You can grab a four-pack Apple AirTag for $85 again, if you missed your chance to do so last time it was on sale for the same price. Previously, you could get the discounted AirTags on Amazon and Walmart — now, it's Adorama that's selling the four-pack bundle for $14 less its usual price. It's a great deal if you're looking to keep track of several belongings, such as keys, wallets or the suitcases you're checking in whenever you fly. The bundle also remains unavailable on Apple's Amazon storefront, so here's your chance to grab it at a discount.

AirTags are, perhaps, the best tracking device you could use if you have an iPhone. They come with Apple's U1 ultra-wideband chip that enables precise tracking for more recent iPhones, so you could more easily find where your item is with the distance and directions you'll see on your phone. You could also make the AirTag play sounds through the Find My App to help you look for your items. If whatever you're tracking is out of range, a network of nearby Apple devices can detect the Bluetooth signals AirTags give out and upload their location to iCloud, allowing you to find your belonging. 

One downside to using AirTags is that they don't come with a built-in keyring or attachment. They're coin-like devices with no holes or hooks either. To help with that issue, we put together a list of accessories you could use to attach the AirTag to your items, including cases, holders and mounts with clips and straps. 

Follow @EngadgetDeals on Twitter and subscribe to the Engadget Deals newsletter for the latest tech deals and buying advice.

After announcing the deal last year, Intel will no longer acquire Tower Semiconductor for $5.4 billion, the company announced in a press release. It was unable to "obtain in a timely manner the regulatory approvals required under the merger agreement" it wrote — specifically in China, according to Bloomberg. Tower produces various types of chips for clients across multiple industries, and Intel made the acquisition to expand its foundry business and better compete with rivals like Taiwanese giant TSMC. 

Tower has seven fabrication facilities (located in Israel, Italy, the US and Japan) that build 6-inch, 8-inch and 12-inch chip wafers. While the company doesn't manufacture cutting edge mobile and other process, its clients don't necessarily need the latest technology. Instead, Tower focuses on reliably manufacturing large volumes of chips for automakers, equipment manufacturers, medical industries and others. 

Before announcing its Tower acquisition, Intel was reported to be in talks to purchase the much larger chip manufacturer and AMD spinoff GlobalFoundries for around $30 billion. Intel launched its foundry services as a separate business unit back in 2021, committing $20 billion to build two Arizona factories. It also revealed plans to build a massive semiconductor facility in Ohio designed to become "the largest silicon manufacturing location on the planet."

Intel said its still executing its roadmap "to retain transistor performance and power performance leadership by 2025," with the aim of becoming the second-largest global external foundry by 2030. "Our respect for Tower has only grown through this process, and we will continue to look for opportunities to work together in the future." As part of its merger agreement, Intel will pay a termination fee of $353 million to Tower. 

Tweetdeck (now called X Pro), a version of Twitter (X) used by social media pros, journalists and others, has become part of the paid Blue subscription, TechCrunch has reported. Multiple users reported that they were prompted to subscribe when attempting to log on to the service. "Blue subscribers with a verified phone number will get a blue checkmark once approved," the message reads. 

The news is not unexpected, as code spotted earlier this year by researcher Jane Manchun Wong indicated that X Pro was on its way to becoming a paid app. However, the change only appears to apply to the new version, but not the older (better) version, at least until it's deprecated. 

Engadget

X has been testing the new version since last year, with changes like "a full Tweet composer, new advanced search features, new column types, and a new way to group columns into clean workspaces," the company said at the time. Some power users aren't loving it, though, as it's more difficult to clear columns and perform other common actions. 

The company has previously pondered a TweetDeck subscription service, asking users in 2017 if they'd be willing to pay up to $20 per month for a "more advanced TweetDeck experience." The new action is part of Elon Musk's attempt to monetize the app with Blue subscriptions, but the service only had about 700,000 subscribers as of last June. Musk has also vowed to make X an "everything app" used for content creation, payments and shopping on top of social networking. 

Despite having trouble obtaining former President Donald Trump's records from Twitter (now X), the special counsel investigating the events of January 6th, 2020 was ultimately able to get his hands on vast amounts of information from the website. Based on the newly unsealed court filings (PDF) shared by Politico, though, it took a a lot of back and forth and a massive amount of effort before the counsel was able to secure the data requested from Twitter. 

Apparently the search warrant required Twitter to produce all the tweets the @realdonaldtrump created, drafted, liked or retweeted, including any post that he eventually deleted. The website also had to turn over DMs "sent from, received by, stored in draft form in, or otherwise associated with" @realdonaldtrump, as well as all other accounts related to it that the former president might have used on the same device. All devices used to log in to @realdonaldtrump had to be noted and given to the counsel, as well, along with the account's privacy settings and IP address history from October 2020 to January 2021. In addition, Twitter was required to hand over all records of searches done by @realdonaldtrump and the account's location information from the same time period.

