TikTok has announced layoffs affecting hundreds of employees worldwide, with a significant impact on its workforce in Malaysia. The social media giant confirmed the cuts on Friday, citing a strategic move to enhance using artificial intelligence (AI) in its content moderation processes. The decision aligns with TikTok’s goal to streamline operations and increase efficiency in managing its vast user-generated content.

Impact in Malaysia: Hundreds of Jobs Affected to enhance the use of AI

Sources familiar with the situation told Reuters that TikTok initially planned to lay off more than 700 employees in Malaysia. However, the company later clarified that fewer than 500 workers in the country were actually affected by the decision. Most of those losing their jobs were part of the firm’s content moderation team.

TikTok informed the affected employees of the layoffs through email late Wednesday. The job cuts align with the company’s broader plan to enhance its global moderation efforts using AI technology.

The company uses both automated systems and human moderators to review its content. The company aims to increase AI’s role in moderation. It believes this shift will improve its efficiency and effectiveness in handling content.

A TikTok spokesperson said the company is committed to strengthening its content moderation model globally. “We’re making these changes as part of our ongoing efforts to further strengthen our global operating model for content moderation,” the spokesperson said. The shift will allow TikTok to handle the growing volume of content more effectively, with the help of advanced AI technologies.

Investment in Trust and Safety

Despite the layoffs, TikTok has indicated its continued investment in trust and safety initiatives. The popular social media app plans to allocate $2 billion globally this year to enhance its content moderation and ensure a safer user experience. According to TikTok, around 80% of content that violates guidelines is already being flagged and removed through automated technologies.

ByteDance, TikTok’s parent company, currently employs over 110,000 people in more than 200 cities worldwide. This restructuring move, however, is expected to lead to further job cuts as TikTok consolidates some of its regional operations.

The company’s efforts to bolster AI-based content moderation come amid increasing pressure from governments around the world, including Malaysia. In the first half of 2023, TikTok and Meta faced a record number of content restriction requests from the Malaysian government. These requests often targeted posts related to sensitive topics, such as race, religion, and royalty.

October Prime Day may have come and gone, but you can still save on Lego sets at the moment. While you may think it’s early to think about holiday shopping, these kits make amazing presents — for the young and full-grown adults alike. At Engadget, we’re most interested in specific sets from lines like Super Mario, Star Wars and Harry Potter, and some of these are still on sale for up to 41 percent off. These are the best Prime Day Lego deals you can still get today.

Prime Day Lego deals still available

On the Star Wars side of things, this Spider Tank set is 36 percent off and down to only $32, which is the lowest it’s ever been. It includes 526 pieces that replicate the spider tank from season three of The Mandalorian, plus three minifigures: Din Djarin, Bo-Katan Kryze and Grogu. Once built, the spider tank has grabbing claws, flexible legs and a little cockpit in which one of the figures can sit. Also on sale is this Boarding The Tantive IV set in which you recreate the iconic scene from Star Wars: A New Hope. That will set you back $44, which represents a 20-percent discount.

The Star Wars Lego Advent calendar (complete with Luke and Leia in holiday sweaters) is down to $36, after a 20 percent discount, and we don’t expect this one to last long.

In the Mario space, this Dixie Kong’s Jungle Jam expansion set has the biggest discount: 41 percent off and down to $16. It has 174 pieces along with buildable Dixie Kong and Squawks figures that both come with musical accessories. Mario fans who are old enough to have a work-from-home setup might appreciate this displayable Piranha Plant set that would look great in the background of any video conference call. It’s 20 percent off and down to $48.

Rounding things out with Harry Potter sets, this Hogwarts Castle and Grounds set is down to $136 and has never been cheaper. It includes 2,660 pieces that create a final product that’s over eight inches high, 13 inches wide and 10 inches deep. Plus, it comes with a cute, golden Hogwarts architect statue minifigure.

If you’re looking for more general Lego sets, the best deal we found was on the Classic Medium Creative Brick Box, down to $21. The box includes 484 pieces in all different sizes and colors, and would make a great gift for anyone who just likes to build with Lego without following a set of instructions.

