Open AI's ChatGPT is turning lots of heads on the internet. There are numerous articles out there from reputable sources, so I won't try and write a full in-depth article here. I just want to highlight some of the more amazing examples I've seen and talk about some of the implications of such a tool. One of the biggest questions out there is if we are seeing the birth of the tool that displaces Google. Google, the champion of search, is where most people turn to ask questions. The idea of searching for answers on the internet via a search engine essentially gave everyone an augmented intelligence of all of the trivia you might hear at a bar, or every cat video worth its salt. ChatGPT is working to raise the game. Instead of asking google and getting a list of links back that may get you the answer you want if you read through the links, imagine if you could talk with an expert on just about anything and get a well educated answer back on the topic. (and yes Google does surface some answers directly to certain queries now too) \ Here are some simple examples of how ChatGPT could take the concept way beyond Google. The queries can be how to write a program, or to simply be a request to go ahead and be said program so that you can interact with it. It can explain concepts at different levels. Heck, it can even make decent jokes. I got to ask a couple of queries before the system became overloaded with demand. Now that I see what it is capable of, I can't wait to go back and further refine my queries. (Of course I asked it about scenarios I'm working on for the sequel to The Quantum Contingent.) One of my queries is below. (It continued beyond the image shown... I was testing it from my phone) So, yes, my next novel will have a car chase in the self-driving future... it makes for some interesting drama...
Want to learn more? I recommend checking out the NYT's article at: https://www.nytimes.com/2022/12/05/technology/chatgpt-ai-twitter.html This is where I saw the above examples first... (before going to explore further on twitter) Drop you comments/questions below! Greg
4 Comments
MaryBeth Creamer
12/9/2022 04:22:04 am
Interesting.
Reply
2/13/2023 02:11:01 pm
Excellent questions MaryBeth! Unfortunately, I don't think AI will know when it needs re-educated. We'll need people in the loop to catch when it is going off the rails and try and keep it working in humanities self interest. Otherwise, we risk making some of the dystopian portrayals in the movies come true!
Reply
Mark W
1/22/2023 04:38:58 pm
I just tried ChatGPT for the first time. It gave an unexpected answer to my question, and when I pressed it a bit it admitted to being wrong and was even gracious enough to apologize.
Reply
2/13/2023 02:12:56 pm
Thanks for sharing Mark! Figuring out how to keep large language models like Chat-GPT honest is hard! they are predicting the next word based on their training and don't always get it right! They do an amazing job, considering they don't really know what they are talking about!
Reply
Leave a Reply. |
The BlogGreg's blog will cover some of the things he learned as well as some of the tech and locations he used in his new novel, The Quantum Contingent. Archives
August 2024
Categories |