Throughout the transcript of the proceedings, you'll see how the prosecutors insisted that Twitter had to adhere to a nondisclosure order that prohibits the company from telling Trump about the warrant. They explained that they had evidence that disclosing the warrant could jeopardize the probe, warning the court that "there would be actual harm and concern for the investigation, for the witnesses going forward." Meanwhile, Twitter's side argued that some of the information the investigators were requesting could be covered by executive privilege, which the prosecutors and judge questioned, since they didn't deem it likely that Trump discussed official government matter via DMs. 

As Politico reports, US District Judge Beryl Howell notably called attention to Twitter's efforts to give Trump advance notice about the search warrant. She had called the action "extraordinary" and something the company had never done for other users. The judge wondered whether Elon Musk was trying to "cozy up" to the former president and make him feel "particularly welcomed" on the platform. Trump's Twitter account had already been reinstated after Musk took over the website, but he has yet to start posting again. 

If you have an older Samsung foldable phone, tablet or watch, you may soon see some new features trickling down from the latest devices. The company announced today that the One UI 5.1.1 update — Samsung’s latest custom UI sitting atop Android — will roll out soon, bringing enhancements from the Galaxy Z Fold 5, Z Flip 5, Tab S9 series and Watch 6 series to pre-2023 gear. The features will support most of Samsung’s foldable phones and its recent tablets and smartwatches.

The One UI 5.1.1 update will support a long list of recent Samsung devices, but some may have to wait. It will roll out to Samsung’s stable of foldable phones, scheduled to reach the Galaxy Z Flip 4 and Z Fold 4 “starting this month,” while the Z Fold 3, Z Flip 3, Z Fold 2 and Z Flip will receive it “sequentially” later. Tablet multitasking updates will arrive for the Galaxy Tab S8, Tab S8+, Tab S8 Ultra, Tab S7, Tab S7+, Tab S7 FE, Tab S6 Lite, Tab A8, Tab A7 Lite, Tab Active 3 and Tab Active 4 Pro. Finally, new smartwatch features will support the Galaxy Watch 5, Watch 5 Pro, Watch 4 and Watch 4 Classic.

Samsung’s foldable phone updates in One UI 5.1.1 make it easier to show or hide the Flex Mode Panel (the controls appearing on the bottom half of the screen when the phone is semi-folded). “When folding your phone in Flex Mode, the panel’s icon will float up the screen, allowing you easy access,” the company wrote. “Enjoy a more personalized Flex Mode Panel experience by simply dragging and dropping the icon on the toolbar.” In addition, Samsung says the media play bar button (now including 10-second fast forward and rewind buttons) will be more prominent and simpler to find when the Flex Mode Panel activates. You’ll also be able to press down on the timeline to scrub through media more efficiently.

Cherlynn Low / Engadget

Samsung is tweaking multitasking on recent tablets and foldables, making switching to multi-window from pop-up window modes easier by long-pressing the handle. It also adds two-handed drag and drop, which lets you hold your finger down on a document from one app, use your other hand to open something else and then drag the file onto it for sharing or storage. The company also says the Z Fold 4 and Tab S8’s taskbar will be faster and include more recent apps. “Now, the Taskbar enables lightning-fast task switching with access to up to four recent apps,” Samsung said.

On the smartwatch front, the company is adding sleep-tracking upgrades from the Galaxy Watch 6 to older wearables. The changes include “a more intuitive UI” with the additions of Sleep Score Factors, Sleep Consistency and Sleep Animal Symbols. Sleep Coaching will also be available on the watch without looking at your paired phone. The Galaxy Watch can also automatically adjust SmartThings home devices for a more rest-inducing ambiance at bedtime.

Sam Rutherford / Engadget

Smartwatches will also receive fitness updates. The Watch 6’s Personalized Heart Rate Zone feature trickles down; it analyzes your physical abilities and “defines five optimal running intensity levels to help users set their own goals based on their ability — from burning fat to high intensity workout.” The Galaxy Watch 5 and Watch 4 series will also add Irregular Heart Rhythm Notifications in the update, which could be a lifesaver for some.

Meanwhile, Track Run, which (as its name suggests) records your runs on a track, also makes its way from the Watch 6. The Galaxy Watch 5 Pro will have access to a GPX File Database for downloading map data to your watch. Finally, One UI 5.1.1. lets you control the Flip 4’s camera with a paired Galaxy Watch. A new quick-access button will let you use your watch as a remote shutter, and it also lets you switch camera modes and control zoom levels.

