Tales from the jar side: AI is taking our jobs, The four quadrants of American politics, A cliché at a memorial service (but a good one), a New Mockito video, and Tweets and Toots
Q: What do you call someone who points out the obvious? A: Someone who points out the obvious. (rimshot)
Welcome, fellow jarheads, to Tales from the jar side, the Kousen IT newsletter, for the week of March 12 - 19, 2023. This week I taught the second week of my Spring in 3 Weeks course on the O’Reilly Learning Platform, and my Functional Java basics course as an NFJS Virtual Workshop.
Regular readers of (and listeners to, and now video viewers of) this newsletter are affectionately known as jarheads, and are far more intelligent, sophisticated, and attractive than the average newsletter reader (or listener, or viewer). If you wish to become a jarhead, please subscribe using this button:
As a reminder, when this message is truncated in email, click on the title to open it in a browser tab formatted properly for both the web and mobile.
AI Took ‘Er Jobs!
This week I read an excellent blog post by Justin Searls entitled How to Tell if AI Threatens YOUR Job. It’s very well written and I recommend you take a look, but here’s a summary of his major points. In the section called Three simple rules for keeping your job, Searls points out that there are three categories of tasks that AI is particularly bad at, so if your job involves one or more of them, you’re probably safe, at least for now. Those task types are:
Novel: If what you do has never been done before, so its not well represented in existing training data, you’re good.
Unpredictable: It helps if the solutions you devise are not obvious from the descriptions of the problems.
Fragile: AIs are notoriously imprecise, given that they’re just matching patterns. If changing a single detail affects the entire solution, AIs are going to struggle.
As a software person, I use Google Copilot a lot, and it has been enormously helpful. Any developer who uses it, however, quickly realizes that you have to keep an eye on it, because it’s wrong a lot, too. You have to know what you want and how to describe it for an AI to come close. My teaching job tends to involve a lot of similar examples that I present repeatedly, so it’s ideal for AI, and yet it’s still wrong quite often.
I recently saw another AI-related video on Nebula, however (which I’d link to, but I don’t remember which one) that talked about a lawyer discussing whether he was worried about being replaced or not. His conclusion was that it’s worth learning how to use the AI tools, because, like most productivity tools, a lawyer with the tool will likely be able to compete better than a lawyer without the tool.
In my opinion, the real danger, at least in the business world, is not that an AI can do your job. It’s that a manager who doesn’t understand what you do will think an AI can do your job.
Let me emphasize that:
The danger isn’t that an AI can do your job — it’s that some manager above you will think an AI can do your job.
That would lead to a devaluation of your job, possibly to the point the manager eliminates you in favor of an AI and then suffers the consequences.
My suggestion when communicating with such non-technical people is to mention the points above, but emphasize that AI tools are like dumb, inexperienced assistants. They produce a lot and do it quickly, but only a fraction of their results are usable, and they need to be monitored at all times. You could also get into the assorted legal and bias issues, but a lot of companies are willing to ignore those if they believe they’ll get away with it and save a lot of money.
Incidentally, the blog post was written by a consultant, and that reminds me of an old story about consultants. A plant manager was having trouble with a particular machine, which had a tendency to vibrate so much it frequently fell apart. He eventually hired a consultant to fix it. The consultant looked the machine over, then tightened one screw (which fixed the problem), and sent the manager a bill for $10K. The plant manager was outraged that 10 minutes worth of work would cost so much, so the consultant followed up with an itemized bill:
Turn one screw, $5.
Knowing which screw to turn, $9995.
The advantage an experienced person will have over AIs, at least for a while, is knowing which screw to turn.
(By contrast, my other favorite story about consultants is this definition: A consultant is someone who borrows your watch, tells you what time it is, pockets the watch, and sends you a bill for it. Good work if you can get it.)
The Four Quadrants of American Politics
Another interesting article I ran across this week is The Four Quadrants of American Politics, by Ronald Brownstein in The Atlantic. It discusses a statistical study of all 435 districts in the (US) House of Representatives and evaluates them along two axes:
The percentage of minorities (defined as anyone nonwhite), relative to the national average, and
The percentage of college-educated whites, relative to national average.
The article divides the four “quadrants” as lo-lo, lo-hi, hi-lo, and hi-hi. Usually such and oversimplification has so many exceptions it’s not useful, but that wasn’t the case here. The conclusions are not terribly surprising: Republicans dominate the lo-lo category, while Democrats do well (but not as dominant) in the hi-hi group and lead in both of the other two groups.
