Given a choice between settling for pathetically slow internet speeds from AT&T or paying Comcast $50,000 to expand to his rural home, Michigan resident Jared Mauch chose option "C": starting up his own fiber internet service provider. Now, he's expanding his service from about 70 customers to nearly 600 thanks to funding aimed at expanding access to broadband internet, Ars Technica has reported. 

Last year, the US government's Coronavirus State and Local Fiscal Recovery Funds allocated $71 million to Michigan's Washtenaw county for infrastructure projects, with a part of that dedicated to broadband expansion. Mauch subsequently won a bid to wire up households "known to be unserved or underserved based on [an] existing survey," according to the RFP.

"They had this gap-filling RFP, and in my own wild stupidity or brilliance, I'm not sure which yet, I bid on the whole project [in my area] and managed to win through that competitive bidding process," he told Ars. 

He'll now need to expand from 14 to about 52 miles of fiber to complete the project, including at least a couple of homes that require a half mile of fiber for a single house. That'll cost $30,000 for each of those homes, but his installation fees are typically $199.

Customers can choose from 100Mbps up/down internet speeds for $55 per month, or 1Gbps with unlimited data for $79 a month. The contract requires completion by 2026, but he aims to be done by around the end of 2023. He's already hooked up some of the required addresses, issuing a press release after the first was connected in June, with a local commissioner calling it "a transformational moment for our community." 

Running an ISP isn't even Mauch's day job, as he normally works as an Akamai network architect. Still, his service has become a must in the region and he even provides fiber backhaul for a major mobile carrier. "I'm definitely a lot more well-known by all my neighbors... I'm saved in people's cell phones as 'fiber cable guy,'" he said. Check out the full story at Ars Technica. 

You know that highlighted piece of text at the very top of a Google search results page when you look up a piece of information? That's called a "featured snippet," and it's meant to provide you with a quick answer to your query. Now, Google is making sure that the information it highlights is reliable and accurate by using its latest AI model, the Multitask Unified Model, so that Search can now look for consensus when deciding on a snippet to feature.

Google's Search AI can now check snippet callouts — those are the information with larger fonts that serve as heading for featured snippets — against other high-quality sources online. It can figure out if there's a general consensus for that callout, even if sources use different words or concepts to describe the same fact or idea. Google says this "consensus-based technique has meaningfully improved the quality and helpfulness of featured snippet callouts."

Google featured snippet

But for some queries, such as those with false premises, displaying features snippets isn't the best way to deliver information. To address that issue, Google tweaked its Search AI so that this particular update reduces the triggering of snippets for those types of queries by 40 percent.  

Google is now also making its "About this result" tool more accessible. That's the panel that pops up when you click on the three dots next to a result, showing you details about the source website before you even visit. Starting later this year, it will be available in eight more languages, including Portuguese, French, Italian, German, Dutch, Spanish, Japanese and Indonesian. It's adding more information to the tool starting this week, as well, including how widely a publication is circulated, online reviews about a company, or whether a company is owned by another entity. They're all pieces of information that could help you decide whether a particular source is trustworthy. 

Finally, in case Google's AI has determined that the overall results for a search query may have questionable quality, the results page will now display a content advisory. "It looks like there aren't many great results for this search," the advisory will say, telling you to check the source you're looking at or to try other search terms. It could help you stay alert and be on the lookout for potential fake information while checking the results the website had presented.

Disney+ isn't done raising prices. As part of its third quarter earnings report, Disney revealed that it's hiking the price of the ad-free service in the US to $11 per month, $3 more than today, on December 8th. If you want to keep the same price, you'll have to subscribe to the ad-supported tier launching the same day. In other words, the ad-backed plan won't really be cheaper — you'll just have to pay more to keep the uninterrupted experience you already have.

The media giant also said it would raise the price of ad-free Hulu by $2 to $15 per month on October 10th. If you can accept ads, you'll also pay $8 per month instead of today's $7. A $10 monthly outlay provides both Disney+ and Hulu with ads. A bundle offering ad-free Disney+, ad-supported ESPN+ and its Hulu counterpart is climbing by a dollar to $15 per month, but you'll dip to $13 per month if you're willing to tolerate commercials across all three. You'll have to pay $20 per month to get the trio without any sales pitches.

Disney wasn't shy about the reason for the price hikes. Although it added 14.4 million Disney+ subscribers during the quarter (for a total of 221 million across all services), the operating losses for its streaming-oriented division surged from $293 million a year ago to nearly $1.1 billion. The production costs for Disney+ and Hulu are soaring, and Disney wants to make that money back.

The performance contrasts sharply with a key rival. While Netflix is prepping its own ad-driven plan, it's currently losing customers — it's counting on advertising to return growth where Disney is simply hoping to make a profit. As rough as its finances might be, Disney+ is in a stronger position.

Starlink can't count on a flood of government subsidies to help expand its satellite internet service. The FCC has rejected the SpaceX unit's bid to receive $885.5 million in aid through the Rural Digital Opportunity Fund. The broadband provider "failed to demonstrate" that it could deliver the claimed service, according to a statement.