Astronomers have used advanced instruments to calculate a more accurate age of Maisie’s galaxy, discovered by the James Webb Space Telescope (JWST) in June 2022. Although the star system isn’t quite as old as initially estimated, it’s still one of the oldest recorded, from 390 million years after the Big Bang — making it about 13.4 billion years old. That’s a mere 70 million years younger than JADES-GS-z13-0, the (current) oldest-known system.

A team led by the University of Texas at Austin astronomer Steven Finkelstein discovered the system last summer. (The name “Maisie’s galaxy” is an ode to his daughter because they spotted it on her birthday.) The group initially estimated that it was only 290 million years after the Big Bang, but analyzing the galaxy with more advanced equipment revealed it’s about 100 million years older than that. “The exciting thing about Maisie’s galaxy is that it was one of the first distant galaxies identified by JWST, and of that set, it’s the first to actually be spectroscopically confirmed,” said Finkelstein.

The spectroscopic confirmation came courtesy of the JWST’s Near InfraRed Spectrograph (NIRSpec) conducted by the Cosmic Evolution Early Release Science Survey (CEERS). The NIRSpec “splits an object’s light into many different narrow frequencies to more accurately identify its chemical makeup, heat output, intrinsic brightness and relative motion.” Redshift — the movement of light towards longer (redder) wavelengths to indicate motion away from the observer — held the key to more accurate dating than the original photometry-based estimate. The advanced tools assigned a redshift of z=11.4 to Maisie’s galaxy, helping the researchers settle on the revised estimate of 390 million years after the Big Bang.

ASSOCIATED PRESS

The astronomers also examined CEERS-93316, a galaxy initially estimated at 235 million years pre-Big Bang — which would have made it astonishingly old. After studying this system, it revealed a redshift of z=4.9, which places it at a mere one billion years after the Big Bang. The first faulty estimate about CEERS-93316 was understandable: The galaxy emitted an unusual amount of light in narrow frequency bands associated with oxygen and hydrogen, making it appear bluer than it was.

Finkelstein chalks up the miss to bad luck. “This was a kind of weird case,” he said. “Of the many tens of high redshift candidates that have been observed spectroscopically, this is the only instance of the true redshift being much less than our initial guess.” Finkelstein added, “It would have been really challenging to explain how the universe could create such a massive galaxy so soon. So, I think this was probably always the most likely outcome, because it was so extreme, so bright, at such an apparent high redshift.”

The CEERS team is now evaluating about 10 more systems that could be older than Maisie’s galaxy.

Only a few months after pulling back the curtain on its Revuelto plug-in hybrid, Lamborghini is just about ready to debut its first fully electric supercar. The automaker offered a glimpse of the vehicle's upper body in a teaser image, albeit in silhouette form.

You won't have to wait long to learn more about this mysterious EV, though. Lamborghini will show off what's expected to be a concept on August 18th, during Monterey Car Week. What we do know is that it will be an all-new fourth model in the lineup and not a version of an existing model with an electric powertrain.

Don't expect this to be the finished product, though. Lambo's first fully electric vehicle could still be years away from production as it only promised to release such a model by 2030. Like many other automakers, Lamborghini is on the path toward full electrification, but it's likely that it will take the brand several more years to reach that point.

Despite the US ethos that you’ll be innocent until proven guilty in a court of law, law enforcement finding an excuse to search your digital devices only requires a presumption of wrongdoing. The tech to do this already exists, and murky legislation lets it happen, speakers from the Legal Aid Society said at DEF CON last Friday.

“Technically and legally there's not much really truly blocking the government from getting the information they want if they want it,” Allison Young, digital forensics analyst at The Legal Aid Society, told Engadget. It’s easy, too. Without picking up any new skills or tools, Young was able to find sensitive data that could be used to, for example, prosecute someone being targeted for getting an abortion as it becomes increasingly illegal across the country.

The problem isn’t just the state of local law either, but it’s embedded in the Constitution. As Diane Akerman, digital forensics attorney at the Legal Aid Society explained, the Fourth Amendment hasn’t been updated to account for modern problems like digital data. The Fourth Amendment intends to protect people from “unreasonable searches and seizures” by the US government. This is where we get legal protections like warrants, where law enforcement needs court approval to look for evidence in your home, car or elsewhere.

Today, that includes your digital belongings too, from your phone to the cloud and beyond, making way for legal loopholes as tech advancements outpace the law. For example, there’s no way to challenge a search warrant prior to it being executed, Akerman said. For physical evidence that makes some sense because we don’t want someone flushing evidence down a toilet.

That’s not how your social media accounts or data in the cloud work though, because those digital records are much harder to scrub. So, law enforcement can get a warrant to search your device, and there’s no process to litigate in advance whether the warrant is appropriate. Even if there’s reason for the warrant, Akerman and Young showed that officers can use intentionally vague language to search your entire cell phone when they know the evidence may only be in one account.

