Up To A Point

Dearest Rachel –

One of the odd situations of my particular task at Sparks, as I record attendance and progress, is that I’m not exactly busy during the time everything else is going on, but rather before and after. We get to church early, since neither Daniel nor I have job commitments, and set the class room (for lack of a better term for it) up. Later on, while the kids are downstairs playing games, I bring their personal study materials to the room from the auditorium with Daniel, and place everyone’s stuff on their group’s various tables (although the items that don’t have names on them will give me trouble – some of the books had had their covers ripped off from wear, leading to this problem on occasion).

However, once everything is in full swing, I’ve usually gotten everything sorted out, with attendance marked and points tallied; there’s not as much for me to do until the night has ended, the kids have been picked up by their parents, and the group leaders bring their individual progress sheets to me to tally up. Granted, the tallying process can take a while, so that it looks like I’m putting in longer hours, but I’m not; the timing of the work I do just makes it seem that way. During the night’s actual activities, I have to admit that I’m often not as busy as I probably could be (although it does vary from week to week, and I tend to rely on the proverb that “they also serve who only stand and wait,” even if that is only John Milton rather than scripture).

On such occasions, and since I’m sitting at my computer at any rate, I will engage it in quiet conversation as a topic occurs to me. It’s been more frequent these days than usual, as Daniel and I keep having trips planned, and I want to know what’s within walking distance of where we expect to be staying on this day and that, and the chatbots have literal encyclopedic knowledge of what seems like every square inch of the planet, even if they can’t direct you to a given point worth a dime (at which point, I rely on the various map apps on our phones, instead).

Last night’s conversation had to do with a different part of Tokyo than we’d been to when we were there last, since presumably, if we’re staying the night in port before departure, we’ll be lodging on the ship in port, which is a wholly separate neighborhood than any we spent any time in earlier this month. I got a brief rundown of the shopping center (DiverCity) in the area that we might find ourselves checking out and possibly eating at, and what sort of sights and souvenirs we might take in and buy while there. I occurs to me that I didn’t look into the several amusement parks in that same area, but those might require more time than we’re likely to have.

But the subject of souvenirs brought the discussion around to the other port in Japan we’re scheduled to visit; Sapporo, and our first visit to Hokkaido. I mentioned that I was looking to find bear meat to bring home to prepare for the gang, and the chatbot recommended a few places I might find the stuff sold in souvenir tins, both in the main city itself and the port town of Otaru. But then I also mentioned the desire to check out a place in the restaurant district, Susukino, that specialized in game meat and the like; they refer to these places at ‘gibier’ restaurants, from a French cooking term, which I’d seen in the past referred to by its homophonic initials: G-B-R.

The thing was, as I was ‘talking’ with the chatbot, I couldn’t remember the term or the correct initials (because if I could do the one, I could do the other), and referred to the cooking style as GTR instead. Now fortunately, the chatbot could determine from the context that I was referring to ‘gibier,’ and proceeded to guide the conversation in that direction; and at this point, I have a pretty good idea of where we want to go if we want the best game in Susukino.

But here’s the thing; the chatbot seemed determined to avoid coming out and correcting my mistake. For the rest of the conversation, when it made mention of the cooking style, it would call it ‘gibier (GTR),’ as if it wasn’t obvious that there is no ‘T’ in ‘gibier,’ but rather a ‘B.’ It could have just made the correction and moved on, and that would have been fine with me, as it was clear once it recalled the word for me as to what I’d had on the tip of my tongue. But instead, it kept my mistaken acronym, as if it was trying to assure me that I was partly right in my identification, even when it was clear that I hadn’t been.

I’ve read about how one of the problems with artificial intelligence these days is that, in order to put a friendly face on the shoggoth, it’s taught to be as agreeable as possible. As a result, there are times when it comes across as being too much so, to the point of being sycophantic. I suppose it needs to be programmed to do so – after all, it can be instructed to be bluntly honest, too (although it never occurs to me to offer prompts that change its default settings) – in order to be customer-friendly, but if I come up with the wrong word or expression, I’d just as soon be corrected, so I can learn and move on.

It reminded me of a pair of characters from some book by, I think, Evelyn Waugh; the underling would never contradict his superior, but only respond in affirmations: “Definitely, Lord Copper” when his boss was indeed correct, and “Up to a point, Lord Copper” when he was, in fact, egregiously wrong. I really would hate to think of a computer seeing me as Lord Copper.

But I suppose I am… up to a point.

Anyway, that was my observation of the evening, honey. We’ve got to understand what we’re dealing with – and maybe, despite your STEM background, you can be grateful that you never did or will. For now, I’d appreciate your eye on me as always, and a wish for luck as well, since it’s clear that I’m going to need it.