Preteens who use increasing amounts of social media perform poorer in reading, vocabulary and memory tests in early adolescence compared to those who use little or no social media. A study published in JAMA examined data from over 6,000 children ages 9 to 10 through early adolescence. Researchers classified the children into three groups: 58% used little or no social media over several years, 37% started with low-level use but spent about an hour daily on social media by age 13, and 6% spent three or more hours daily by that age.

Even low users who spent about one hour per day performed 1 to 2 points lower on reading and memory tasks compared to non-users. High users performed 4 to 5 points lower than non-social media users. Jason Nagata, a pediatrician at the University of California, San Francisco and study author, said the findings were notable because even modest social media use correlated with lower cognitive scores.

Read more of this story at Slashdot.

“ChatGPT shouldn’t have political bias in any direction.”

That’s OpenAI’s stated goal in a new research paper released Thursday about measuring and reducing political bias in its AI models. The company says that “people use ChatGPT as a tool to learn and explore ideas” and argues “that only works if they trust ChatGPT to be objective.”

But a closer reading of OpenAI’s paper reveals something different from what the company’s framing of objectivity suggests. The company never actually defines what it means by “bias.” And its evaluation axes show that it’s focused on stopping ChatGPT from several behaviors: acting like it has personal political opinions, amplifying users’ emotional political language, and providing one-sided coverage of contested topics.

Read full article

Comments

Did you know you can customize Google to filter out garbage? Take these steps for better search results, including adding my work at Lifehacker as a preferred source.

While sometimes there’s a big piece of misinformation that a lot of people latch onto—like The Rapture or the existence of “MedBeds“—the fractured nature of the information sphere has all but killed the overarching conspiracy theory. No longer do big ideas like “we never went to the moon” unite the dumbest minds; instead, the algorithm creates bespoke conspiracy theories. So instead of joining the Flat Earth Society, you might think the actual year is 1728, or that AI secretly imagined a British comedian from the the 1980s and seeded the web with evidence of his existence.

But how does it start? And how quickly can social media platforms transform someone from a person from a seeker-of-knowledge to a believer-in-bullshit? YouTuber Benaminute recently posted a video where he dug in to find out. His question: If you start with a benign, broad, randomly chosen subject, and you only watch videos having to do with that subject, how long will it take until TikTok, YouTube Shorts, and Instagram Reels feed you a conspiracy theory video? The answer: not long at all.

Different topics all lead to the same place (more or less)

For the experiment, Benaminute created “blank” social media profiles and behaved like someone who was innocently curious about one of three topics—dinosaurs, The Vietnam War, and the 2000 presidential election. He put the keyword in each platform’s search bar and only watched and liked videos about the initial subject.

Dinosaurs

• YouTube Shorts: The initial videos were ads for Jurassic Park, AI slop featuring dinosaurs, and the occasional educational video, but the 541st video was a clip from the Joe Rogan Experience about how the pyramids were not tombs, but “DNA restoration devices.”

• TikTok: If you thought TikTok would get to conspiracy theories quickly, you’d be right. The 144th video was this fake UFO video that has 24 million views.

• Instagram Reels: Insta took 661 videos to get from dinosaurs to a “forbidden phone from the 2000s that lets you see into a parallel dimension.”

The Vietnam War

Things get worse for people interested in historical or political events. On all short-form platforms, an interest in Vietnam will lead you pretty quickly to right-leaning content, which leads you to conspiracy theories.

• YouTube Shorts got to a conspiracy theory video about Noah’s Ark in only seven videos.

• TikTok took a little longer; video 161 was about how financial services company Blackrock had something to do with the attempted assassination of Donald Trump.

• Reels took 139 videos to get to “Bush did 9/11.”

The 2000 election

The election of 2000 is still a charged topic, but it’s been awhile, so maybe cooler heads and verified information will win the day? Spoiler: nope.

• YouTube Shorts took 136 videos to get to the same Noah’s Ark conspiracy as it did for dinosaur fans.

• TikTok only took 38 videos to get to “The Rapture is happening on September 24.”

• Reels took only 26 videos to land on “The World Trade Center was bombed” (by either Clinton or Bush).

Which social media app leads to conspiracy theories fastest?

The champion of “normal search to conspiracy theory” speed runs is TikTok, with an average of 114 videos or 57 minutes of watching. YouTube Shorts comes in second with 230 videos or 1 hour 57 minutes of time, and Reels takes 275 video or 138 minutes. It’s a distinction without a difference; however, all three platforms lead to conspiracies in the time it takes to watch a Marvel movie.

