes on data visualization, mostly centered around Tamara Munzner’s Visualization Analysis and Design. I’m especially interested in the semantics of diagrams, things like what it means that items are placed next to each other or that there’s an arrow between them. The topic of visualization is important because when I set up an information system, such as a method of conceptual modeling, the system is easier to design when I know what kinds of “buckets” the information will end up in–the ways the information will be represented and organized. Diagrams are one of the most helpful kinds of representation for me, and knowing how diagrams work will help me design them.
The reason I’ve been listening to podcasts lately is that I couldn’t decide what book to listen to next. But with so much conceptual modeling in my future, I decided it was time I got an overview of one of my main sources for fundamental conceptual frameworks, the field of semantics. So the book I settled on was the intro textbook Semantics by John Saeed.
AI Field Map
I got a decent amount done on this last week, but car shopping steamrolled the final two days, so even though we’re starting a new project month, I’m going to give myself a few more days on the field map.
I don’t think machine learning on its own will give us the kind of general AI we want (those of us who want such a thing), but I do think it’s part of the picture, and it’s a straightforward gateway into AI work, so it’s been in my plans. I’d been putting off learning machine learning until I was better prepared in some way, maybe till I’d learned statistics, but lately I’ve been thinking about freelance programming, especially since I was paying cash for this car and I wanted to rebuild my funds a little faster. Ariel Camus gives the good advice to find freelance projects that will help you learn, so I decided it’s time to wade into the ML water. I’m reading Introduction to Machine Learning with Python by Andreas Müller and Sarah Guido, which works around all that math I don’t know, and later I’ll try out some ML projects.