Meta launches a standalone AI app to compete with ChatGPT


After integrating Meta AI into WhatsApp, Instagram, Facebook, and Messenger, Meta is rolling out a standalone AI app. Unveiled at Meta’s LlamaCon event on Tuesday, this app allows users to access Meta AI in a standalone app, similar to the ChatGPT app and other AI assistant apps.

To win over users, Meta is trying to leverage what makes it different from companies like OpenAI and Anthropic — Meta already has a sense of who you are, what you like, and who you hang out with based on years of data that you’ve likely shared on Facebook or Instagram.

Meta’s AI app can differentiate itself from existing AI assistants because it can “[draw] on information you’ve already chosen to share on Meta products,” the company said, such as your profile and the content you engage with. So far, these personalized responses will be available in the US and Canada.

You can also give Meta more information about you to remember for future conversations with its AI — for example, you can tell the AI that you are lactose intolerant, which it could remember before recommending that you go to a wine and cheese tasting on your next vacation.

As with any AI product, users should be aware of how Meta may use the data that they share with its chatbots. Meta relies on its wealth of user data to power its targeted advertising business, which makes up the bulk of its revenue.

Image Credits:Meta

Meta’s AI app also introduces a Discover feed, where you can share how you’re using AI with your friends — in a mockup image, Meta shows someone asking the AI to describe them in 3 emojis, which they then shared with their friends. A user’s interactions with Meta AI will only be shared to the feed if they choose to do so.

This feed might amplify certain generative AI trends, like the recent trend in which people tried to make themselves look like Barbie dolls or Studio Ghibli characters. But then again, not every app needs to have a social feed — we’re looking at you, Venmo.

Read what Mark Zuckerberg and Facebook execs said about Instagram before buying it


The first week of the Meta antitrust trial brought new revelations about how the company formerly known as Facebook approached the competitive threat posed by Instagram in the early 2010s.

The U.S. government is accusing Meta of violating competition laws by acquiring companies like Instagram and WhatsApp that threatened the Facebook monopoly. If lawyers for the U.S. Federal Trade Commission (FTC) are successful, the government could force Meta to break up its business by selling off Instagram and WhatsApp.

As part of the trial, the FTC shared compelling evidence to demonstrate that Facebook was very much aware of the risk Instagram created for its business as the photo-sharing app grew in popularity. In documents containing Facebook’s internal emails, Facebook execs fret over Instagram’s growth and discuss how much to pay for the app, if Facebook were to acquire it.

The company execs also discuss other strategies for limiting Instagram’s growth, including copying its functionality and releasing an app of their own, or buying the app and then no longer adding new features to it while working on its own products.

Facebook’s strategy to either buy or bury its competition is on display in these conversations, according to the government’s arguments. In addition to showing how the company was thinking about its competition at the time, the messages are indicative of the cutthroat strategies that have allowed Meta to become the social networking behemoth it is today.

Some of the highlights from these messages are below.

Mark Zuckerberg and others worry about Instagram’s rapid growth

  • “Instagram seems like it’s growing quickly. In 4 months they’re up to 2m users and 30k daily photo uploads. That’s a lot. We need to track this closely. Also, apparently Dropbox’s next big push is going to be in photo sharing.” — Mark Zuckerberg, February 2011
  • “If Instagram continues to kick ass on mobile, or if Google buys them, then over the next few years they could easily add pieces of their service that copy what we’re doing now, and if they have a growing number of people’s photos then that’s a real issue for us. They’re growing extremely quickly right now. It seems like they double every couple of months or so, and their base is already ~5-10m users. As soon as we launch a compelling product a lot of people will use ours more and future Instagram users will find no reason to use them. But at the current rate, literally every couple of months that we waste translates to a double in their growth and a harder position for us to work our way out of.”  — Mark Zuckerberg, September 2011
  • “The photos team is now focused almost exclusively on a new mobile photo app as we gawk at Instagram’s simple photo-sharing app taking off (and even our own app sees fat growth … mobile uploads increased to 17.7M day, +5.3 w/w). Like Beluga, watching these guys explode validates our strategy of de-cluttering our mobile experience and offering standalone messaging and photos products outside from the monolithic app garden.” — Chris Cox, chief product officer, February 2011
  • “One concerning trend is that a huge number of people are using Instagram every day — including everyone ranging from non-technical high school friends to even FB employees — and they’re only uploading some of their photos to FB. This creates a huge hole for us and one that I’m sure anything we’re going to do on platform or with social dynamics will completely solve.” — Mark Zuckerberg, February 2012

