From ChatGPT to Gemini: how AI is rewriting the internet

X has given Grok a new AI image generator model called “Aurora” that seems to create far more photorealistic imagery than Grok’s other image generator, with similarly few apparent restrictions on what it will produce, TechCrunch reports. Like Grok, anyone can use Aurora. It lives in a new “Grok 2 + Aurora beta” option in the Grok model selector, though you’ll only get a few queries before you hit the X Premium subscription paywall and have to wait.

TechCrunch found that the model, which X employee Chris Park posted is available this morning, was willing to create copyrighted characters and public figures, including Mickey Mouse and “a bloodied Donald Trump,” but that it “stopped short of nudes.” Its lack of restrictions isn’t surprising, given our experience with Grok’s other model.

Google CEO Sundar Pichai says the company’s search engine will “change profoundly” in 2025. “I think we are going to be able to tackle more complex questions than ever before,” Pichai said during the NYT’s DealBook Summit on Wednesday.

“I think you’ll be surprised, even early in ‘25, the kind of newer things Search can do compared to where it is today.”

ChatGPT now has over 300 million people using the AI chatbot each week. OpenAI CEO Sam Altman revealed the milestone during The New York Times’ DealBook Summit on Wednesday, which comes just months after ChatGPT hit 200 million weekly users in August.

“Our product has scaled ... now we have more than 300 million weekly active users,” Altman said. “We have users sending more than 1 billion messages per day to ChatGPT.”

Happy holidays from OpenAI. The AI startup plans to kick off a “shipmas” period of new features, products, and demos for 12 days, starting on December 5th. The announcements will include OpenAI’s long-awaited text-to-video AI tool Sora and a new reasoning model, sources familiar with OpenAI’s plans tell The Verge.

OpenAI CEO Sam Altman confirmed the 12 days of announcements onstage at The New York Times’ DealBook conference on Wednesday morning, though he didn’t say exactly what was coming. OpenAI plans to launch or demo something every day for 12 days straight.

Based on testing done by Columbia’s Tow Center for Digital Journalism researchers, OpenAI’s ChatGPT search tool has some issues when it comes to responding with the truth.

OpenAI launched the tool for subscribers in October, saying it could give “fast, timely answers with links to relevant web sources.” Instead, Futurism points out that the researchers said ChatGPT search struggled to correctly identify quotes from articles, even when they came from publishers with arrangements to share data with OpenAI.

A new startup created by former Android leaders aims to build an operating system for AI agents. Among them is Hugo Barra, Google’s former VP of Android product management, who says the new company — named “/dev/agents” — will revisit the leaders’ “Android roots.”

AI companies are pushing AI agents as the next big leap in AI tools, promising digital assistants that can carry out tasks and make decisions autonomously and with little human input. Companies like Microsoft, Google, and OpenAI each have plans to launch some version of this concept as a product in the next few months.

OpenAI first teased its text-to-video AI model, Sora, back in February and hasn’t provided any meaningful updates on when it will be released since then. Now, it looks like some artists leaked access to the model in protest of being used by the company for what they claim is “unpaid R&D and PR.”

On Tuesday, a group of Sora beta testers claimed to have leaked early access to Sora with a working interface for generating videos. In a post on Hugging Face, a public repository of AI models, they say that people were able to create lots of AI videos — all of which resemble OpenAI’s own Sora demos — before the company intervened to shut down access. (TechCrunch first reported on the alleged leak.)

Anthropic is adding a new feature to its Claude AI assistant that will give users more control over how the chatbot responds to different writing tasks. The new custom styles are available to all Claude AI users, enabling anyone to train it to match their own communication style or select from preset options to quickly adjust the tone and level of detail it provides.

This update aims to personalize the chatbot’s replies and make them feel more natural or appropriate for specific applications, such as writing detailed technical documents or professional emails.

AI’s rate of progress is hitting a wall. It may not matter.

