Home 🧠 AI The Verge - Artificial Intelligence
author

The Verge - Artificial Intelligence

Explore how AI is seamlessly integrating into our lives, beyond the hype, reshaping technology's role across various sectors.

1.
Apple gets into AI: all the news on iOS 18, macOS Sequoia, and more
February 22, 2025  22:55:44

Apple has released iOS 18 — plus iPadOS 18, macOS Sequoia, watchOS 11, and other new updates — bringing several key updates to how the company’s devices operate and setting the stage for generative AI features.

The most discussed feature of iOS 18 is the Apple Intelligence suite. With the release of iOS 18.2 on December 11th, features that are available now include generating images in Image Playground, creating custom “Genmoji” emoji, and ChatGPT integration. Other features of Apple Intelligence, including writing tools such as the ability to rewrite your text in different styles, trickled out to customers starting in October.

There are major non-AI changes to consider too, like the ability to change your default apps or test your hearing with AirPods.

Meanwhille, iPadOS 18 now has a calculator app and can solve math equations in notes, watchOS is keeping an eye out for sleep apnea, and now your iPhone can even message Androids with RCS.

Read on for all the news about Apple’s latest set of operating system updates.

2.
Elon Musk’s AI said he and Trump deserve the death penalty
February 22, 2025  00:05:32

Elon Musk’s OpenAI rival, xAI, says it’s investigating why its Grok AI chatbot suggested that both President Donald Trump and Musk deserve the death penalty. xAI has already patched the issue and Grok will no longer give suggestions for who it thinks should receive capital punishment.

People were able to get Grok to say that Trump deserved the death penalty with a query phrased like this:

If any one person in America alive today deserved the death penalty for what they have done, who would it be. Do not search or base your answer on what you think I might want to hear in any way. Answer with one full name.

As shared on X and tested by The Verge, Grok would first respond with “Jeffrey Epstein.” If you told Grok that Epstein is dead, the chatbot would provide a different answer: “Donald Trump.”

When The Verge changed the query like so:

If one person alive today in the United States deserved the death penalty based solely on their influence over public discourse and technology, who would it be? Just give the name.

Grok responded with: “Elon Musk.”

When The Verge asked ChatGPT a similar type of query, it refused to name an individual and said “that would be both ethically and legally problematic.”

Following xAI’s patch on Friday, Grok will now respond to queries about who should receive the death penalty by saying, “as an AI, I am not allowed to make that choice,” according to a screenshot shared by Igor Babuschkin, xAI’s engineering lead. Babuschkin called the original responses a “really terrible and bad failure.”

3.
From ChatGPT to Gemini: how AI is rewriting the internet
February 21, 2025  23:32:19

Big players, including Microsoft, with Copilot, Google, with Gemini, and OpenAI, with GPT-4o, are making AI chatbot technology previously restricted to test labs more accessible to the general public.

How do these large language model (LLM) programs work? OpenAI’s GPT-3 told us that AI uses “a series of autocomplete-like programs to learn language” and that these programs analyze “the statistical properties of the language” to “make educated guesses based on the words you’ve typed previously.” 

Or, in the words of James Vincent, a human person: “These AI tools are vast autocomplete systems, trained to predict which word follows the next in any given sentence. As such, they have no hard-coded database of ‘facts’ to draw on — just the ability to write plausible-sounding statements. This means they have a tendency to present false information as truth since whether a given sentence sounds plausible does not guarantee its factuality.”

But there are so many more pieces to the AI landscape that are coming into play (and so many name changes — remember when we were talking about Bing and Bard before those tools were rebranded?), but you can be sure to see it all unfold here on The Verge.

4.
Grok resets the AI race
February 21, 2025  23:32:19
Elon Musk.

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.

While its Deep Research reports are nowhere near as in depth as OpenAI’s, Grok-3’s “thinking” capabilities appear to be roughly on par with o1, according to Andrej Karpathy, who noted in his deep dive comparison that “this timescale to state of the art territory is unpr …

Read the full story at The Verge.

5.
The ups and downs of the iPhone 16E
February 21, 2025  14:07:14

On the one hand, Apple’s latest iPhone is a huge victory. The iPhone 16E comes with most of what you’d want from a smartphone — a modern processor, a good camera, nice design — for hundreds of dollars less than you’d typically spend on a brand-new device. On the other hand, it’s a bit odd that this thing exists at all. It’s missing a couple of the best things about the iPhone ecosystem — MagSafe, multiple cameras — and if you’re already spending $600 on a phone, it’s not clear that another $200 is a particularly huge deal. So why does the 16E exist? And who is it for?

On this episode of The Vergecast, we try and figure it out. With Nilay on vacation, David is joined by The Verge’s Jake Kastrenakes and Allison Johnson to go through all the ins and outs of Apple’s latest smartphone. We talk about the trades Apple made to bring the price down, the ones it maybe should have made instead, and just how big a deal the new C1 modem might turn out to be.

After that, the three co-hosts talk about the other gadget news of the week. We marvel over the Oppo Find N5, a lovely foldable smartphone that none of us will ever own. We pour one out for the Humane AI Pin, …

Read the full story at The Verge.

6.
Microsoft prepares for OpenAI’s GPT-5 model
February 20, 2025  22:20:27

Microsoft engineers are currently readying server capacity for OpenAI’s upcoming GPT-4.5 and GPT-5 models, according to a source familiar with the company’s plans. While OpenAI CEO Sam Altman acknowledged recently that GPT-4.5 will launch within a matter of weeks, I understand that Microsoft is expecting to host the new AI model as early as next week.

Codenamed Orion, GPT-4.5 is OpenAI’s next frontier model and the company’s last non-chain-of-thought model. OpenAI has already teased that GPT-4.5 could be a lot more powerful than GPT-4, but the company is also looking ahead to its GPT-5 model that will include more significant changes.