Facebook considers an Instagram acquisition, stopping its development and growth

  • “I wonder if we should consider buying Instagram, even if it costs ~500M. Right now they seem to have two things that we don’t: a really good camera and a photo-centric sharing network.” — Mark Zuckerberg, February 2012
  • “I think it’s quite possible that our initial thesis was wrong and theirs is right — that what people want is more to take the best photos than to put them on FB … we might want to consider paying a lot of money for this.” — Mark Zuckerberg, February 2012
  • “I actually think that there is a serious argument to be made that we should buy Path, Pinterest, Instagram, Evernote, and whomever else we really admire/are doing great things right now if (1) we can structure it in a way that we keep their products up & running but transition the teams to working on FB proper; (2) we think the people deeply care about building great things and we think we can lock them up for 4+ years to work on our platform.” — Samuel W. Lessin (former Facebook VP of Product), corresponding with Mark Zuckerberg in February 2012
  • “I think what we’d do is keep their product running and just not add more features to it, and focus future development on our products, including building all of their camera features into ours. By not killing their products we prevent everyone from hating us and we make sure we don’t immediately create a hole in the market for someone else to fill, but all future development would go towards our core products.” — Mark Zuckerberg, February 2012
  • “One way of looking at this is that what we’re really buying is time. Even if some new competitors springs [sic] up, buying Instagram, Path, Foursquare, etc now will give us a year or more to integrate their dynamics before anyone can get close to their scale again.” — Mark Zuckerberg, February 2012

Meta’s new board members include former Trump security advisor


Meta has announced that Patrick Collison and Dina Powell McCormick are joining its board of directors on April 15. Collison is the co-founder and CEO of Stripe, the payment processor and financial services company that he started with his brother. He is also the co-founder of the Arc Institute, a biomedical science and technology institution. Meanwhile, Powell McCormick was a partner at Goldman Sachs and ran its Global Sovereign investment banking business. She also worked for the US government and served as Deputy National Security Advisor to President Donald Trump during his first term.

Powell McCormick helped shape the Trump administration’s foreign policy, especially in regards to the Middle East as an Egyptian-American. She served as an Assistant to the President and Senior Counselor for Economic Initiatives during Trump’s first term, as well. Her Trump-era appointments weren’t the first time she worked in the US government, though: Back during the George W. Bush administration, she served as the Assistant to the President for Presidential Personnel, Assistant Secretary of State for Educational and Cultural Affairs and Deputy Undersecretary of State for Public Affairs and Public Diplomacy. She is married to Republican senator Dave McCormick.

“Patrick and Dina bring a lot of experience supporting businesses and entrepreneurs to our board. Patrick is deeply committed to expanding economic opportunity, and Dina has a long career advocating for economic development and supporting entrepreneurs. Their perspective will be extremely valuable to businesses that rely on our services to grow,” Meta chief Mark Zuckerberg said in a statement.

In January, Meta also welcomed UFC CEO Dana White, who’s a known friend and associate of Donald Trump and a supporter of his re-election bid, to its board of directors. White’s and Powell McCormick’s additions to Meta’s board are signs that the company is tacking right now that Trump is back in power, and perhaps one of the ways the company is ingratiating itself to the current administration.

Meta’s benchmarks for its new AI models are a bit misleading


One of the new flagship AI models Meta released on Saturday, Maverick, ranks second on LM Arena, a test that has human raters compare the outputs of models and choose which they prefer. But it seems the version of Maverick that Meta deployed to LM Arena differs from the version that’s widely available to developers.

As several AI researchers pointed out on X, Meta noted in its announcement that the Maverick on LM Arena is an “experimental chat version.” A chart on the official Llama website, meanwhile, discloses that Meta’s LM Arena testing was conducted using “Llama 4 Maverick optimized for conversationality.”

As we’ve written about before, for various reasons, LM Arena has never been the most reliable measure of an AI model’s performance. But AI companies generally haven’t customized or otherwise fine-tuned their models to score better on LM Arena — or haven’t admitted to doing so, at least.

The problem with tailoring a model to a benchmark, withholding it, and then releasing a “vanilla” variant of that same model is that it makes it challenging for developers to predict exactly how well the model will perform in particular contexts. It’s also misleading. Ideally, benchmarks — woefully inadequate as they are — provide a snapshot of a single model’s strengths and weaknesses across a range of tasks.

