Playing with what is all the rage and below of a short conversation. I can think of couple forum members better suited to train the AI on the subject of audiophiles haha.
A very high tech, expensive, example of the old computer maxim “garbage in, garbage out”
My buddy had the genius idea of asking the AI to “rewrite the Golden Girl’s theme song, but make it about enemies.”
These chatbots generate a word salad with no thought, nuance, or emotion. It is purely the result of a predictive algorithm for word choice based on machine learning and probabilities. We assign it meaning because the words are in the proper order to represent a thought. It is amusing but meaningless.
That’s an oddly narrow understanding. As a counter-example, and useful to me, it can tell you what a piece of code is doing. Have you seen the art it can create yet? It’s so much more more than word salad. I started studying AI in the late 90s…we cared about fuzzy logic at that time, which is more than word salads.
Many years ago while touring Berkeley National Laboratory, the visitor center was running an early version of an AI chatbot. I found it to be a silly curiosity and quickly lost interest. I see little has changed over the decades. Which is commentary on how little AI has advanced relative to early predictions on pace of development. Or maybe just commentary on chatbots being a toy of choice for the bored.
Maybe word salad was not the best choice of words. Salads are made from ingredients selected for their taste and color. The words an AI chatbot uses are selected via programming algorithms rather than thought. I did not say the salad was not pretty or tasteless. My point is we, the reader of the output, are the ones who assign meaning. The bot only generates the words. I was speaking only in the area of chat AI. That’s why it seems narrow. It was intended to be.
That being said, it will be very useful for aggregating and analyzing written words. It will be the death of search engine lists, see the latest bing announcement.
I think it’s interesting and potentially helpful tech when applied with benevolent intent but honestly I’d like to find someone who can convince me that it shouldn’t scare the hell outta me.
It’s where the human condition meets math, coupled with big data. It’s a natural progression that we have been thinking about with science fiction for many years. I’m excited and scared!
Here was a great little piece on the CBS Sunday Morning show:
This is a great list of ideas to seperate the bored from the super-productive:
This may not have not gotten any national press and it goes well beyond chatbots but the 3rd White House AI advisory committee convened at SAS Institute last Friday.
“ On Friday, AI experts from the world of business, academia and nonprofits gathered on the SAS Institute campus in Cary to discuss how the United States government can best harness the expanding capabilities of artificial intelligence…
“We have an incredibly important and also relatively broad mandate,” said Victoria Espinel, CEO of the software industry group BSA, at the start of the meeting. “AI can lead to enormous societal benefits if it is used responsibly. It is also clear that it can lead to significant adverse consequences if it is not used responsibly.””
Raleigh News and Observer 12Feb23
Her was and is a bizarre movie but getting to be less of a stretch.