The article further suggests that the last few “wave” elections drove out most of the politicians that might cross those boundaries, like Democrats in lo-lo districts or Republicans in hi-hi ones, so that our national divide these days is really a chasm, making it very hard to ever compromise on anything. The real conclusion is that the district numbers are so balanced (for assorted reasons like redistricting, voter suppression, and simple gerrymandering), that the House is likely to swing back and forth over the next few elections, just as it has done over the last few.
Anyway, it was an interesting article, and you might want to take a look. For the record, I live in Connecticut’s 2nd district, which covers nearly half the state:
Such a large district spans many areas on the economic scale. From what I was able to gather, we’re in the lo-hi category, with very low diversity (only 22% vs 40%), but higher on the education measure that the national average (39% vs 36%), though not as high as I thought. Our representative is Joe Courtney (D), who first won a very narrow election back in 2006 (by only 83 votes out of more than 242,000 (!), by far the closest election that year), but has become an institution since then. So that tracks, pretty much.
Memorial Service
This week I was hired to sing at a memorial service at our local church. That happens occasionally. I’m well-known there, and people who know me call sometimes to have me perform at special events.
(If you want a sample, see my 7-part Ave Maria (Franz Biebl) with Justin Halley.)
Every once in a while, a memorial service is more than just comfort for the family. Sometimes, like yesterday, it’s downright uplifting. This person lived into his 80s and was a member of the church for over 50 years. I learned that he grew up under very difficult circumstances, but somehow managed to be upbeat, positive, and friendly to everyone in his life anyway. I knew him only slightly — we’d met several times, and I talked to him whenever we ran into each other, but I couldn’t tell you much about him other than that he smiled a lot and he and his wife were inseparable.
It turned out that in addition to a long career as a teacher and being very fond of Dad Jokes (which you know I can identify with), apparently he had a favorite saying he used to tell the kids:
Nobody cares how much you know, but everybody knows how much you care.
When one of the eulogy speakers said this, I thought, wow, what a cliché, but one I’d never heard before, and it’s a good one.
These polarized days when some people view empathy not simply with contempt but as an opportunity for them to take advantage of you, it’s good to see that there are those precious individuals who choose to live their lives caring anyway. I agree with that, and at my age I know that’s just who I am and I couldn’t change it if I tried, but it’s nice knowing I wasn’t alone. I’m really glad I was able to attend the service and contribute in some small way.
Mocking Final Types
I uploaded a new Mockito video this week to the Tales from the jar side YouTube channel. It’s called Mocking Final Classes and Methods:
As usual, it’s based on a section of my Mockito Made Clear book, but I fleshed it out a bit. I also wrote up a companion Medium article that will appear in a few weeks in the Pragmatic Programmers publication there.
Once again, I released it late on a Friday afternoon, which wasn’t ideal, but it’s when I had time to do it. So be it. :)
Tweets and Toots
But what does the fox say?
Huh. That explains a lot.
Susan being lazy
Might want to try harder next time.
But it promised!
It seemed so sure of itself.
Speaking of kindness
Admittedly, that doesn’t make being laughed at any easier, but it’s still worth it.
Great Joke for 1979
I saw the original Alien movie when it came out in 1979, when I was a high school kid spending my first summer away from home. A friend and I watched it during a matinee and emerged into the daylight totally freaked out. Even now the movie still holds up, more or less, but back then it was terrifying.
(On the off chance you didn’t get the joke, the tag line for the movie was “In space, no one can hear you scream.” Brilliant, though debatable. Lots of people heard screaming in the movie.)
Daylight Saving Time
I think I’m over it by now, but I hate that adjustment every year.
Speaking of AI
Exactly. I have to find somewhere to use that gag.
Lab Tested
Works for me.
Keep your appointments
Maybe they started without you.
Where have you gone, Karen Carpenter?
Must be the breadcrumbs I keep in my pockets. Yeah, that’s it.
Indiana Jones and the Comfort of Tenure
“Beware March 15” doesn’t have the same ring to it
And finally, in celebration:
Et tu, developer?
Have a great week, everybody. :)
The video version of this newsletter will be on the Tales from the jar side YouTube channel tomorrow.
As a reminder, you can see all my upcoming training courses on the O’Reilly Learning Platform here and all the upcoming NFJS Virtual Workshops here.
Last week:
Week 1 of Spring and Spring Boot in 3 Weeks, on the O’Reilly Learning Platform
Modern Java: Functional Programming with Streams, Lambdas, and Method References, an NFJS Virtual Workshop
This week:
Week 2 of Spring and Spring Boot in 3 Weeks, on the O’Reilly Learning Platform
Gradle Concepts and Best Practices, an NFJS Virtual Workshop.
The joke about 80-year-old professors hit a little to close.