The premiere GPT from Musk's xAI is capable of answering spicy questions with politically incorrect responses.
Two former Apple employees (and Sam Altman) promise their AI pin will revolutionize our relationships with the internet. But what is it?
New research shows how popular LLMs are able to accurately guess a user’s race, occupation, or location, after being fed seemingly trivial chats.
OpenAI hopes to make its API the obvious choice for developers at its DevDay conference on November 6th.
Top tech and business leaders begged generative AI companies to pause new developments for six months. AI companies clearly didn't listen.
Training and running AI needs a lot of power and a lot of water, but Microsoft may want to use microreactors to fuel its big artificial intelligence ambitions.
OpenAI said ChatGPT will soon 'see' your images and 'hear' your words, but the company is warning users the chatbot could go off the rails.
Take a quick look at the newest and weirdest happenings in AI.
The AI model is reportedly aimed at assisting businesses with services.
August marked a decline in active users on ChatGPT for the third month in a row, but back-to-school season could turn things around.
Code Llama includes several different versions that can write in C++, Java, Bash, and there's even a specialized version for writing Python.
DEF CON's big AI hack-fest. Plus: OpenAI's new moderation efforts, NYT lawsuit drama, and Werner Herzog.
That pirated set of works contained 37 GB worth of text for training AI. Companies like Meta have already used it to train their language models.
Researchers found lefty leanings in some of the AI's responses, which ChatGPT likely picked up from its training data.
Zoom executives promised Monday that the company wouldn't use your calls as AI fodder, even while its TOS said otherwise.
IBM researchers convinced large language models to play a multi-layered "game" of offering incorrect advice to prove they are "ethical and fair."
Pamera uses an object identifier and GPT-4 to generate poems based on photos in a matter of seconds. We tested it out on pics of a day in New York City.
As companies like Chegg struggle against instantaneous ChatGPT answers, textbook company Pearson wants AI to tutor students instead.
UCLA researchers’ new study says the GPT-3 language model can reason with analogy, but they can’t say the answers aren’t already in the training data.
ChatGPT can now remember users' previous commands, and OpenAI is promising it will spend more time fine-tuning its GPT-3.5 and GPT-4 models.