“You litigate the issues once they already have the data, which means cat is out of the bag a lot of the time and even if it's suppressed in court, there's still other ways it can be used in court,” Akerman said. “There's no oversight for the way the government is executing words on digital devices.”

The issue only exacerbates across the third-party apps you use. According to the Fourth Amendment, if you give your information to a third party you’ve lost any sense of privacy, Akerman said. The government can often very easily get information from the cloud because of that, even if it’s not entirely relevant to the case. “You would be furious if police busted down your door and copied five years of texts for you walking out on a parking ticket five years ago, it's just not proportional,” Young said.

There are no easy ways for an individual to better protect themselves from these searches. On a case by case basis, there are ways to lock down your device, but that changes with every update or new feature, Young said. Instead, both speakers pushed to put the onus back on the systems and structures that uphold this law, not the individuals affected by it.

“I live in a world where I have to opt out of modern society to not have other people housing my data in some way,” Akerman said. “The question really should be like, what responsibility do those people have to us, since they have made us into their profit, rather than forcing me to opt out in order to protect myself?”

It appears that Elon Musk is, once again, punishing websites run by his perceived enemies. The website formerly known as Twitter seems to be interfering with links to The New York Times, Mastodon, Bluesky, Threads and Substack to make them load noticeably slower.

Clicking links to those websites from within X causes a significant delay in loading the web pages compared with navigating to the links from another browser or website. The delay, which seems to be about 5 seconds, was noticed by users on the Hacker News forum and later reported byThe Washington Post.

It appears to be widespread and only affecting a handful of high-profile websites that Musk has publicly attacked in the past. The user who first reported the issues on Hacker News estimated the inexplicable delay first cropped up on August 4th, noting that it’s “so consistent it's obviously deliberate.”

For now, it’s unclear if there is a cause for the delay other than Musk’s personal grievances with these companies. But the lag only seems to be affecting websites that Musk has previously retaliated against. He previously blocked links to Substack, Mastodon, Threads and other competitors. Musk has also publicly attackedThe Times, revoking the publication’s verified status shortly before the rollout of Twitter Blue (now called X Premium). The owner of X is also currently feuding with Meta CEO Mark Zuckerberg over his launch of Twitter competitor, Threads.

As usual, X didn’t respond to a request for comment. The company has largely stopped replying to press inquiries since Elon Musk fired its communications staff shortly after taking over the company.

Content moderation has been one of the thorniest issues on the internet for decades. It's a difficult subject matter for anyone to tackle, considering the subjectivity that goes hand-in-hand with figuring out what content should be permissible on a given platform. ChatGPT maker OpenAI thinks it can help and it has been putting GPT-4's content moderation skills to the test. It's using the large multimodal model "to build a content moderation system that is scalable, consistent and customizable."

The company wrote in a blog post that GPT-4 can not only help make content moderation decisions, but aid in developing policies and swiftly iterating on policy changes, "reducing the cycle from months to hours." It claims the model can parse the various regulations and nuances in content policies and instantly adapt to any updates. This, OpenAI claims, results in more consistent labeling of content.

"We believe this offers a more positive vision of the future of digital platforms, where AI can help moderate online traffic according to platform-specific policy and relieve the mental burden of a large number of human moderators, " OpenAI's Lilian Weng, Vik Goel and Andrea Vallone wrote. "Anyone with OpenAI API access can implement this approach to create their own AI-assisted moderation system." OpenAI claims GPT-4 moderation tools can help companies carry out around six months of work in about a day.

It's been well-documented that manually reviewing traumatic content can have a significant impact on human moderators' mental health, particularly when it comes to graphic material. In 2020, Meta agreed to pay more than 11,000 moderators at least $1,000 each in compensation for mental health issues that may have stemmed from reviewing material that was posted on Facebook.

Using AI to lift some of the burden from human reviewers could be greatly beneficial. Meta, for one, has been employing AI to help moderators for several years. Yet OpenAI says that, until now, human moderators have received help from "smaller vertical-specific machine learning models. The process is inherently slow and can lead to mental stress on human moderators."

AI models are far from perfect. Major companies have long been using AI in their moderation processes and, with or without the aid of the technology, still get big content decisions wrong. It remains to be seen whether OpenAI's system can avoid many of the major moderation traps we've seen other companies fall into over the years.

In any case, OpenAI agrees that humans still need to be involved in the process. “We’ve continued to have human review to verify some of the model judgements,” Vallone, who works on OpenAI's policy team, told Bloomberg.

"Judgments by language models are vulnerable to undesired biases that might have been introduced into the model during training. As with any AI application, results and output will need to be carefully monitored, validated and refined by maintaining humans in the loop," OpenAI's blog post reads. "By reducing human involvement in some parts of the moderation process that can be handled by language models, human resources can be more focused on addressing the complex edge cases most needed for policy refinement."