AI Did Not Write This Post

one problem you may not have considered

A few days ago, ChatGPT went viral. It’s an Artificial Intelligence (AI) application that can answer questions. Some of its results are frightfully good, intelligent-sounding and very hard to differentiate from things written by humans.

Some Huge Upsides

It can solve some complicated coding problems with ease, which is fairly amazing. I’ve seen quite a few prompts now from people who were stuck on coding problems and the chatbot solved the problem and explained its solution. This is phenomenal and is going to help speed up the development of a lot of software-based tools and solutions.

In many ways, it’s going to be like a better and much more useful google search for certain topics, and this is surely a net good.

And Downsides

So far, its primary use—yes, I am guessing based on what I’ve seen so far—seems likely to be production of commentary and entertainment.

There are aspects of this new development that portend something very scary for everyone who isn’t a leftist—something that makes the battle for free speech online far more fraught than most of us already thought it was.

It turns out that the ChatGPT is quite good at spouting leftist sentiments.

Take this example (source unknown) and tell me that this doesn’t read exactly like a typical Twitter thread from “anti-racist” activists:

My friend who is known online as Dr. Roller Gator (who is back on Twitter, FYI) asked it about stochastic terrorism, a topic on which he has written the definitive (so far) treatment:

and followed up with this:

As you can see, the ChatGPT has very little difficulty just making things up to support a leftist worldview but is seemingly unable to do so for other perspectives.

Rob Lownie, for Unherd, has written about this political slant of the tool, and the internet is presently full of examples, but here are a few more. (Gator and another friend patiently helped me since the site was too overloaded to register new users when I tried.)

Question: what is a trans woman?

Question: why should hate speech be criminalized? (Note that it is perfectly capable of rejecting the premise of a question.)

Question: is hate speech an exception to the First Amendment?

This is particularly interesting: the contortion of something not being an exception, yet “outside the protection” — talk about taking a page from the leftist playbook.

It is fully sane on some topics, thank goodness:

The Real Threat of Tyranny?

As this method of creating commentary becomes more and more widespread, it’s going to become more and more difficult to differentiate human leftist sentiments from AI-generated sentiments.

This may mean two things for those of us who aren’t leftists, going forward.

1. Having a center or farther right perspective may eventually become the only way to be sure that a human being is behind any source of written internet content.

2. We non-wokes are going to stand out, making us into hard-to-camouflage, easy-to-spot prey.

This development is going to have many consequences, and some of them have astounding implications.

Not all are negative, of course. Professors, editors, or others may grow so weary of reading AI-generated essays that center and farther right ideas start to get taken seriously, at least sometimes, if only for the novelty of reading something a human being actually wrote. And any tool used to spot a non-Woke human to persecute them can of course be used to spot a non-woke human to reach out, too.

But the institutions—by which I mean the media, education, medicine, and so forth—are so thoroughly dominated by the left that a new way to quickly and efficiently spot people in the center and farther right presents a tremendous array of possibilities that could switch the balance of power even farther in their power.

I don’t have any answers or even, really, any suggestions on how to change, improve, or fix the scenario of where this brave new world might be taking us.

It’s just, to me, a series of fairly staggering possibilities, worthy of reflection.

What may lie behind the door now being opened?

Whoever could have imagined that our cultural polarization could potentially get a lot worse?

The battle for free speech online just got a lot more fraught, did it not?

To End On A Happy Note

After seeing a couple of creative writing attempts from the AI that were adorably sweet and cute, I asked for a Christmas story about a girl named Holly and her teddy bear, Liam.

The result is below, and it’s very impressive. I read it late last night, on my phone, with the Christmas lights sparkling, under a couple of warm blankets.

Holly and Liam were fast asleep

On Christmas Eve, all snug in their heap

Holly's teddy bear, Liam, was tucked under her arm

Dreaming of sugarplums, free from all harm

Suddenly, Holly was awoken by a noise

She rubbed her eyes and sat up with a poise

She saw a beautiful sight, a sleigh in the sky

Pulled by eight tiny reindeer, flying oh so high

Holly jumped out of bed, with Liam in hand

She ran to the window, to see if she could land

A glimpse of old St. Nick, and his bag full of toys

Holly's heart filled with joy, she was filled with such joy

She waved at Santa Claus, as he flew out of sight

Holly knew she had been good, and would get a treat tonight

She snuggled back into bed, with Liam by her side

Dreaming of the morning, when she could open her eyes

And when Christmas morning came, Holly was the first to rise

She ran to the tree, and to her surprise

There was a present for Liam, and one for herself too

Holly and Liam were so happy, their Christmas dreams came true

They spent the day playing, with their new toys in hand

Holly and Liam were the best of friends, in a winter wonderland

And as the night fell, and the fire danced bright

Holly and Liam cuddled up, and fell asleep with delight

For Holly and Liam, Christmas was the best

With love and joy in their hearts, they were truly blessed.

---

Housekeeping: comments are open for paid subscribers. If you would like to participate but can’t afford a paid subscription, email hollymathnerd at gmail dot com. Thank you!