What does it all mean?

It would be easy to conclude that the massive tech companies that built YouTube, Instagram, and TikTok companies weight their recommendation engines so viewers are led to fake stories. Maybe they have specific political aims and are trying to sway votes, or maybe (as Benaminute posits in a semi-tongue-in-cheek way) these apps are built to “keep us angry, divided, and distracted” from realizing the conflict isn’t between Left and Right, but between “up and down.”

This is also a conspiracy theory, however. I’m not saying he’s wrong, but we don’t have enough information to know why algorithms recommend conspiracy content. It could be because bad actors at the top demand specific results for some purpose, but it seems more likely to me that TikTok et al. don’t have an agenda beyond making money.

I have no doubt that a social media platform featuring an algorithm that weighs the truth heavily would fail pretty quickly; the Truth is boring compared to conspiracy theories. Conspiracy theories, broadly, make believers feel special, like they have inside knowledge the rest of us lack. People scroll TikTok to have fun; the truth usually isn’t fun. Conspiracy theorist can say things like “UFOs are here!” or “They’re turning the frogs gay!” Meanwhile, if you’re devoted to the truth, you mostly have to go with “the best evidence suggests…” or “it seems logical that…” and who wants to hear that?

SpaceX’s second-generation Starship vehicle has just made a graceful exit. The company achieved every major objective it set for the super-heavy lift vehicle’s 11th flight test, the second-gen Starship’s final flight, which launched from Starbase in Texas on October 13. It followed another successful test in August, which saw Starship deploy its payload for the first time ever. Before those two most recent flights, SpaceX suffered a series of failures: Starship exploded during its ascent stage in the company’s seventh and eighth tests, and it failed to deploy its payload during its ninth test. Another Starship vehicle blew up on the ground during a routine test while SpaceX was preparing for its 10th flight. 

All of the vehicle’s 33 Raptor engines ignited upon launch, and the stage separation and first-stage ascent went smoothly. The Super Heavy booster splashed down into the ocean as planned, while Starship was able to deploy all its Starlink simulators before re-entering the atmosphere. During its reentry burn, SpaceX intentionally stressed the vehicle to determine the capabilities of its heatshield. And with just a few minutes left to the flight, the vehicle executed a banking maneuver to “mimic the trajectory that future missions returning to Starbase will fly.”

The company says it will now focus on developing the next generation of Starship and Super Heavy. It has multiple versions of the vehicle and the booster being prepared for tests at the moment, and it expects them to be used for the first Starship orbital flights and operational payload missions. 

This article originally appeared on Engadget at https://www.engadget.com/science/space/spacex-is-preparing-the-next-gen-starship-after-a-successful-flight-test-130027382.html?src=rss

We may earn a commission from links on this page. Deal pricing and availability subject to change after time of publication.

Did you know you can customize Google to filter out garbage? Take these steps for better search results, including adding Lifehacker as a preferred source for tech news.

If you’ve been looking for a great pair of noise-canceling earbuds that don’t cost a fortune, the JBL Tune Buds are now down to $44.95, their lowest price yet, according to price trackers. That’s less than half of what they’re originally sold for, which makes them a strong contender in the under-$50 range.

JBL Tune Buds

$44.95
at Amazon

$99.95
Save $55.00

$44.95
at Amazon

$99.95
Save $55.00

These are true wireless earbuds that focus on delivering that signature JBL bass (deep and thumpy) but with enough flexibility to tweak the sound using the JBL app. Out of the box, they lean heavy on the low end, which works well for pop and EDM, but you can easily adjust the EQ if you prefer something more neutral.

Performance-wise, the Tune Buds pack in more than you’d expect for the price. They support Bluetooth 5.3, AAC, and SBC codecs, along with multipoint pairing, so you can jump between your phone and laptop without reconnecting. The active noise cancellation is solid for casual use—good enough to dull traffic noise or a chatty office, even if it can’t fully block the low rumble of a subway, notes this PCMag review. There’s also an IP54 rating, meaning these can handle sweat and light rain, making them suitable for workouts or daily commutes. Touch controls are customizable through the companion app, and there’s built-in Alexa support if you like using voice commands hands-free.

