Google's Bard gained a handful of new features and functions Thursday in the chatbot AI's latest round of updates, including expanded linguistic knowledge, more nuanced response controls and the ability to respond with spoken word in addition to text. In all the AI can now converse in nearly four dozen languages. 

Users can now converse with the AI in Arabic, Chinese, German, Hindi and Spanish, among others as well as access the platform from more places on the planet, such as Brazil and "across Europe," Jack Krawczyk, Bard Product Lead, and Amarnag Subramanya, Bard's VP of Engineering, wrote in a blog post Thursday. "As we bring Bard to more regions and languages over time, we’ll continue to use our AI Principles as a guide, incorporate user feedback, and take steps to protect people’s privacy and data."

Bard now literally speaks. Users will have the option to either read or listen to the AI's generated responses, which Krawczyk and Subramanya believe will help immensely when users want to hear the correct pronunciation of words in those 40 newly-added languages. Users have also been afforded more robust controls over how friendly Bard will be with five distinct options for the AI's tone: simple, long, short, professional or casual. Those are only available on English-language requests for the moment but the company is already working to expand it out to more of the 40, "soon."

The chatbot also has some fancy new multimodal eyes, gaining the capacity to interpret images dropped into the chat through the prompt field. Faster and easier than uploading it as a document, users can request more information about the contents of the image or generate content like captions based on it. This also is currently English-only. 

Getting the information and code that Bard generates out of the chat window and into the hands of collaborators is no longer quite such a slog. Starting Thursday, users will be able to export Bard-generated Python code to Replit, in addition to Colab. They'll also be able to copy and share portions of individual chats with other users. The process of organizing and revisiting old conversations being streamlined as well with the addition of pinned conversations, which are what they sound like, and the ability to rename them.

Twitter has published an update on how its "Freedom of Speech Not Reach" moderation approach is working, and according to the company, it has seen some encouraging results. In April, the website started limiting the reach of tweets violating its hateful conduct policy and applying a label to them that reads: "Visibility limited: this tweet may violate Twitter's rules against hateful conduct." Apparently, Twitter has applied the label to more than 700,000 posts since then and has proactively prevented ads from appearing adjacent to those content. 

The company also said that the label reduces the reach of a post by 81 percent, thereby effectively limiting the visibility of posts that potentially exhibit hateful conduct. In addition, Twitter revealed in its update that more than one-third of users choose to delete labeled tweets themselves once they've been notified that they have violated the website's policy and only four percent of authors have appealed labels. 

The company charging for API access means most researchers studying hate speech can't independently verify these claims. But Twitter is clearly claiming that its approach has been effective so far. In fact, the website is pushing through with its plan to expand its labels and include more types of policy violations. According to its announcement, it will now also label and downrank posts that violate its Abusive Behavior and Violent Speech policies. Tweets that will be labeled in the coming weeks include posts with malicious content targeting individuals, those that encourage others to harass an individual or group of people, those that threaten to inflict physical harm on others, and tweets that encourage others to commit acts of violence or harm. 

The Federal Trade Commission isn't giving up on its attempt to halt Microsoft's pending $68.7 billion purchase of Activision Blizzard. The agency has appealed Judge Jacqueline Scott Corley's denial of its request for a preliminary injunction to temporarily stop the deal from going through.

The FTC has sued to prevent the merger from happening over antitrust concerns. An administrative trial is set to start in August, but the companies have a merger deadline of July 18th. The agency was concerned Microsoft and Activision would close their deal by then despite a UK regulator blocking the deal in that country.

Bloomberg first reported that the agency was considering an appeal against Corley's decision. The FTC told Engadget after Tuesday's ruling that it would announce its "next step to continue our fight to preserve competition and protect consumers" in the following days.

Corley ruled that, unless the FTC obtains an emergency stay from the Ninth Circuit Court of Appeals by 11:59PM PT on July 14th, a temporary restraining order that's currently preventing Microsoft and Activision from closing the deal will be dissolved. The restraining order was put in place until Corley made a decision on the preliminary injunction.

