From ChatGPT to Gemini: how AI is rewriting the internet

Yesterday morning, billionaire Los Angeles Times owner Patrick Soon-Shiong published a letter to readers letting them know the outlet is now using AI to add a “Voices” label to articles that take “a stance” or are “written from a personal perspective.” He said those articles may also get a set of AI-generated “Insights,” which appear at the bottom as bullet points, including some labeled, “Different views on the topic.”

“Voices is not strictly limited to Opinion section content,” writes Soon-Shiong, ”It also includes news commentary, criticism, reviews, and more. If a piece takes a stance or is written from a personal perspective, it may be labeled Voices.“ He also says, “I believe providing more varied viewpoints supports our journalistic mission and will help readers navigate the issues facing this nation.”

Read Article >

Microsoft has announced Microsoft Dragon Copilot, an AI system for healthcare that can, among other things, listen to and create notes based on clinical visits. The system combines voice-dictating and ambient listening tech created by AI voice company Nuance, which Microsoft bought in 2021.

According to Microsoft’s announcement, the new system can help its users streamline their documentation through features like “multilanguage ambient note creation” and natural language dictation. Its AI assistant offers “general-purpose medical information searches from trusted content sources,” as well as the ability to automate tasks like “conversational orders, note and clinical evidence summaries, referral letters, and after visit summaries.”

Read Article >

Anthropic is releasing Claude 3.7 Sonnet, its first “hybrid reasoning model” that can solve more complex problems and outperforms previous models in areas like math and coding.

In addition to a new model, Anthropic is also releasing a “limited research preview” of its “agentic” coding tool called Claude Code. While Anthropic already powers AI coding tools like Cursor, it’s pitching Claude Code as “an active collaborator that can search and read code, edit files, write and run tests, commit and push code to GitHub, and use command line tools.”

Read Article >

Just a few weeks after everyone freaked out about DeepSeek, Elon Musk’s Grok-3 has again shaken up the fast-moving AI race. The new model is ending the week at the top of the Chatbot Arena leaderboard, while the Grok iOS app is at the top of the App Store, just above ChatGPT. Even as Musk appears to be crashing out from his newfound political power, his xAI team has managed to deploy a leading foundational model in record time.

It’s one thing to have the leading model; it’s another to build the biggest user base around it. Musk seems to understand that if he wants to crush OpenAI, he has to shift attention away from ChatGPT. Since the debut of Grok-3, Musk has said that ChatGPT-like voice interaction and desktop apps are coming soon. Where his product roadmap appears to differ considerably from OpenAI’s is xAI’s nascent efforts to build an AI gaming studio, though the details there are scarce.

Read Article >

The Humane AI Pin has collapsed, but Rabbit is still kicking. The company published a blog post and video today showing off a “generalist Android agent,” slowly controlling apps on a tablet in much the same way that Rabbit claimed its R1 device would over a year ago. (It couldn’t, and can’t.) The work builds on LAM Playground, a “generalist web agent” Rabbit launched last year.

The engineers don’t use the Rabbit R1 at all for the demonstration. Instead, they type their requests into a prompt box on a laptop, which translates them to actions on an Android tablet. They task it with things like finding a YouTube video or locating a whiskey cocktail recipe in a cocktail app, gathering the ingredients, and then adding them to a Google Keep grocery list. At one point, they ask it to download the puzzle game 2048 and figure out how to play it, which it does, albeit slowly.

Read Article >

Google Meet’s Gemini-powered note-taking feature is getting a bit more useful for Workspace teams. It will now generate a checklist of suggested next steps at the end of your meetings, going as far as assigning due dates and attaching a primary stakeholder to the task.

The note-taking feature was initially launched in August last year. I’ve played around a bit with it here at The Verge since then, and thus far, it hasn’t produced totally disastrous results.

Read Article >

Amazon won’t launch the AI-powered upgrade for Alexa for at least a month after its showcase at an event set for February 26th, according to The Washington Post. The delay is reportedly at least partly because the updated assistant has issues with giving inaccurate answers to test questions.

An anonymous Amazon employee told the outlet that the upgrade won’t come “until March 31 or later” due to the issues. The new Alexa could be tied to a subscription, with features like “the ability to adopt a personality, recall conversations, order takeout or call a taxi,” and was originally set to launch later this month as a free trial, the Post writes, citing internal documents and messages.