Indeed, researchers on X have observed stark differences in the behavior of the publicly downloadable Maverick compared with the model hosted on LM Arena. The LM Arena version seems to use a lot of emojis, and give incredibly long-winded answers.

We’ve reached out to Meta and Chatbot Arena, the organization that maintains LM Arena, for comment.



Meta introduces Llama 4 with two new AI models available now, and two more on the way


Meta has released the first two models from its multimodal Llama 4 suite: LLama 4 Scout and Llama 4 Maverick. Maverick is “the workhorse” of the two and excels at image and text understanding for “general assistant and chat use cases,” the company said in a blog post, while the smaller model Scout could tackle things like “multi-document summarization, parsing extensive user activity for personalized tasks, and reasoning over vast codebases.” The company also introduced Llama 4 Behemoth, an upcoming model it says is “among the world’s smartest LLMs” — and CEO Mark Zuckerberg said we’ll be hearing about a fourth model, LLama 4 Reasoning, “in the next month.”

Both Maverick and Scout are available to download now from the LLama website and Hugging Face, and they’ve been added to Meta AI, including for WhatsApp, Messenger and Instagram DMs.

A text slide describing three models from the Llama 4 family: Llama 4 Behemoth, Llama 4 Maverick and Llama 4 ScoutA text slide describing three models from the Llama 4 family: Llama 4 Behemoth, Llama 4 Maverick and Llama 4 Scout

Meta

Scout has 17 billion active parameters with 16 experts, Meta says. According to Zuckerberg, “It’s extremely fast, natively multimodal, and has an industry leading, nearly infinite 10 million token context length, and it is designed to run on a single GPU.” Maverick on the other hand has 17 billion active parameters with 128 experts. The company says it beats competitors like GPT-4o and Gemini 2.0 on coding, reasoning, multilingual, long-context and image benchmarks, and stacks up against DeepSeek v3.1 on reasoning and coding.

Zuckerberg is already calling the upcoming Behemoth model, which is still training, “the highest performing base model in the world,” with 288 billion active parameters, according to the company. It may not be here yet, but it’s likely we’ll be hearing a lot more about that and the Reasoning model soon; Meta’s big AI developer conference, LlamaCon, is just a few weeks away.



Meta’s next Llama models may have upgraded voice features


Meta’s next major “open” AI model may have a voice focus, per a report in Financial Times.

According to the piece, Meta is planning to introduce improved voice features with Llama 4, the next flagship in its Llama model family, which is expected to arrive in “weeks.” Reportedly, Meta has been particularly focused on allowing users to interrupt the model mid-speech, similar to OpenAI’s Voice Mode for ChatGPT and Google’s Gemini Live experience.

In comments this week at a Morgan Stanley conference, Meta chief product officer Chris Cox said that Llama 4 will be an “omni” model, capable of natively interpreting and outputting speech as well as text and other types of data.

The success of open models from the Chinese AI lab DeepSeek, which perform on par or better than Meta’s Llama models, has kicked Llama development into overdrive. Meta is said to have scrambled to set up war rooms to decipher how DeepSeek lowered the cost of running and deploying models.

Meta is building Oakley smart glasses for athletes: report


Meta’s Reality Labs is expanding its lineup of AI smart glasses with a new pair of Oakley-branded spectacles targeted at athletes, according to Bloomberg’s Mark Gurman.

Bloomberg reports the new smart glasses, with which Meta seeks to capitalize on the Ray-Ban Meta’s success, will be based on Oakley’s Sphaera glasses, and will shift the camera toward the center of the glasses frame.

It seems Meta wants to tap into Oakley’s cult-like following among various athlete groups, including cyclists, runners, and rowers. This could help popularize smart glasses with more mainstream consumers.

Beyond new glasses, Meta is also exploring a wide range of AI-powered devices, including camera-equipped earbuds and a smart watch. The company is also developing a $1,000 pair of smart glasses based on the Ray-Ban Meta’s design, with an AR display projected on the bottom portion of the right lens, according to the report.

Budget-friendly gadgets that are good


It’s a slower October than usual in the tech industry, thanks mostly to Google and Microsoft having held their typical fall hardware announcements earlier this year. Still, we’ve seen a fair number of companies reveal new devices in the last two weeks, while Amazon’s October Prime Day raged on. Whether you were busy shopping or watching Elon Musk talk up robotaxis and cybervans, the Engadget team continued to review recently (and not-so-recently) launched products. As usual, this bi-weekly roundup is here to help you catch up, though because I missed last week’s edition (as I was out on time off), the cadence is just a bit off.

