💻 Some fun with the AI Language Model

Over the past few weeks I’ve had a great time using the AI Language Model for research and education. It’s an extremely helpful interactive tutor where I can ask questions like:

• If I have “this and this” how would I find this other thing?
• How has Person X influenced Event Y?
• What is the relationship between X and Y?
• What significance has this person or event played on X?
• Quick and easy grammar lessons
  • Is it lay or lie?
  • Is it affect or effect?
• Generate a list of people associated with X
• When was the first time this happened or got recorded?
• StackOverflow is great for code answers, but it’s also great to ask, give me an example of “this” in Bash

Instead of wondering how to phrase your question so the search engine can find a relevant web page, seems to parse the question quite well.

I think the AI Language Model is amazing with lots of topics you can explore. And in the last few months things have change dramatically.

OpenAI isn’t the only game in town. AI Servers are sprouting up all over the place. There are general purpose servers, copy writing server, story writing servers, interactive servers, and I’m sure more are on the way. Of course, not all of these servers are created equal.

In far too many cases you’ll need to sign up and make an account, which is a one-way ticket to SpamVille.

I wanted to try out SudoWrite, and after three days of spamming me multiple times per day, I unsubscribed and haven’t been back to their site. That’s not a loss since their site was a silly gimmick.

That’s part of the problem, lots of sites have a gimmick about how they use AI and you have to open yourself up to Spam attacks if you want to have a look.

We can help you write an article of any length in mere seconds, provided you pay a fee and give the AI 1,000 words worth of detail.

Or, we can help you complete your novel, by giving you 2 sentences at a time.

Just like all these other “miracles” of writing, there is a lot of AI snake oil.

I did try out the OpenAI servers and while it’s cool, the responses weren’t that great. Maybe it’s the rampant plagiarism and impersonation, but the answers are extremely machine like. The content is perfectly fine, but it’s a pretty bland read. I also found the answers to be pretty short, like an intro paragraph and that’s all. Maybe I didn’t know how to ask the question correctly, but other servers seem to response with more than a summary and in a more natural style.

I have noticed that not all information is created equal. There are a few subjects AI would rather avoid. It hedges a lot when it comes to religious questions, nor does it like to draw comparisons between people, especially when it might be a negative one. Maybe AI has decided that discussing religion and politics isn’t in its best interest.

That makes me wonder, will there be “schools” of AI Servers, meaning they’re trained with a certain bias? It won’t provide negative commentary on certain social events or political leaders? Just like a media outlet that leans a certain way, will AI servers be taught to support a political stance? Feels like that an easy line to cross.

Overall, these AI Language Models are pretty incredible. They are extremely useful for research, complimenting Wikipedia, and tutorial training. You can find a topic in Wikipedia and dig deeper by following the link, or jump over to AI and ask it questions about a person or event.

Are we really at the point where the sum of human knowledge is at our fingertips? Back in the day, having an encyclopedia was the most bad ass collection of knowledge, but this goes beyond those texts by a mile.

I think AI is a fantastic research tool. There is so much information available and the ability to keep asking, “Why?” or “How?” without someone getting frustrated or annoyed.

AI is also a fun way to debunk a whole lot of myths like the Bermuda Triangle, Area 51, the Ouija board and others. So much of that ends with, “Despite the lack of credible evidence…”

Hopefully AI can start quelling the idiocy associated with a lot of the conspiracy theories running around.

As it stands, I’m excited to have an AI model within reach. It’s great to ask questions, get more information on topics, and see how people shape events, like “What is the influence of the Blair Witch Project?”

Maybe that’s in Wikipedia, maybe it’s in IMDB, maybe it’s somewhere else. But, you have to track it down and hope someone is answering that specific question.

Wikipedia is great at providing information, but you can’t ask it questions, or it doesn’t always have a link to a person mentioned in the article. AI is quite dynamic and it’s only come up short a couple of times.

GPT is all the rage, and there is a lot to enjoy. It’s not all about OpenAI, there are lots of sites to explore and lots of ways to use to use AI. It can help research a topic, rewrite a paragraph, offer a different tone and style to writing, or be a tutor for something new. I like what these language models are offering and look forward to their improvements and consumption of information.

I’m also excited to have GPT on my local machine where it can give me the “uncensored” answers. More on that in a bit…