

It's not intelligent and doesn't understand (it just manipulates tokens) anything. It's just a stupid program with bugs like any other program. A self-hallucinating bucket of bits.On Wednesday evening, OpenAI declared the ChatGPT writing nonsense issue (what they called "Unexpected responses from ChatGPT") as resolved, and the company's technical staff published a postmortem explanation on its official incidents page:
On February 20, 2024, an optimization to the user experience introduced a bug with how the model processes language.
LLMs generate responses by randomly sampling words based in part on probabilities. Their “language” consists of numbers that map to tokens.
In this case, the bug was in the step where the model chooses these numbers. Akin to being lost in translation, the model chose slightly wrong numbers, which produced word sequences that made no sense. More technically, inference kernels produced incorrect results when used in certain GPU configurations.
Upon identifying the cause of this incident, we rolled out a fix and confirmed that the incident was resolved.
Something tells me that the program will end up needing constant maintenance to its database, plus being babysitted by a numerous group of human "handlers"https://arstechnica.com/information...ting-out-shakespearean-nonsense-and-rambling/
ChatGPT goes temporarily “insane” with unexpected outputs, spooking users
Reddit user: "It's not just you, ChatGPT is having a stroke."
It's not intelligent and doesn't understand (it just manipulates tokens) anything. It's just a stupid program with bugs like any other program. A self-hallucinating bucket of bits.
It'll learn to babysit itself.Something tells me that the program will end up needing constant maintenance to its database, plus being babysitted by a numerous group of human "handlers"
It'll learn to babysit itself.
What, you've never left children alone together?
It's a common problem with these products.
Being technically correct while being effectively wrong.
https://www.dailymail.co.uk/sciencetech/article-13108515/Google-Gemini-racist-white-people.html
Google Gemini is accused of being racist towards white people: Users claim the AI bot refuses to create images of Caucasian people - after asking for photos of Popes, Vikings, and country music fans
View attachment 315871
https://www.bbc.com/news/business-68364690
Google to fix AI picture bot after 'woke' criticism
View attachment 315872
“After review, it has been determined that an outside consultant for the reelection campaign utilized ChatGPT in support of initiatives that were in fact completed by the Philadelphia Sheriff’s Office under the administration of Sheriff Rochelle Bilal,” read an unsigned statement released by Friends of Rochelle Bilal.
The campaign did not respond to follow-up questions seeking the name of Bilal’s campaign manager or the outside consultant.
“It is now clear that the artificial intelligence service generated fake news articles to support the initiatives that were part of the AI prompt,” the statement read. “Our campaign provided the outside consultant talking points which were then provided to the AI service.”
Not sure I'm following ... was aspect is it that is technically correct?It's a common problem with these products.
Being technically correct while being effectively wrong.
https://www.dailymail.co.uk/sciencetech/article-13108515/Google-Gemini-racist-white-people.html
Google Gemini is accused of being racist towards white people: Users claim the AI bot refuses to create images of Caucasian people - after asking for photos of Popes, Vikings, and country music fans
It generates fictional images by design so I'm not shocked. You ask for vikings, you get fictional generic vikings that technically are correct as a fictional cartoon viking but are a historical warping of representative reality to create that inclusive past fiction by IMO obvious human designed racial select intervention in some phase of the image generation process that's randomly seeded from any possible fact based source of actual vikings images. It generates mindless race biased cartoons that greatly reduce the utility of the program.Not sure I'm following ... was aspect is it that is technically correct?
Here would be a Catch-22 for Google Gemini (actually, LOTS of mines in this field) -- ask it to generate images of people that are generally associated with both negative stereotypes and white people, such as "evil capitalist". If it generates predominantly white people, or even better, white males, while only rarely depicting whites with positive stereotypes also generally associated with white people, then it becomes hard to claim it isn't strongly biased.
Also from the article was an example where it it was prompted to specifically generate a Caucasian pope image -- it did so, but accompanied it with a lecture on it being inappropriate to as assume that all popes are Caucasian. Makes me wonder, when asked to specifically generate an image of a black person in a role that is generally associated with a positive black stereotype, if it would give a similar lecture about it being inappropriate to assume that all "whatever" are black.
The more and more I hear about Altman, the more convinced I am that he is (or has become) a snake-oil salesman.
"Coding" != "Programming". You don't really program in a computer language. You express ideas from human intelligence in hardware or software codes some non-thinking, so called AI computer can handle.At the recent World Government Summit in Dubai, Nvidia CEO Jensen Huang made a counterintuitive break with what he thinks is a long tradition of tech CEOs advising youngsters to learn how to code. Huang argued that, even at this early stage of the artificial intelligence (AI) revolution, programming is no longer a vital skill. With coding taken care of by AI, humans can instead focus on more valuable expertise like biology, education, manufacturing, or farming, reasoned the Nvidia head.
Google is quickly losing something priceless called, trust in results.Google Mired in Controversy Over AI Chatbot Push
Search company is betting on AI for its next big business, but the tech’s pitfalls are playing out in real time
Google has less room for mistakes because of the trust it has built with users of its search engine over many years, Sheth said. “The world trusts it implicitly with giving them the truth.”
IMO the common definition of hacking is getting a system to do something it wasn't designed to do.OpenAI did not name the "hired gun" who it said the Times used to manipulate its systems and did not accuse the newspaper of breaking any anti-hacking laws.
"What OpenAI bizarrely mischaracterizes as 'hacking' is simply using OpenAI's products to look for evidence that they stole and reproduced The Times's copyrighted work," the newspaper's attorney Ian Crosby said in a statement on Tuesday.
Representatives for OpenAI did not immediately respond to a request for comment on the filing.
The truth, which will come out in the course of this case, is that the Times paid
someone to hack OpenAI’s products. It took them tens of thousands of attempts to generate the
highly anomalous results that make up Exhibit J to the Complaint.
She does not deny that AI might someday reach a similar level of intelligent understanding. But machine understanding may turn out to be different from human understanding. Nobody knows what sort of technology might achieve that understanding and what the nature of such understanding might be.
Large language models aren't headed toward ANY understanding -- that have exactly ZERO understanding.
Buy-now-pay-later lender Klarna said its AI assistant, powered by OpenAI, is doing the equivalent work of 700 full-time agents and has had 2.3 million conversations, equal to two-thirds of the company’s customer service chats, within the first month of being deployed. The AI tool resolved errands much faster and matched human levels on customer satisfaction, Klarna said.
Teleperformance shares fell as much as 29% in Paris trading, the steepest drop since November 2022, amid regular halts for volatility. The company already uses AI to manage simple processes on behalf of its clients, it said in a statement in response to the stock drop.
“The group’s current activity in no way reflects the negative conclusions in its business that could be drawn from the technological developments mentioned in this communication,” Teleperformance said.
| Thread starter | Similar threads | Forum | Replies | Date |
|---|---|---|---|---|
|
|
asked Chatgpt to recommend some upgrades/modifications for my Akai MPC3000 | General Electronics Chat | 1 |