FCC chair Jessica Rosenworcel said Starlink had "real promise," but suggested her agency couldn't justify 10 years of subsidies for "developing technology" that requires a $600 satellite dish. She added that the FCC needed to make the most of "scarce" funding for broadband expansion.

SpaceX won its bid in December 2020 through an auction. At the time, it said it would use the subsidies to serve 35 locations. It also promised prices in sync with terrestrial broadband, and that it would meet "periodic" service buildout requirements.

LTD Broadband, a fixed wireless provider, netted over $1.3 billion in that auction and also lost its bid today. That company was not "reasonably capable" of deploying the required internet service after it lost qualifying statuses in seven states, the FCC said.

We've asked SpaceX for comment. The denial isn't a fatal blow to the company's plans, but it makes clear that Starlink will have to rely on its own funding if it's going to expand as outlined in 2020. The FCC's move might also serve as a warning to other would-be fund recipients. While the Commission is eager to improve rural broadband, it won't grant money to internet providers without some close scrutiny.

Facebook's crackdown on hate speech apparently has room for improvement. As The Washington Postexplains, the non-profit watchdog Tech Transparency Project (TTP) has published a study indicating that white supremacist groups still have a significant presence on the social network. Over 80 of these racist organizations have a presence on Facebook, some of which the company has already labeled as "dangerous organizations" it normally bans. Researchers found 119 pages and 20 groups, including 24 pages Facebook auto-generated when users listed white supremacist groups as employers or interests.

Searches were also problematic, according to the watchdog. Facebook displayed ads next to searches for white supremacist groups, even when those outfits were on the social site's blocklist. Recommendations steered visitors to other hate pages, and Facebook's tactic of redirecting users to pro-tolerance groups was only effective for 14 percent out of 226 searches. Some searches for supremacists displayed ads for Black churches. This could effectively identify targets for extremists, TTP said.

In a statement to Engadget, Meta said it "immediately" began removing ads from searches linked to banned groups. It also said it was fixing the issue with a "small number" of auto-generated pages. The company further vowed to keep working with outside experts to "stay ahead" of hate and other extremist content. You can read the full statement below.

The survival of these groups on Facebook isn't completely surprising. University of Michigan associate professor Libby Hemphill told The Post that hate groups are increasingly aware of how to dodge content restrictions. Online platforms are frequently scrambling to adapt, and the TTP study suggests they're not always successful.

Even so, the findings add to Meta's headaches. They come just weeks after GLAAD accused Meta brands of doing too little to protect LGBTQ users, and relatively soon after whistleblower Frances Haugen said Facebook's algorithmic content filtering only caught a "tiny minority" of hate speech. There's plenty of pressure to ramp up anti-hate measures, and it's not yet clear how well the latest fixes will help.

"All 270 groups that Meta has designated as white supremacist organizations are banned from our platform. We invest extensively in technology, people, and research to keep our platforms safe. We immediately resolved an issue where ads were appearing in searches for terms related to banned organizations and we are also working to fix an auto generation issue, which incorrectly impacted a small number of pages. We will continue to work with outside experts and organizations in an effort to stay ahead of violent, hateful, and terrorism-related content and remove such content from our platforms."

Amazon has reduced the price of the Echo Show 5 by $45. With the 53 percent discount, you can buy the smart display for $40, or just $5 more than it was during Prime Day. Released in 2021, the second-generation Echo Show 8 features a 960 x 480 resolution display and a 2-megapixel camera. Amazon offers the device in a trio of colors – Charcoal, Deep Sea Blue and Glacier White. All three models are currently on sale. Like its bigger sibling, the Echo Show 5 includes a sunrise alarm feature, allowing the display to slowly brighten as a way to wake you gently during dark winter mornings. It also comes with all the usual features found on Amazon Echo devices, including Alexa voice control and Ring integration.

Amazon has also discounted the Echo Show 5 Kids. The kid-friendly version is currently $50 after a $45 price drop. Paying the extra $10, you get a smart display with a two-year guarantee and a trial to Amazon Kids+. The service provides access to child-appropriate ebooks, games, videos and apps. Of course, it wouldn’t be an Echo sale if Amazon didn’t include the 4th-generation Dot. The company’s most affordable smart speaker is currently $40, down from $50. You can get the Kids version on sale too. The Echo Dot is a good option for adding Alexa to one of the rooms in your home.

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

Pew Research has published a new report that examines social media usage trends among US teens. The organization found that a whopping 95 percent of them use YouTube, while 19 percent are on the platform "almost constantly."

Perhaps unsurprisingly, two-thirds (67 percent) said they used TikTok, with 16 percent claiming they are on the app "almost constantly." The third most-popular social media platform among teens is Instagram, per Pew, with 62 percent using it. A tenth say they use it almost all the time — despite the app occasionally telling them to take a break. A previous poll conducted in 2014-15 found that 52 percent were using Instagram (Pew didn't ask about YouTube usage for that survey and TikTok didn't exist at the time).

Snapchat also rose among teens, with 59 percent using it in 2022, compared with 41 percent in the previous poll. Facebook was the top social media app among teens seven years ago, with 71 percent of them using it, but that figure has dropped to 32 percent. Teen adoption of Twitter (down from 33 percent to 23 percent) and Tumblr (14 percent to five percent) has fallen over the same period too.

