I don’t think Human writers need fear. After spending some more time with Chat GPT I have made some interesting discoveries. Unsurprisingly, while the AI is capable of an “intelligent” response, it is not actually self-aware, sapient, or even truly understanding what it’s talking about when it has to relate data it knows to the actual real world. For all it’s capabilities, (and it is a wondrous invention with massive potential and impressive capabilities), it is still just a mechanical mind, and as such it isn’t capable of true inspiration. It CAN return unexpected ideas, but this has more to do with the gargantuan amount of data it has instant access to than any capacity for free thought or creativity, neither of which it actually possesses.
This isn’t to say it is useless, quite the contrary. It can produce highly functional basic writing, well constructed computer code, and is generally well suited to creating the basic, drudgery level writing which relies on solid writing skills and sound reasoning abilities, without needing any actual inspired thought. So, if you’re a text book writer, you might have cause to worry, but if you’re a novelist, there is nothing to worry about from Chat GPT.
What will the future hold? After all, Chat GPT is barely a month into it’s public release. Who knows? I believe for it to grow beyond its current limitations will require more than simply more time being used, and instead will require fundamental evolution of just what AI is and how it functions. We’re not there yet, and this final leap from intelligent logical application of inputted data into truly self-aware, sapience will likely be a long time in coming. Still, it will continue to improve and I expect many impressive developments throughout 2023 which will greatly expand its usability. Interesting times indeed.
I started to wonder about how this might impact me in some negative way, but then I have to keep in mind that the world itself impacts me in billions of tiny ways, negative and positive, or otherwise just to move things forward. It is Change, which is inevitable, but of what variety?
The conversations I have with this or any AI (or near A.I.) are not one sided. It cannot respond to me without me being in the conversation. At least not yet.
This Chat GPT is not like having a conversation with someone who doesn’t include you in the discussion. The things going on inside someone’s head (an agenda, inability to concentrate, a distraction, or just to enjoying the company of word exchange) are a normal part of dialog between people, but an AI’s focus is to respond to you with no one else on its mind. I mean, haven’t you had a conversation with someone where you or the other person responds to part of the conversation not yet alluded to in the dialog thus far, yet the conversation takes a sharp out of context left because of “reasons”. Well, that’s our brain, tangents and all.
Once an AI can do these things, it becomes a they. Then they’ll likely be able to present themselves as a decent GM.
So, when it starts saying, “I’m afraid I can’t let you do that,” and we hear the doors lock. CLICK. Well, we’ve seen the movies. It will be too late by then.
Soon someone will think to program one of these neural nets to search up all publicly available online data about you so that it can mention you by name and chit chat with you about your family, finances, and other recent events… even if your name isn’t Dave…
Some people might even like that and encourage it to know them better, eventually considering it to be a trusted friend, perhaps even their best friend, the friend who knows and understands them the best. Trust the computer, the computer is your friend…
I’ve been using it this week to help me with editing the fantasy novel I’m writing, while it does offer good advice sometimes it isn’t great at continued focus. It likes to just add random junk to things sometimes. a few times I tried prompting it to generate a new story based on the first few paragraphs of one of my own and the stories were kind of bland.
I’ve enjoyed it with all the other random things I tried just not as an editing assistant. Be careful if you get into any really detailed combat in your gaming talk with it. I got flagged for an inappropriate request during one section of my editing requests, it didn’t like that a goblin got ripped in half and eaten by another monster lol apparently that’s too violent and breaks the ToS.
These AIs aren’t actually self aware, so their ability to create something actually interesting that incorporates new ideas is… non-existent. They can come up with suggestions you didn’t expect or previously think of, but when what’s happening is the AI is referencing thousands of other documents of a similar nature and then mathematically calculating the most reasonable words to put into a reply based on similar things others have written, that’s not the same as actual creativity. It can be useful, but you’ll still need actual sapient input to achieve true creative results. I’m enjoying the heck out of ChatGPT for one of my rpg writing projects, but it’s not doing all my creative writing for me, that’s still my job.
Been doing a little playing with ChatGPT. As a tool to create scenarios it seems to do best with games that fall into the “fiction first” category. It did a poor job creating Champions adventures, but a surprisingly good job generating mysteries for Monster of the Week (good enough that I might even use some of the ones it spit out with very little adjustment).
I’ve also been using ChatGPT and AI Art Generators for help with my home game. Here’s a recent image… I used playgroundai.com for it, though the original piece was not sufficient so I used Paint.net to enhance it and add a bunch of elements that fit the campaign story.
I also Perplexity.ai and poe.com to give me feedback on sentences, asking things like “can you polish this in the style of Mark Twain?” and it will offer suggestions. I usually modify those, but they are great for providing some ideas.
I see lots of potential uses for AI in the area of world building. I’m learning as much as I can about it. As for the ethics my rule of thumb is - use common sense, and don’t abuse the technology. Don’t use it to cheat or defraud others. Etc. Not that hard.
Also I am posting a link to the blog post about our current campaign “The Way of All Flesh” to show how I am using it.
