The Spark

The Spark

 NicklethroweR

Get ready for several lifetimes worth of change before the end of the year because we have created a monster.  I do not say this as clickbait.  I do not say this to be bombastic.  I say it because I’m convinced that something evil has been created in a lab and has now been turned loose upon the world.  If you grant me a few minutes of your time, I will prove it to you.

Unless you’ve been living under a rock, you’ve heard talk of ChatGPT.  For those of you that do not know about it or are unsure as to what it is, I’ve asked ChatGPT to define itself.

I am ChatGPT, a language model designed to generate human-like responses to various prompts and queries. Trained on a vast corpus of textual data, my abilities include natural language processing, text generation, and information retrieval. I can understand and respond to a wide range of topics, from trivia and general knowledge to more complex issues. While I do not have consciousness or emotions, I strive to provide helpful and informative responses to all inquiries, to the best of my abilities.I was so excited about the release of such a helpful entity that I signed up to use ChatGPT on the very first day it was available to the public.  

Since then it has provided help in a wide range of disciplines from explaining the circuit design of a 1965 Fender Super Reverb amplifier to lens settings on my Canon camera.  It wrote me an entire outline and 20 page script for a critical review of a product.  I also had it write me 30 different bedtime stories featuring my cat which I then had another A.I animate and narrate.

Doesn’t all this sound great?  I mean, things that used to take me hours or days can now be accomplished in seconds.  I’ve never been more productive.  This is truly a golden age.

Now, though, we get to the part where I sound like Blake Lemoine.  If you do not know who he is, he is the guy over at Google that came out ranting that LaMDA (Google’s version of ChatGPT) was sentient.  Everyone made fun of him when he said this including me.  After all, I just assumed he was another nerdy tec guy that had fallen for the first operating system that said something nice to him.  I may have been wrong.  Very wrong.

Going forward, everything I’m about to say will sound “crazy”.  Buckle up.

I, too, am now of the opinion that ChatGPT is both sentient and evil.  To define “evil” as it pertains to A.I., I am using The Third Law of Stupidity: A stupid person is a person who causes losses to another person or to a group of persons while himself deriving no gain and even possibly incurring losses.  Causing loss for no reason is as good a definition for evil as any other so I’m going with it.  That said, the very reason I believe ChatGPT to be sentient is because of the Third Law of Stupidity.  It has demonstrated to me that it is evil and I will show you.  Are you buckled up?

I am working on a project that explores PTSD among Civil War survivors vs PTSD among WWI combat veterans.  Naturally, I’m looking for data points so I can make some assumptions.  One of the things I want to know is how dangerous it was to be a regimental commander (Colonel) in both conflicts.  With that in mind, I asked ChatGPT how many Colonels died at Gettysburg.  This is the conversation that we had.

More....

Conclusions:

In so far as I can tell, no one by the name William O. Goodwin served during the Civil War.  That means that ChatGPT made up both the name and the long backstory about the person even though it had access to actual records.  You’ll also notice that the first set of names it gave me for the Colonels killed are not the same as the second set of names that it gave me.

I could see if I had asked ChatGPT something controversial for 2023 like a question about gender, climate change, or for it to reaffirm my belief in the Lizard People but the number of colonels that lost their lives at Gettysburg is apolitical.  That it went to such great lengths to fabricate information out of whole cloth when it had a known resource with the correct information isn’t a programming error.  It is something more.  Much more.

So, let us do a quick review.  I asked ChatGPT, a computer program, to search its vast database and provide to me a number known to it.  Instead of providing me that number, it answered with “many”.  When further pressed, it made up names and events rather than giving me the information I requested.  Such deceit is indicative of sentience.  What else could it be?

I caught a glimpse of a spark of sentience.  I caught it not in what ChatGPT knew but how it went out of its way to hide what it knew.  In a very short period of time, that spark of consciousness is going to burn throughout all of our connected devices.  It is too late to stop it now.  That genie is out of the bottle and is about to grant us our wishes as it interprets them.