The Figure 02 runs 50% longer with three times the compute and inference power of its predecessor.

Nvidia helped itself to "a human lifetime visual experience worth of training data per day," according to company execs.

OpenAI has finally released its highly anticipated Advanced Voice feature to a select few of its ChatGPT-Plus subscribers, but everybody will have it by fall.

ChatGPT has reportedly been considering implementation of a text-detection tool for over a year now.

If you hadn't noticed the end of Meta's celebrity chatbot program, don't worry, neither did anyone else.

Mark Zuckerberg expects the compute requirements for next year's Llama 4 to be "10 times" that of last month's Llama 3.

ChatGPT's Advanced Voice mode arrived on Tuesday for a select few OpenAI subscribers chosen to be part of the highly-anticipated feature's alpha release.

This could signal the start of a wider industry revolt against the so-called "Nvidia tax."

The Coca-Cola Company announced Monday that it has partnered with WPP Open X in an effort to incorporate AI from Nvidia into its global ad campaigns. 

Engineering researchers have developed and demonstrated a new computer memory design that could drastically reduce the amount of energy AI systems consume.

ChatGPT's Advanced Voice feature could arrive soon, but only a lucky few will be part of July's alpha test.

Kling can generate videos up to 2 minutes in length, though they start at just 5 seconds.

Google announced Thursday that it is releasing Gemini 1.5 Flash, it's snack-sized large language model, to all users, including the free tier.

ChatGPT and Copilot are two of the most popular platforms today. Both run on GPT-4, but which is best for you?

The new LLM will be available in three sizes — 8B, 70B, and 405B parameters, with the latter being the largest open-source AI built to date

Musk has stated that Grok 3 will be a huge improvement and topple its competitors when it supposedly launches in late 2024.

The company is promising "genuinely useful humanoid robots" if they ever do arrive.

These text-to-video and image-to-video AI generators are really fun to play with, and let you try them out for free.

GPT-4o mini is both less resource intensive and cheaper to operate than its the standard GPT-4 model.

According to a new report from Proof News, Apple has been using a dataset containing the subtitles of 173,536 YouTube videos to train its AI.

Mere months after departing OpenAI, AI researcher Andrej Karpathy announced the launch of his new AI+Education startup, dubbed Eureka Labs.

Google's senior adviser on AI governance made a disturbing discovery: His Gemini AI was accessing and summarizing PDF files that it shouldn't be able to.

An Apple research team has developed a new method to ease users into the quirks of upgraded AI models.

The company has also revealed a five-tier scale for measuring its AI system capabilities.

Microsoft has announced that its new VALL-E 2 voice synthesizer AI is too capable to safely release to the public.

Lost in an unfamiliar office building, big box store, or warehouse? Just ask the nearest robot for directions.

Barely a week after news broke that Apple's Phil Schiller would join the OpenAI board in a non-voting observer role, the two companies are nixing that plan.

OpenAI's Sam Altman and HuffPo founder Arianna Huffington announced they are teaming up to develop an “AI health coach” via Thrive AI Health.

Stability AI has improved the quality of its models, as well as updated the licensing agreement for an expanded free tier.

Both Claude 3.5 Sonnet and GPT-4o represent the best that AI assistants currently have to offer. We'll tell you which one is ultimately superior.

A hacker infiltrated OpenAI's internal messaging system last year and stole details about the company's AI designs.

The visionary tech leader that brought us the iPhone also had some strong opinions on AI.

The ElevenLabs Reader app now gives you access to the AI voices of Judy Garland, James Dean, Burt Reynolds, and Sir Lawrence Olivier to narrate PDFs.

Meta's new "3D Gen" model is a "state-of-the-art, fast pipeline" for transforming input text into high-fidelity 3D images in under a minute.