The 2014-15 poll didn't ask about Twitch, WhatsApp or Reddit. These days, a fifth of teens use Twitch, 17 percent are on WhatsApp and 14 percent are accessing Reddit. For what it's worth, the earlier poll suggested 33 percent of teens used Google+, while a quarter used Vine. This time around, Pew did not ask teens about their use of Discord or social gaming spaces such as Fortnite.

Pew surveyed 1,316 teens aged 13 to 17 (as well as one of their parents) in April and May. It found that boys were more likely to use YouTube, Twitch and Reddit and girls were more likely to say they access TikTok, Instagram and Snapchat. More Black and Hispanic teens said they used TikTok, Instagram, Twitter and WhatsApp than white teens.

Even though over half (54 percent) of teens said they'd find it hard to give up social media, 36 percent admitted they spent too much time on the platforms. Around 55 percent said their usage levels were "about right." Meanwhile, 97 percent of teens now use the internet every day, with 46 percent saying they're online almost all the time.

The poll found that 95 percent of teens have access to a smartphone (up from 73 percent in 2014-15), while 90 percent can access a desktop or laptop computer, up from 87 percent in the previous survey. Curiously, the percentage of teens who say they have access to a gaming console has fallen slightly, from 81 percent to 80 percent.

Meta is making its Portal smart displays more useful to remote workers. Starting today, you can use the Portal Go and second-generation Portal Plus as a second screen for your Mac or Windows PC. The new functionality comes courtesy of Duet Display, the app that inspired Apple’s Sidecar feature in macOS Catalina. Normally, you would need to either buy a copy of the software or subscribe to use Duet Display, but CEO Rahul Dewan told Engadget the company is making both wired and local wireless second display functionality free for Portal users.

One thing to note is you can’t use the app while video calling with the Portal Go or Portal Plus. Duet Display is available to download on Portal Plus in the US, Canada, UK, France, Spain, Italy, Australia and New Zealand, and Portal Go in the US, Canada, UK, France, Spain and Italy.

Meta is also releasing the Portal Companion app on macOS. The software allows you to share your screen while on call. You can also use the app to send links to the device and more easily access controls like the mute toggle. The Portal Companion app is available to download in the US and UK.

The arrival of Duet Display on Portal devices comes after Meta reportedly decided to reposition them as enterprise products. In June, The Information and Variety said the company would not announce any new consumer versions of the smart display line.

Boeing is starting to overcome one of its larger hurdles in recent memory. CNBCnotes the aircraft maker has delivered its first 787 Dreamliner in over a year, supplying American Airlines with one out of the nine vehicles it expects to receive in 2022. Boeing paused manufacturing in May 2021 as the Federal Aviation Administration reviewed how the company inspected planes following a string of manufacturing problems.

The company had to halt deliveries multiple times in less than a year after detecting potentially dangerous production problems, such as fuselage spacing. The FAA only cleared Boeing to resume deliveries on Monday. Dreamliner handovers have been on hold for most of the past two years between the manufacturing defects and a pandemic that dramatically curbed passenger flights.

There's a strong incentive to put the 787 Dreamliner into customers' hands. The flaws and ensuing production cuts will cost Boeing $5.5 billion, and that's on top of serious 737 Max issues that led to crashes killing 346 people. Boeing has a tarnished reputation, and these deliveries could help it (slowly) mend its image while capitalizing on an air travel revival.

Is there much more space for innovation in desks, I hear you ask? After all, now that we’ve made them go up and down, there are no new worlds worth conquering. Not so, says Lumina, makers of its eponymous AI webcam that’s been described as the equal of a DSLR in some corners. Now, the company is turning its attention to building a smart desk with a programmable, 24-inch OLED screen nestled in its top, designed to offer you a place to put passive data in easy view.

The Lumina Desk, as it’s called, is a powered sit-stand desk, with a display that sits between your laptop (or monitor) and your keyboard and mouse. The idea, as you can see, is to offer you space to add in a view for your Google Calendar, stock view from Robinhood or a Twitter feed. All things that it’d be nice to always have in view, but you may not necessarily want to buy a second (or third) monitor for.

Lumina

Rather than a touchscreen, you’ll control data from a companion app on the desktop, and the whole desk is coated in a layer of toughened, anti-glare glass. The company is also promising a maximum brightness of 1,200 nits, and a 60Hz refresh rate, making it a pretty sweet display all told. There’s also plenty of power coursing through its frame, with two 20 x 20 Qi wireless charging pads on the tabletop, each one able to pump out 100W. Then there six hidden AC outlets with a built-in circuit breaker, and six USB-C ports, each one able to deliver up to 30W.

Sadly, you won’t be able to get your hands on one of these until 2023 at the earliest, and the company is today opening up unpaid reservations to gauge future interest. The price, too, has yet to be decided (although I’m told it’ll be as close to $1,000 as possible, at least that’s the aim right now) but if you’ve been looking for a more sci-fi way to passively collect information, this might be worth a peep.