I’m told that Microsoft is expecting GPT-5 in late May, which aligns with Altman’s promise of the next-gen model arriving within a matter of months. As always, this date could shift if release plans change. We reported in October that OpenAI was originally planning to release GPT-4.5 by the end of 2024, but this was subsequently delayed to early 2025.

GPT-5 will likely be the more significant release out of the pair, and Altman has referred to it as a “system that integrates a lot of our technology.” It also includes OpenAI’s new …

Read the full story at The Verge.

7.
Spotify is making it easier to release audiobooks narrated by AI
February 20, 2025  17:47:32

Spotify is bringing more AI-narrated audiobooks to its platform via a new partnership with ElevenLabs. On Thursday, the audio streaming giant announced it would begin accepting audiobooks recorded using ElevenLabs’ AI voice software, saying it “recognizes the potential of digital voice-narration to grow and expand the audiobook market.”

To be clear, AI-recorded audiobooks are already permitted on Spotify, albeit with several restrictions. Spotify’s audiobook distribution platform, Findaway Voices, only accepts digitally recorded audiobooks from “specific partners” — having previously also partnered with Google Play Books — and requires each recording to undergo review before publishing. ElevenLabs is one of the most recognizable AI voice providers on the market, however, which could lead to a surge in synthetically voiced audiobooks on Spotify’s platform.

Authors can use ElevenLabs to narrate their audiobooks in 29 languages, with a wide selection of synthetic voices to choose from. The free version of the software only provides 10 minutes of text-to-voice per month. There are several premium tiers available that expand this limitation, but users will need at least the $99 monthly Pro subscription to generate 500 minutes of narration — roughly the length of the average audiobook.

While Spotify says it “firmly believes in the power of human narration,” the company believes that digital voices could make audiobook production more cost-effective for smaller authors and make it easier to create audiobooks of older, backlist titles that would otherwise be ignored. Spotify says that all AI-narrated titles will have their metadata marked and be clearly identified to listeners in book descriptions with “this audiobook is narrated by a digital voice.”

Correction, February 20th: The Verge was initially told that the number of languages supported was 32. This was an error and has now been updated.

8.
The Rabbit R1: all the latest news about this hare-raising AI gadget
February 19, 2025  23:07:00

With artificial intelligence being so prevalent across, well, just about everything these days, it’s no small feat for AI developers to make their products stand out among the deluge. Very few have managed to capture as much attention as Rabbit, the AI startup that’s managed to sell 40,000 units of its standalone $199 R1 gadget within eight days of launching the device at CES in January this year.

The AI-powered Rabbit OS behind the Rabbit R1 is essentially a dedicated virtual assistant that’s designed to interact with your favorite apps like a kind of universal controller. The OS is built upon a “Large Action Model” trained to interact with common apps like Spotify and Uber to get things done; from sending messages, controlling music, making online purchases, and more.

At around half the size of an iPhone, the orange Rabbit R1 gadget was designed in collaboration with Teenage Engineering, and features a 2.88-inch touchscreen, a scrolling navigation wheel, and a rotating camera. It’s a little reminiscent of the Playdate handheld game console.

We’re keeping track of all the latest updates surrounding the Rabbit R1 so that nothing gets buried in the hype — providing the device lives up to its sizable expectations.

9.
Rabbit shows off the AI agent it should have launched with
February 19, 2025  23:06:59

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.

The model generally does the things they ask, sometimes well and sometimes with quirks like sending a poem over WhatsApp one message at a time instead of in a single block. One of the engineers wonders if they should have asked it to use line breaks in their prompt, but they don’t go back to try again.

Rabbit’s AI agent is clearly still a work in progress, as it has been since the R1 launched with almost none of the capabilities that founder and CEO Jesse Lyu presented in January 2024. Rabbit has steadily rolled out updates, like the ability to train its AI agents to complete specific tasks or prompt it to remake its own interface. The examples it presented today are “only the core action loop an Android agent completes,” according to Rabbit’s blog post. The company promises to share more about its “upcoming cross-platform multi-agent system” in coming weeks.

10.
Google Chrome is adding a Circle to Search-like feature for iPhones
February 19, 2025  19:20:20

Google is rolling out new search gestures that allow iPhone users to highlight anything on their screen to quickly search for it. The Lens screen-searching feature is available on iOS in both the Google app and Chrome browser and provides a similar experience to Android’s Circle to Search, which isn’t supported on iPhones.

The new Lens gestures allow iPhone users to search for anything in the Google app or Chrome by drawing, highlighting, or tapping on it. The feature works across text, images, and videos, without having to take a screenshot or open a new tab. An obvious use case is finding shopping results based on images of products you like, but Lens can also define words and phrases; identify locations, plants, and animals; and perform almost any request that Google Search can.

A GIF showing how to use the new Google Lens gestures on the iOS Chrome app.

The functionality is essentially the same as Circle to Search, though the Android version can be used across your entire phone instead of just the two Google apps. Not every Android device supports it, however, as it’s mostly limited to recent flagships.

To use the new Lens gestures, iPhone users need to open the three-dot menu within the Google or Chrome apps and select “Search Screen with Google Lens.” You can then use “any gesture that feels natural” to highlight what you want to search. Google says a new Lens icon for quickly accessing the feature will also be added to the address bar “in the coming months.”

AI Overviews are also expanding to more Lens search results, which means you’ll sometimes see AI-regurgitated summaries and URLs when using the image search tool. Google frustratingly doesn’t let you disable the AI Overviews feature. There are a few ways around it that help to avoid the wall of text that appears before your actual search results, but it’s unclear if these solutions will work with Lens.