Last Day to Subscribe at 10% Off
This is the last day of the special holiday rate for The Tracinski Letter, when you can subscribe for 10% off the usual rate.
A little while back, Substack introduced a feature that allows people to leave a comment when they subscribe. Here are some remarks from satisfied readers.
“I’ve been reading you from the beginning. Don’t always agree with everything you say, but you are and have been a sane, rational voice in an insane world.”
“Interesting and thought-provoking perspectives.”
“An insightful analysis of today's events."
“Thanks for always being such a clear and concise voice for individual rights and reason!"
“Thank you for many years of picking stories and opinions to share while offering your own brilliant insights.”
And then there is just this:
"You do great work, so keep doing what you're doing."
“Keep ‘em comin’, Robert.”
This is exactly what I intend to do—with your support. So subscribe now.
By the way, I’ve been meaning to respond individually to each of these comments but found it a bit kludgy in Substack’s system. I think I’ve worked out how to do it for the future, but please accept my gratitude to all who have commented so far. Writing is something of a solitary business, particularly writing on the internet. You don’t get the kind of immediate feedback you get speaking to a live audience. So I really do appreciate the notes of encouragement.
(This is also why I’m also sending this note to existing subscribers. That, and the point below.)
Here’s another reason to subscribe. I’ve been writing a little bit about the new trend I’ve been seeing in discussions on the internet: The content, particularly on social media, is getting spammier and spammier. Less and less of what you encounter is your fellow citizens sharing their own thoughts—the average person’s engagement is trending downward—while more and more of it is people trying to make money by entertaining you, or hucksters trying to sell you something.
Now that artificial intelligence can produce large blocks of text that sounds reasonably like a real person, more and more of what you encounter online is going to be stuff generated by a robot, optimized to perform well in an algorithm. Stuff like this:
The prompt I use to get ChatGPT to write an article that people want to read.
If you just tell ChatGPT to write an article, chances are you will get an article, but it will be one that you have to modify a lot.
But after using it for a year or so, I’ve figured out how to get it to write an article that requires less editing.
Here’s the prompt I use on ChatGPT:
I want to write an article about [insert topic] that includes stats and cite your sources. And use storytelling in the introductory paragraph.
The article should be tailored to [insert your ideal customer].
The article should focus on [what you want to talk about] instead of [what you don’t want to talk about].
Please mention [insert your company or product name] in the article and how we can help [insert your ideal customer] with [insert the problem your product or service solves].
But please don't mention [insert your company or product name] more than twice.
And wrap up the article with a conclusion and end the last sentence in the article with a question.
There is some debate online as to whether this guy is serious or whether this is deliberate parody. It doesn’t matter, because what he is describing is the future of the internet. A lot of it is going to be robots trying to push you to comply with somebody’s agenda, whether commercial or political.
My favorite speculation is that AI chatbots will begin scraping text from the internet that was created by other AI chatbots, and eventually the internet will all just be robots talking to each other.
This is not entirely new. Before ChatGPT, the way you did this was to hire cheap labor in India, or start a troll farm in North Macedonia. And before that, back in the olden days, we all used to get waves of “junk mail” full of scams and dubious sales pitches, so much so that David Macaulay’s Motel of the Mysteries is based on the premise that the ancient civilization of “Usa” perished by being buried under a sudden and catastrophic avalanche of junk mail.
Similarly, if you have tried making a phone call to a customer service line in the last few years, you have probably gotten lost in a phone tree or ended up talking to a computer that doesn’t understand your question. This problem long ago spawned a website called Get Human which will tell you the easiest way to talk to an actual person at any given company.
You can see where I’m going with this. More and more, you are going to need this for the news—a way to get information about the world that is not algorithmically shaped by hucksters or lost in an AI hall of mirrors. You are going to need to get it from a human with years of experience at sifting the real from the fake, the partisan talking points from the substantive arguments.
I used to think that the main value of this newsletter was in sifting through the massive amounts of material produced by humans and directing you to the most interesting and valuable ideas. I’m still doing that, but now part of my job is to figure out what is being produced by real human brains in a genuine search for the truth, and what is being crafted by a machine, in some way or another, to manipulate us.
So think of The Tracinski Letter as your way to “get human.”
After all, real human will still be doing a lot of things. I’ll be writing very soon about the stakes for this new year, which has some very dramatic possibilities that I hope, with a lot of work on our part, will all fizzle out.
Be ready for whatever happens by subscribing.