Meanwhile, after Corley's ruling, Microsoft, Activision Blizzard and the UK's Competition and Markets Authority said they agreed to pause their legal battle and see if they could reach a compromise. The CMA later clarified that although "merging parties don’t have the opportunity to put forward new remedies once a final report has been issued, they can choose to restructure a deal." It added that doing so could lead to a fresh merger investigation, which would likely delay the takeover beyond July 18th.

It's easy to look at every new macOS release and wonder where the upgrades actually are. Unlike iOS and iPadOS, Apple doesn't often push for bold and visible changes on the desktop (last year's Stage Manager was a surprising addition, but even that was brought over from iPadOS). Instead, most of the enhancements to the Mac experience are usually under the hood — don't forget, the company completely rebuilt macOS for its new Apple Silicon chips.

So it goes with macOS Sonoma, an update that doesn't truly make itself known until you see its new high-res video screensavers (something the Apple TV has done for years), or try to drag over a widget to your desktop. If widgets are a highlight in a new OS, you know it's a slow year.

This isn't a knock against Sonoma, exactly. It shouldn't be a surprise to anyone that Apple is saving its bolder UI concepts for iPhones, iPads, and the Vision Pro. Macs are computers that don't even have touchscreens yet! Playing it safe makes the most sense.

Photo by Devindra Hardawar/Engadget

So let's talk about screensavers. While not exactly Earth shattering, Apple's new aerial footage looks great in macOS Sonoma. It adds a dose of elegance to an already refined operating system — having footage from Hawaii or Patagonia pop up in the middle of the workday is like a mini-vacation. (Don't judge me, I know it sounds sad.)

I also liked being able to turn aerial screensavers into wallpapers. Doing so lets them play as videos on your lock screen, but once you log in, the video slows down and smoothly settles on a static image. I'm sure that's done to avoid motion sickness from some users, but I'm hoping Apple or some intrepid hacker unlocks the ability to let the video constantly play on your desktop. I'll risk a bit of squeamishness for more of that slick drone footage.

As for the widgets, well... they're widgets. On your desktop. It's a pretty straightforward concept. I'll admit, being able to quickly see the weather or an upcoming appointment under my windows is easier than popping open the notification tray. But it's not exactly a killer feature. But hey, if you really do love these things, at least you can easily bring over your iOS widgets. And if you also have an iPad, I'm sure it'll be helpful to have all of your widgety goodness across your devices.

Beyond those two features, Sonoma's upgrades are a bit more subtle. If you jump into a video chat, you can use hand gestures to kick off some fun special effects. A double thumbs up triggers fireworks behind you, while a peace sign surrounds you with floating balloons. They're certainly fun at first, but thankfully you can turn them off easily for more professional video chats.

Photo by Devindra Hardawar/Engadget

The new presenter overlay features, on the other hand, are a lot more useful. In supported screen sharing apps, like FaceTime or Zoom, you can place a small or large window of yourself on top of whatever window you're sharing. It's the sort of feature I'm surprised Apple didn't roll out sooner, as it can help to humanize boring video chats. Instead of just sharing a spreadsheet, for example, your colleagues can also see what you think of your data as you're discussing it.

Here are a few other Sonoma updates to keep an eye out for:

• Safari supports profiles for different use cases. For example, one profile could house all of the bookmarks and settings that you need for work.

• You can also turn any website into a web app with Safari. It's the perfect way to read Engadget without any distractions!

• Stickers and Memoji now synchronize with all of your Apple devices over iCloud.

• A new Game Mode will prioritize resources while gaming. I didn't notice any major differences with Apple Arcade titles during my testing, though.

• Autocorrect should be better too, but again, I didn't notice any major changes.

You'll want to act quickly if you're shopping for a truly stand-out TV set. Samsung's 2023 Frame TVs are on sale at Amazon for Prime Day, and the discounts are particularly steep for some models. The line now starts at $548 for the bedroom-friendly 32-inch model, but the sweetest deal is for the 55-inch model, which has dropped to a record low of $988 — 34 percent off. You won't get an add-on bezel at those prices, but the savings are large enough that you might not mind.

