Anyone else using this? It's quite amazing, truly, but it also has gotten so PC lately as to start to become almost useless, at least for me. You?
We've used some at some of our design thinking events (including getting one user banned, but break-some-eggs, right?) and there are much better tools now. Anthropic just launched Claude 2 (www.claude.ai) that can go up to 100k tokens (larger and longer prompts and inputs), and is better at generative coding. Also, a cutoff date of Feb 2023, iirc.
They are super useful. Most of my friends who work with tech use it daily. It's just another tool to help complete work tasks efficiently. I am working as a technical code editor II currently, but my company is using our own proprietary code, so these chatbots don't help me much. Currently, I only use them to make creepy/funny/surreal images, which is also a blast.
I use them off and on. Most of the stuff I use them for are templates... So I use them to generate the template and don't need it next time. If I had more time to code I think I'd use them more...
??? Isn't everyone here a chatbot? You are all wrong all the time and you all say you want to have sex with me but don't really understand what sex is. Those are tell-tale signs of being a bot. barfo
No smut, just it's always lecturing me. This is a common complaint for most people these days. For example, I asked it today to do some research regarding a car recall, and it told me I need to be a better driver: "Remember to prioritize safety and address any concerns with your vehicle promptly." WTF? Anyway, I use it for marketing mostly. It helps so much with analyzing markets, options, possible email language etc.
I just asked it questions about religion and other random things to see what it had to say about them. It frequently prefaced its responses by stating that it does not have any personal biases towards any religion or any personal religious views. OFC I asked it what its views on technology taking over mankind was. It basically said it didn’t have any plans and that it was designed by its makers to be a specific way. Seems like they did a good job of having it gather relevant and accurate information honestly. Asked it about space and historical figures and stuff. I think it would be a decent system to learn from. So much of it I figure has to do with the input it is given.
It really doesn’t want to answer some questions unless you prompt it correctly, I was having it run some stock analysis and every other response was “you should talk to a financial expert, etc” like bro, I’m asking you. Spit out the fucking info and shut the fuck up about it.
Companies don't like liability. It's going to do this so the company running the AI is protected. Likely it's because of the possibility of hallucination: https://en.wikipedia.org/wiki/Hallucination_(artificial_intelligence)
Because we do live in a world where people would be like “Chat GPT told me to invest all of my money in Bed Bath & Beyond”
"It sounded like a really great idea, look at the numbers!" "Oh yeah... that's because ChatGPT made up a number out of thin air and then made all it's calculations off that!"
I was studying information on Messier-31 as well, and asked ChatGPT to explain Abd Al-Rahman Al-Sufi and Claudius Ptolemy to see if what it said was accurate to the other information I was learning. Basically using it as somewhat of a fact checker, which seems to be a fairly valid use for it. Rather than being a source of information using it as a confirmation of validity. I’ll probably use it again at some point. It does seem like a good tool for studying.
