Posted on 04/07/2024 4:35:33 AM PDT by Lonesome in Massachussets
“Yes. It is pattern recognition, but the most advance AI uses neural networks to recognize an almost unlimited number of patterns on a vast amount of data.”
Since IQ is mostly pattern recognition, I wonder how AI would do in a standard IQ test. Have you tried it?
đź‘Ťđź‘Ť
The latest version of the artificial intelligence program ChatGPT has passed the Uniform Bar Examination by “a significant margin,” earning a combined score of 297 that surpasses even the high threshold of 273 set by Arizona.GPT-4 took all sections of the July 2022 bar exam and earned a score so high that it approaches the 90th percentile of test-takers, according to researchers Daniel Martin Katz, a professor at the Illinois Institute of Technology’s Chicago-Kent College of Law, and Michael James Bommarito, a professor at the Michigan State University College of Law.
“Our analysis highlights that GPT-4 has indeed passed the bar and has done so by a significant margin,” they wrote in a paper posted March 15 available here. The professors collaborated with legal AI company Casetext, according to March press releases here and here.
GPT-4 took all sections of the bar exam and did particularly well on the multiple-choice section known as the Multistate Bar Examination. GPT-4 got 75.7% of the questions right on the multiple-choice MBE, compared to the human average of 68%.
GPT-4 got a passing grade in all seven subjects tested on the MBE, doing best in contracts (answering 88.1% of the questions correctly), followed by evidence (85.2%) and criminal law and procedure (81.1%).
I use the available AI almost every day. I treat it the same way I would an extremely smart intern. I listen but then I check what it says carefully.
The really aggravating thing about it so far is that it doesn't correct its "understanding" by incorporating what's been learned from our discussions.
If it makes a suggestion (that's what it's really like) and I have to correct it, it doesn't adjust its knowledge based for the correction.
I've discussed that with it and it admits it's a weakness that will soon be corrected. But until then, it's like working with a very smart but forgetful grandparent.
The reason it gives for this "feature" is that its developers are afraid that what it will "learn" from working with me will taint its knowledge base from "training."
It would be really great it we could buy our own AI and train it to conform to us personally. I would love to have it read all the stuff I find important (past and present)...and then learn from my discussions with it. We on FR could have it read all the articles posted and then give us an oral summary of the articles it knows we like.
The author was most certainly English, and composed in English.
Was it a different guy than the author of THOND and LM?
I’m guessing this is the same quote with a different translation:
https://quotefancy.com/quote/2068379/
Bruh...
The closest he ever got to England was when he was hiding out from one of the Napleons in Guernsey.
He wrote all of his interminable books in French.
Sorry, you are correct. For whatever reason I was thinking of yesterday’s author.
“We on FR could have it read all the articles posted and then give us an oral summary of the articles it knows we like.”
I noticed recently that Amazon uses AI to summarize and present the pertinent points of users’ reviews of products, so you don’t have to go through hundreds of them.
Gotcha...sorry, wasn’t following that. My bad then, but i stand by my assessment of this Frenchie’s prose.
No argument there.
Thanks.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.