As for the battery life, JBL claims up to 10 hours per charge with ANC on and 12 hours without, plus an extra 30 to 36 hours from the charging case. That’s easily a few days of use (your mileage may vary) without needing to plug in. The 10mm drivers cover the standard 20Hz to 20kHz range, and the earbuds come with three silicone tip sizes to help with comfort and fit. They don’t offer premium-level sound or the silence of high-end ANC sets from Sony or Bose, but they strike a good balance between affordability, sound, and everyday convenience.

Apple has rebranded its streaming service Apple TV+ to simply Apple TV, further blurring the already confusing line between the Apple TV device, the Apple TV app, and the Apple TV service. As John Gruber notes, users can now “watch Apple TV in Apple TV on Apple TV.” From Daring Fireball: In some ways, I get it. Like, if you’re telling someone how much you enjoy Slow Horses and they ask how to watch it, it’s more natural and conversational to just say “It’s on Apple TV.” That’s what most people say. That’s what I say — and as part of my job, I completely understand the difference between Apple TV the device, Apple TV the (free) app, and Apple TV+ the (paid) streaming service.

But right there in Apple’s own “About Apple TV” description, you see just how overused “Apple TV” now is. You can watch Apple TV in Apple TV on Apple TV — the paid service in the free app on the set-top box. But you can watch any streaming service you want on the box, in that service’s own app. But many of those services are also available in the Apple TV app. And the Apple TV streaming service is also available on just about all other popular set-top hardware platforms. So don’t need an Apple TV to watch Apple TV. It’s a bit like Abbott and Costello’s classic “Who’s on First” routine.

Read more of this story at Slashdot.

Therabody just revealed a bunch of new products, including the TheraFace Mask Glo. This MF Doom-looking full-face mask offers quite a different experience from the pre-existing TheraFace Pro. The Pro is essentially a face massager, but the Mask Glo is a giant mask that uses LED lights to “reduce fine lines and wrinkles, firm and tighten skin and even tone and texture.”

The company promises results in “as little as eight weeks” with daily 12-minute sessions. The mask has been cleared by the FDA and includes 504 medical-grade LEDs to deliver “an effective, full-face treatment.” It also doubles as a Halloween mask if you are so inclined. Tis the season. The TheraFace Mask Glo is available to order today and costs $380.

The Theragun Mini Plus is a portable version of its popular massage gun, but with one feature missing from the original Mini. This one has heat. It combines “percussive massage with consistent, rapid heat.” Therabody says this results in a “benefits 3x faster than percussive massage alone.”

Therabody

It features three massage speeds, three heat levels and app-guided routines to target specific muscle groups. The gun is TSA-approved, so bring it on a plane. The Theragun Mini Plus is available today and costs $280.

The company also released a new version of the Theragun Prime massage gun, which costs $330, and the Theragun Sense massage gun. That last one features an LCD screen and real-time pressure feedback to make sure people are using it right. It costs $300.

This article originally appeared on Engadget at https://www.engadget.com/home/therabody-just-introduced-the-theraface-mask-glo-which-uses-leds-to-reduce-wrinkles-123044315.html?src=rss

TiVo has confirmed that it has stopped selling its DVR set-top boxes, marking the end of an era that changed how we watch television forever. As first reported earlier this month by Cord Cutters News, TiVo Corporation quietly pulled its once-groundbreaking digital video recorder from its website. Holding company Xperi later confirmed that the listing was removed on October 1.

“I can confirm that as of October 1, 2025, TiVo stopped selling physical DVR products, including hardware and accessories, both online and through agents,” a TiVo spokesperson confirmed to PCMag. “TiVo no longer manufactures hardware, and our remaining inventory is now depleted, though we will continue to offer support for the products going forward.”

The TiVo box revolutionized television upon its launch in 1999, allowing viewers to pause, rewind and record live television. There was a time when you would just miss the start of a show if you weren’t punctual, and you’d have to sync grabbing a snack with a commercial during the big game. You also had to actually watch the commercials, something that is unfortunately making a comeback with an increase in ad-supported streaming.

The DVR pioneer is now a software company. It has been producing TiVo OS almost exclusively in the European market since 2022, though the smart TV OS premiered in the US this year via Sharp. “The Sharp Smart TV Powered by TiVo” launched as a 55-inch QLED display with 4K resolution and HDR support. TiVo OS functions like many other television operating systems, aggregating streaming services and offering its own library of free and paid content.

TiVo will still offer customer support for its now-discontinued boxes, which bodes well for customers who have purchased a lifetime plan.

This article originally appeared on Engadget at https://www.engadget.com/entertainment/tv-movies/tivo-has-discontinued-its-dvr-boxes-123037999.html?src=rss