Read Article >

On Tuesday, US District Court of Delaware judge Stephanos Bibas issued a partial summary judgment in favor of Thomson Reuters in its copyright infringement lawsuit against Ross Intelligence, a legal AI startup. Filed in 2020, it’s one of the first cases that will deal with the legality of AI tools and how they are trained, often using copyrighted data scraped from somewhere else without license or permission.

Similar lawsuits against OpenAI, Microsoft, and other AI giants are currently winding their way through the courts, and they could come down to similar questions about whether or not the AI tools can claim a “fair use” defense of using copyrighted material.

Read Article >

AI chatbots struggle with factual inaccuracies and distortions when summarizing news stories, research from the BBC has found. The study, which examined whether OpenAI’s ChatGPT, Google Gemini, Microsoft Copilot, and Perplexity can accurately summarize news, found more than half of all the AI-generated output had “significant issues of some form.”

As part of the study, the BBC asked ChatGPT, Copilot, Gemini, and Perplexity to provide summaries of 100 BBC news articles, while journalists reviewed their answers. In addition to finding major issues in 51 percent of responses, the BBC found that 19 percent of answers citing the BBC included incorrect statements, numbers, and dates. Meanwhile, 13 percent of quotes from the BBC were “either altered from the original source or not present in the article cited.”

Read Article >

ChatGPT no longer requires you to log in to use the AI chatbot’s search engine, OpenAI announced on Wednesday. With the feature, ChatGPT will surface responses based on information from the web while presenting a list of sources it used to inform its answer.

OpenAI first launched its search engine to paid ChatGPT subscribers last October and later rolled it out to everyone in December. But now that you no longer need an account to use it, ChatGPT search will compete directly with search engines like Google and Bing.

Read Article >

Google is bringing its experimental “reasoning” artificial intelligence model capable of explaining how it answers complex questions to the Gemini app. The Gemini 2.0 Flash Thinking update is part of a slew of Gemini 2.0 AI rollouts announced by Google today, including its latest Gemini 2.0 Pro flagship model.

This comes as the search giant is expecting to invest $75 billion on expenditures like growing its monotonously named family of AI models this year. That’s a considerable jump from the $32.3 billion on capital expenditures it spent in 2023, with Google now racing to keep up with AI competitors like OpenAI, Microsoft, Meta, and the Amazon-backed Anthropic.

Read Article >

OpenAI has revealed another new agentic feature for ChatGPT called deep research, which it says can operate autonomously to “plan and execute a multi-step trajectory to find the data it needs, backtracking and reacting to real-time information where necessary.”

Instead of simply generating text, it shows a summary of its process in a sidebar, with citations and a summary showing the process used for reference.

Read Article >

Gemini has some new abilities that could make it more helpful in Sheets, Google announced in a post on the Workspace blog. Now, Gemini can respond to questions about your data with details about trends or by creating static charts that you can insert into your spreadsheet as images. The new capability is rolling out now to most Workspace plans and to users on the $19.99-per-month Google One AI Premium plan.

Google says Gemini does all of this by creating and running Python code, then producing an analysis of the code’s results. For simpler requests, it may use normal spreadsheet formulas, but the bottom line is that it could save you the tedium and headache that normally comes with creating data visualizations. Before this, Gemini was limited to simpler tasks like telling you how to do things in Sheets or creating tables for you.

Read Article >

Meta is widely launching the ability for its AI chatbot to “remember” certain details about you, such as your dietary preferences or your interests, the company said in a blog post on Monday. It will then use your past conversations, in addition to details from Facebook and Instagram accounts, to provide more relevant recommendations.

Meta first started rolling out a memory feature for its AI chatbot last year, but now it will be available across Facebook, Messenger, and WhatsApp on iOS and Android in the US and Canada. Though you can tell Meta AI to remember certain things, like that you love traveling, it will also “pick up important details based on context.”

Read Article >

After surging to the top of Apple’s App Store charts in the US, DeepSeek’s AI Assistant is now restricting new user sign-ups. According to an incident report page, registrations are being temporarily limited “due to large-scale malicious attacks on DeepSeek’s services,” though it’s unclear how these limitations are being applied.

“Existing users can log in as usual,” DeepSeek said in its update. “Thanks for your understanding and support.” An alert banner on the DeepSeek web sign-up page says that “registration may be busy,” rather than entirely restricted, however, and encourages users to wait and “try again” if their application is unsuccessful.

Read Article >