The appeal of Frame TVs is as clear as ever: it's a set that doubles as artwork when it's not in use. Mount it on your wall (the kit is thankfully included) and it blends into the decor using either professional images or your own. You'll usually have to subscribe to the Samsung Art Store or buy individual works if you want classics, but it could be worthwhile to have a Van Gogh or Vermeer in your living room. The 2023 models have a matte finish that cuts reflections and gives them a canvas-like appearance.

These aren't Samsung's highest-end TVs, so you won't get the crispness of an 8K panel or the vivid contrast of OLED. They still offer an accurate 4K picture, however, and they're full-featured sets with voice assistant control, a game mode and a wide range of apps. At these prices, they're easy choices if you either want a wall-mounted TV or were otherwise resigned to buying an ordinary screen.

Your Prime Day Shopping Guide: See all of our Prime Day coverage. Shop the best Prime Day deals on Yahoo Life. Follow Engadget for the best Amazon Prime Day tech deals. Learn about Prime Day trends on In the Know. Hear from Autoblog’s car experts on must-shop auto-related Prime Day deals and find Prime Day sales to shop on AOL, handpicked just for you.

A Congressional investigation concluded that several tax prep providers shared sensitive filing data with Meta and Google. It follows a 2022 report from The Markup highlighting the practice in which TaxSlayer, H&R Block and TaxAct used Meta’s Pixel tracking tool to harvest info like filing status, approximate adjusted gross income, refund amount, names of dependents and which text-entry fields users clicked on. Meta is already facing a lawsuit connected with the initial reporting.

The panel sent the conclusions to the IRS, FTC, DOJ and Treasury Inspector General for Tax Administration (TIGA), urging the agencies to investigate and prosecute if applicable. “Big Tax Prep has recklessly shared tens of millions of taxpayers’ sensitive personal and financial data with Meta for years, without appropriately disclosing this data usage or protecting the data, and without appropriate taxpayer consent,” the report reads. “The findings of this report reveal a shocking breach of taxpayer privacy by tax prep companies and by Big Tech firms that appeared to violate taxpayers’ rights and may have violated taxpayer privacy law.”

The review found the Meta Pixel tracker also gathered data about “whether taxpayers had visited pages for many revealing tax situations, such as having dependents, certain types of income (such as rental income or capital gains), and certain tax credits or deductions.” In addition, it transmitted the full names, email, country, state, city, zip codes, phone numbers and gender as hashed values. The information was also collected from taxpayers using TaxAct’s Free File service — which is through a partnership with the IRS.

Congressional investigators listed in the report include Senators Elizabeth Warren (D-MA), Ron Wyden (D-OR), Richard Blumenthal (D-CT), Tammy Duckworth (D-IL), Bernie Sanders (I-VT) and Sheldon Whitehouse (D-RI) and Rep. Katie Porter (D-CA).

“The tax prep firms were shockingly careless with their treatment of taxpayer data,” the investigation concluded. “They indicated that they installed the Meta and Google tools on their websites without fully understanding the extent to which they would send taxpayer data to these tech firms, without consulting with independent compliance or privacy experts, and without full knowledge of Meta’s use of and disposition of the data.” The panel also chided Meta and Google for acting “with stunning disregard for taxpayer privacy.”

The report cites laws that say, “a tax return preparer may not disclose or use a taxpayer’s tax return information prior to obtaining a written consent from the taxpayer,” while mentioning that the tax prep companies failed to do that. Although tax-filing companies can legally hand data to “auxiliary service providers in connection with the preparation of a tax return,” the panel said Meta and Google don’t meet that definition since the tracking was used for advertising. Violations can lead to fines of up to $1,000 per instance (likely pocket change for these companies) and up to a year in prison.

Microsoft is doing more to tackle toxicity in multiplayer Xbox games. The company is introducing a feature that allows Xbox Series X/S and Xbox One players to capture a 60-second video clip of abusive or inappropriate voice chat and submit it for moderators to review.