From Meta’s Quest 3S VR headset and the DJI Air 3S drone, to Sony’s midrange suite of audio gear, these weeks have coincidentally been about the less premium, more affordable “un-flagships,” if you will. And it turns out you don’t have to throw chunks of your retirement savings at companies to get solid devices that are well worth the money.

by Billy Steele

Sony

Sony flexes its mobile audio muscle on the LinkBuds Fit, combining great sound with tons of features.

Pros

  • Tiny, comfy design
  • Surprisingly big sound
  • Lots of handy features
Cons

  • ANC performance is lacking
  • Battery life
  • Call quality is fine, but not great

$200 at Sony

The main thing I learned from Billy’s recent reviews of Sony and Bose headphones is that flagship, premium-level noise cancelation is making its way down to the midrange products. And I, for one, am happy about it. Looks like most of the things you’ll miss if you opt to save some money are features like spatial audio, head-tracking and stuff that’s supposed to be make for a more immersive, intuitive listening experience. Personally, I’m okay missing out on those things for now — I really just want decent sound, good voice quality and competent reduction of background noise.

Definitely check out Billy’s review of Sony’s LinkBuds trio of devices even if you’re not in the market for headphones, because he put his ears through literal pain to test the strangely designed earbuds for us. In fact, a rejected headline for the review was “Why am I in pain?” We salute you, Billy.

by Devindra Hardawar

Meta

The Meta Quest 3S delivers immersive virtual reality for just $300. It’s fast, comfortable to wear and it has a wealth of games and VR experiences to choose from.

Pros

  • Fast performance
  • Comfortable to wear
  • Excellent controllers
  • Large app library
Cons

  • Older Fresnel lenses lead to artifacts
  • No headphone jack
  • Average mixed reality cameras

$299 at Amazon

The Quest is arguably the industry’s leading VR headset for consumers, especially considering it costs a fraction of alternative options while offering a relatively high-quality immersive experience. Unlike the Apple Vision Pro, which costs $3,500, the new Quest 3S is much more palatable at $300. And, as Devindra observes in his review, it’s comfortable, fast and allows wearers to access Meta’s large library of VR apps and content. To quote Devindra, “It’s not Meta’s first $300 VR headset, but it’s still a tremendous accomplishment.”

by Steve Dent

DJI

DJI’s Air 3S is the company’s first consumer drone with LiDAR designed to make it safer to fly at night. 

Pros

  • Excellent image quality
  • LiDAR improves obstacle detection at night
  • ActiveTrack 360 produces cinematic tracking shots
  • Long battery life
  • Now has 42GB internal storage
Cons

  • Sometimes fails to detect small obstacles

$1,099 at DJI

Continuing in the trend of excellent products that cost less than usual, the DJI Neo is an excellent drone for just $200. Calling it the best $200 drone ever made may sound hyperbolic, but we truly can’t think of any other drone that cost the same that came close to delivering similar performance as the Neo. And though one of its main drawbacks, according to our reviewer Steve, is that it sounds like a banshee, I honestly think that would be a pro for me just for the pure comedy of it all. I never thought I’d be in the market for a banshee-sounding drone but here I am thinking $200 would be a fair price to pay for the aerial videos I could get, and I could easily lay upbeat, operatic singing over the footage anyway.

On the other end of the spectrum, Steve’s also tested the higher-end DJI Air 3S drone and Canon’s EOS R5 II, both of which will cost you a lot more money. The good news is that Canon’s EOS R5 II is now so much improved that it better takes on Sony’s rival offerings, while bringing intriguing features like eye-control autofocus. Sure, it’s still too early to be reliable. But if you have the money for these premium products, you’ll get to experience bleeding-edge tech, maybe part of the appeal is that they may not always work well.

by Dan Cooper

reMarkable / Engadget

reMarkable adds a color screen to its e-paper tablet, creating a best-in-class distraction-free writing experience.

Pros

  • Color is a welcome and useful addition
  • Backlight lets you work in dark environments
  • Vastly improved performance

$579 at reMarkable

I know Dan was a bit worried about giving a device as niche as the reMarkable Paper Pro such a high score, but after some discussion, we both agreed it fit. As it stands, the reMarkable Paper Pro is arguably the best e-paper writing tablet with a color screen, even after this week’s announcement of Amazon’s first color Kindle. Per Dan’s review, the Paper Pro not only levels up with the addition of color, but also got much faster and reliable in the process.