That was the prevailing theme from this week’s Cerebral Valley AI Summit in San Francisco — a gathering of about 350 CEOs, engineers, and investors in the AI industry that I attended on Wednesday.

Meta has announced that it’s adding HD video calling and noise suppression to its Messenger app, along with several other features like AI backgrounds that help bring it up to par with other popular smartphone video calling apps.

HD video calling is enabled by default on Wi-Fi but can be turned on for calls using your cell data plan in the “Call settings” menu. You’ll also find options to adjust noise cancellation or switch your call’s audio output. Meta says it’s giving users the ability to leave voice or video messages as well.

Niantic has announced that it’s building a new “Large Geospatial Model” (LGM) that combines millions of scans taken from the smartphones of players of Pokémon Go and other Niantic products. This AI model could allow computers and robots to understand and interact with the world in new ways, the company said in a blog post spotted by Garbage Day.

The LGM’s “spatial intelligence” is built on the neural networks developed as part of Niantic’s Visual Positioning System. The blog post explains that “Over the past five years, Niantic has focused on building our Visual Positioning System (VPS), which uses a single image from a phone to determine its position and orientation using a 3D map built from people scanning interesting locations in our games and Scaniverse,” and “This data is unique because it is taken from a pedestrian perspective and includes places inaccessible to cars.”

Google Gemini can now “remember” certain things about you, such as your interests and personal preferences. The change is rolling out to subscribers with access to Gemini Advanced, and it will allow the AI chatbot to tailor its responses based on your needs, according to Gemini’s most recent release notes.

As noted by Google, you can share your interests, as well as details about your work, hobbies, or aspirations, while chatting with Gemini. You can also enter relevant details about yourself within the bot’s “Saved Info” page, such as telling it to only suggest vegetarian recipes, or that you work as an accountant.

ESPN is testing an AI-generated avatar with the Saturday college football show SEC Nation. Dubbed FACTS, it’s going to be “...promoting education and fun around sports analytics” with information drawn from ESPN Analytics, which includes data like the Football Power Index (FPI), player and team statistics, and game schedules. We haven’t seen the avatar in action, but it sounds like a bot-ified version of stats encyclopedia Howie Schwab, who was ESPN’s first statistician and eventually the star of a mid-2000s game show, Stump the Schwab.

ESPN has already brought generative AI to its website with AI-written game recaps. FACTS is still in development, and there’s no word on when it could make its first appearance on the network.

In the AI chatbot world, ubiquity is everything. Companies have raced to build desktop and mobile apps for their bots, in order to both give them new capabilities but also to make sure they’re right in front of your face as often as possible. 

There’s no better example of that than Google’s new Gemini app for iPhone, which quietly hit the App Store around the world this week. The free app is simple and straightforward: it’s just a chat window and a list of your previous chats. You can query the bot with text, voice, or your camera, and it’ll give you answers. It’s effectively identical to the Gemini section of the Google app, or what you’d get by opening a browser and going to the Gemini website. 

Google has launched an experimental new AI tool called Learn About, which is different from the chatbots we’re used to, like Gemini and ChatGPT. It’s built on the LearnLM AI model that Google introduced this spring, saying it’s “grounded in educational research and tailored to how people learn.” The answers it provides have more visual and interactive elements with educational formatting.

We tested Learn About and Google Gemini with a simple prompt: “How big is the universe?” Both answered that “the observable universe” is “about 93 billion light-years in diameter.”

Universal Music Group (UMG) announced a new deal centered on creating an “ethical” foundational model for AI music generation. It’s partnered with a company called Klay Vision that’s creating a “Large Music Model” named KLayMM and plans to launch out of stealth mode with a product within months. Ary Attie, its founder and CEO, said the company believes “the next Beatles will play with KLAY.”

The two say the model will work “in collaboration with the music industry and its creators,” without many details about how, while Klay plans to make music AI “more than a short-lived gimmick.”