"This feature is purpose-built to support the broadest arena of in-game interactions between players and works across thousands of games that offer in-game multiplayer voice chat, including Xbox 360 backward-compatible titles," Xbox Player Services corporate vice-president Dave McCarthy wrote in a blog post.

Microsoft designed the tool for both ease of use and to minimize the impact on gameplay. When you capture a clip for reporting, it will stay on your Xbox for "24 online hours." You'll have the option to submit it right away or wait until you finish your gaming session. You'll get a reminder before the 24 hours are up. If you opt not to report the clip, your Xbox will automatically delete it.

No one else will have access to the clip unless you submit it. "Xbox is not saving or uploading any voice clips without you, the player, choosing to start the reporting process," McCarthy said. Any clips you capture through the tool will not appear in your recent captures and it won't be possible to download, share or modify them. These clips will only be used for moderation purposes. After the safety team has reviewed your report, you'll receive a notification to tell you whether it took action against an abusive player.

An Xbox spokesperson told Engadget the safety team will use "a range of moderation tools that leverage AI and human moderators" to analyze the clips. Moderators will review audio and video to determine if someone has violated the community standards.

At the outset, the reactive voice reporting system allows a player to report up to three individuals at once. "If a moderator cannot determine who was talking at a given time and associate it with the reported Xbox Live player, the report will be closed as unactionable, no enforcement action will be taken and the captured video will be deleted within 90 days," the spokesperson said.

The advent of cross-platform play muddies the waters here too. The safety team won't take any action in cases of inappropriate voice chat from those who are on other platforms. "The announced reactive voice moderation feature is specifically for reporting Xbox players to the Xbox Safety Team," the spokesperson noted.

It's encouraging to see Xbox tackle the problem of toxic voice chat at a platform-wide level. The PlayStation 5 has had a similar feature since it debuted in 2020.

Several studios have adopted similar approaches in their own games. In 2021, Riot said it would record Valorant voice communications but only listen to them when a report was filed. It started testing the system last July.

Ahead of Overwatch 2 going live last October, Blizzard said it would automatically transcribe a recording of a match's voice chat following a report from a player. The company's chat review tools analyze the transcription for signs of abusive behavior and both the recording and text file are deleted. (Of note, Blizzard and Overwatch 2 may soon be Microsoft properties.)

To begin with, Xbox's voice reporting feature will be available for Alpha and Alpha-skip Xbox Insiders in the English-speaking markets of the US, Canada, Great Britain, Ireland, Australia and New Zealand. Microsoft hopes that insiders will provide feedback to help improve this feature. It plans to continue investing in voice moderation and to support more languages. Xbox will share data and updates on voice chat moderation in its bi-annual transparency report.

The immensely popular online game creation platform/thingamajig Roblox is coming to VR, thanks to the Meta Quest platform. Meta CEO Mark Zuckerberg made the announcement, which was followed by an official Roblox blog post. A beta version of the app launches in just a few weeks, according to Zuckerberg, which will be accessible via the Quest platform’s dedicated App Lab.

The beta will be available for the Quest 2 and Quest Pro headsets, so OG Quest owners are out of luck. The recently-announced Quest 3 wasn’t mentioned, but that’s likely because it won’t be out in time for the beta release. A finished version of the game, however, would make for an extremely compelling launch title for the Quest 3, as Roblox boasts more than 66 million daily users across multiple platforms.

To that end, this is going to be a fully cross-platform title, so anything you interact with or make on your phone or console should be accessible in VR, though it depends on the developer. The Roblox team is actively encouraging developers to make experiences in VR with a forthcoming toolset accessible via the Creator Dashboard. You’ll soon see VR as a new device type on this dashboard, along with desktop, mobile, tablet and console.

Meta recently lowered the minimum age to operate a Quest headset from 13 to ten, but this change will not impact Roblox. You have to be at least 13 to use the app in VR. That’s probably for the best, given the addictive nature of the title and associated child safety concerns.