I particularly enjoy how much care has been taken to more seamlessly blend handwriting and typed text for a more natural note-taking and annotating experience. And though it falls outside this edition’s general theme of being a less-expensive product with premium specs, I’m still incredibly tempted to find ways to cough up the $600 or so needed to bag myself a Paper Pro with the accessories Dan recommends.

by Jessica Conditt

I’m not the most engaged gamer, but I love learning about games from Jess. She recently reviewed sci-fi mystery Phoenix Springs and dreamy platformer Neva, with pieces written so beautifully that I can’t help but itch to check out each title so I can feel the way she clearly does about them. Phoenix Springs even follows a technology reporter and covers my favorite genres (science fiction and mystery), while Neva’s art looks simply gorgeous. I don’t yet know if I’ll ever find the time to play new games, but these two have certainly been added to my list.

The Engadget team geeks out about a multitude of things, and as we get started on our holiday gift guide writing, we’re thinking about products we like for ourselves and our loved ones. Many of us love e-readers and writing tablets. I certainly have my eye on one of the new Kindles that were announced this week, particularly the new Kindle Scribe and possibly the Colorsoft. Those won’t be available until later this year, so for now I’ll keep hungrily eyeing the reMarkable Paper Pro. I’ll also be wondering what Amazon hardware chief Panos Panay has up his sleeves, other than cans of Diet Coke — something I learned he (like many people) enjoys when I got the chance to speak with him in an interview this week.

Many of us are millennials and have gigantic soft spots for retro gaming. Our executive editor Aaron Souppouris talked to the folks at Analogue, and you can read his beautifully constructed piece to learn more about the upcoming Analogue 3D.

This week, Apple sort of surprised us by announcing the new iPad mini, which has been tweaked to offer more storage and power with an A17 Pro processor. It’ll also support Apple Intelligence, of course, and though it seems like a minor update, some of my coworkers and I still love the idea of a small tablet. If only it had a better screen.

All of those recently announced things are on our review roster, alongside things like the Google TV Streamer and Samsung Galaxy S24 FE that launched in the last few weeks. As we approach November and all the holiday shopping that brings, hopefully the hardware launches properly slow down. We’re still anticipating an Apple event for Macs, and can still remember being surprised by the announcement of Humane’s AI Pin late last year. But for now, we’re chugging along while occasionally bathing in nostalgia and longing for a simpler time.

Utah judge blocks law preventing youth from accessing social media freely


On Tuesday, Chief US District Judge Robert Shelby granted a preliminary injunction to from limiting the social media usage of minors. Republican Governor Spencer Cox had signed the Utah Minor Protection in Social Media Act earlier in March. It was supposed to take effect on October 1, but the court’s decision to block the law is a victory for young social media users in Utah.

This isn’t the first time Utah’s governor has attempted to use among the youths in the state. Last year, he signed two bills that required parents to grant permission for teens to create social media accounts, and these accounts had limitations like curfews and age verification. He in March due to lawsuits challenging their legality.

Under the law, social media companies would have been forced to verify the age of all users. If a minor registers for an account, they are subject to various limitations. The content they share would be seen only by connected accounts. Additionally, minor accounts could not be searched for or messaged by non-followers or friends, effectively nonexistent to strangers.

The primary reason for the is due to NetChoice’s claim that the law constitutes a violation of the First Amendment. NetChoice is a trade association formed by tech giants such as X (formerly Twitter), Snap, Meta and Google. The association has managed to win in court battles and block similar laws entirely or in part in states like , and .

Threads is testing disappearing posts that expire after 24 hours


Threads is testing the option for users to put a 24-hour expiration timer on their posts, after which the post and all replies would disappear, Stories-style. A spokesperson confirmed to that the feature is being tested among a group of users after it was first spotted earlier this summer by developer . It sounds a lot like pre-X Twitter’s Fleets, . But, the ephemeral format could save habitual post-deleters some trouble.

It comes a few months after Instagram head Adam Mosseri shared that Threads was . That optional feature would let users designate a date for their posts to be hidden from the feed. But Threads users in the past have indicated that they of automatic archiving, and such a feature hasn’t yet shown up on a wider scale. Threads at the beginning of August, and recently introduced an for users — particularly those with large followings — to keep track of their account’s performance. Meta also said the option to schedule posts is on the way.