The Perseverance Rover has found evidence of organic compounds in the Jezero Crater on Mars. Although this isn’t the smoking gun proving once and for all that Mars once hosted life — these compounds could have also developed in nonbiological ways — the results hint at surprisingly complex organic conditions for the “key building blocks for life” on Earth’s neighbor. The study was published in Nature.

The Perseverance Rover, the first to explore the Jezero Crater, has been investigating the area since February 2021. Researchers believe the basin once housed an ancient lake, including a delta from a river that once flowed into it. It’s one of the most likely regions to reveal leftover signs of life on Mars.

Organic molecules like those observed in the Jezero Crater contain carbon and often hydrogen atoms. They’re the core components of life as we know it on Earth, although they can also develop abiologically. “They are an exciting clue for astrobiologists since they are often thought of as building blocks of life,” paper co-author Joseph Razzell Hollis, a postdoctoral fellow at London’s Natural History Museum, said toNewsweek. “Importantly, they can be created by processes not related to life as we know it, and so organic molecules are not evidence of life on their own without sufficient extra evidence that cannot be explained by nonbiological — or abiotic — processes.”

The rover observed the compounds using an instrument — the Scanning Habitable Environments with Raman and Luminescence for Organics and Chemicals (SHERLOC) — that maps organic molecules and minerals on rock surfaces. Significantly, it found organic materials in all ten targets it observed on the crater floor. “Our results support observations by previous robotic missions to Mars that the Red Planet was once rich in organic material, compounds made primarily of carbon and hydrogen, and that some of that organic material can still be detected billions of years later,” co-author Joseph Razzell Hollis, a London-based astrobiologist, toldGizmodo. “Each detection, each observation, gives us a little bit more information that brings us closer to understanding the history of Mars and whether it could have supported life in the past.”

Now that the researchers have observed the molecules, they’ll need a better look at them in Earthbound labs to draw further conclusions about their origins. “If these samples are returned to terrestrial laboratories, a more diverse suite of tools can be used to study the samples, including at higher spatial resolution and with much greater specificity and sensitivity,” the authors wrote. They’ll have to wait for the Mars Sample Return (MSR) mission, which isn’t expected to launch from Earth until at least the late 2020s. Still, the trip should be worth the wait. “So far, the only Martian rocks we’ve ever been able to study on Earth have been meteorites. Getting our hands on intact Mars rocks, carefully stored and protected from contamination, will be invaluable to planetary science,” Razzell Hollis told Newsweek.

Chipotle isn't done experimenting with robots that could make life easier for restaurant staff. The chain has teamed up with robotics company Vebu to test a prototype Autocado robot (yes, that's the name) that cuts, cores and peels avocados to be used in guacamole. Kitchen workers only have to fill the bot with up to 25lbs of ripe avocados and choose a size. After the processing is done, they collect the fruit in a bowl, add remaining ingredients and start mashing.

As Autocado improves, Chipotle expects to cut guacamole prep time in half — no mean feat when it usually takes 50 minutes to make a batch. That could ensure a steady supply of guac while eliminating some of the drudgery for employees. The restaurant also believes the robot could reduce food waste (and thus costs) through more efficient processing.

Autocado is currently limited to a testing center in Irvine, California. This is the start of a "long term partnership" with Vebu, Chipotle adds. The two haven't said what comes next, but the focus will be on collaborative robots that "drive efficiencies" and "ease pain points" for workers. Last year, Chipotle began testing Miso Robotics' Chippy robot to aid in making tortilla chips.

Chipotle is far from alone. McDonald's and other restaurants are trying AI, robots and other automated solutions to streamline their experiences. These moves theoretically help employees concentrate on serving customers and other tasks where humans are still preferable. Of course, there are also lingering concerns among critics that restaurants might use this to automate people out of jobs. They may choose to shrink headcounts and cut costs instead of easing the workload. That isn't guaranteed to happen with Autocado when humans are still necessary, but the long-term future isn't quite so clear.