Follow @EngadgetDeals on Twitter for the latest tech deals and buying advice, and stay tuned to Engadget.com for all of the best tech deals coming out of October Prime Day 2024.

The number of AI video generation models continues to grow with a new one, Pyramid Flow, launching this week and offering high quality video clips up to 10 seconds in length — quickly, and all open source.

Developed by a collaboration of researchers from Peking University, Beijing University of Posts and Telecommunications, and Kuaishou Technology — the latter the creator of the well-reviewed proprietary Kling AI video generator — Pyramid Flow leverages a new technique wherein a single AI model generates video in stages, most of them low resolution, saving only a full-res version for the end of its generation process.

It’s available as raw code for download on Hugging Face and Github, and can be run in an inference shell here but requires the user to download and run the model code on their own machine.

At inference, the model can generate a 5-second, 384p video in just 56 seconds—on par with or faster than many full-sequence diffusion counterparts — though Runway’s Gen 3-Alpha Turbo still takes cake in terms of speed of AI video generation, coming in at under one minute and often times 10-20 seconds in our tests.

We haven’t had a chance to test Pyramid Flow yet, but the videos posted by the model creators appear to be incredibly lifelike, high enough resolution, and compelling — analogous to those of proprietary offerings. You can see various examples here on its Github project page.

Indeed, Pyramid Flow is available designed now to download and use — even for commercial/enterprise purposes — and is designed to compete directly with paid proprietary offerings such as Runway’s Gen-3 Alpha, Luma’s Dream Machine, Kling, and Haulio, which can cost hundreds of even thousands of dollars a year for users on unlimited generation subscriptions.

As the race between various AI video providers to gain users continues, Pyramid Flow aims to bring more efficiency and flexibility to developers, artists, and creators seeking advanced video generation capabilities.

A new technique for high-quality AI videos: ‘pyramidal flow matching’

AI video generation is a computationally intensive task that typically involves modeling large spatiotemporal spaces. Traditional methods often require separate models for different stages of the process, which limits flexibility and increases the complexity of training.

Pyramid Flow is built on the concept of pyramidal flow matching, a method that drastically cuts down the computational cost of video generation while maintaining high visual quality, completing the video generation process as a series of “pyramid” stages, with only the final stage operating at full resolution.

It’s described in a pre-reviewed paper, “Pyramidal Flow Matching for Efficient Video Generative Modeling,” submitted to open access science journal arXiv on October 8, 2024.

The authors include Yang Jin, Zhicheng Sun, Ningyuan Li, Kun Xu, Hao Jiang, Nan Zhuang, Quzhe Huang, Yang Song, Yadong Mu, and Zhouchen Lin. Most of these researchers are affiliated with Peking University, while others are from Kuaishou Technology.

As they write, the ability to compress and optimize video generation at different stages leads to faster convergence during training, allowing Pyramid Flow to generate more samples per training batch.

For example, the proposed pyramidal flow reduces the token count by a factor of four compared to traditional diffusion models, which results in more efficient training.

The model can produce 5- to 10-second videos at 768p resolution and 24 frames per second, all while being trained on open-source datasets. Specifically, the paper states that Pyramid Flow was trained on trained on:

• LAION-5B, a large dataset for multimodal AI research.
• CC-12M, a dataset of web-crawled image-text pairs.
• SA-1B, which features high-quality, non-blurred images.
• WebVid-10M and OpenVid-1M, which are video datasets widely used for text-to-video generation.

In total, the authors curated approximately 10 million single-shot videos.

However, many of these “public” or “open source” datasets have in recent years come under fire from critics for including copyrighted material without permission or informed consent of the copyright holders, and LAION-5B in particular accused of hosting child sexual abuse material.

Separately, Runway is among the companies being sued by artists in a class action lawsuit for training on materials without permission, compensation, or consent — allegedly in violation of U.S. copyright. The case remains being argued in court, for now.

Permissively licensed, open source for commercial usage

Pyramid Flow is released under the MIT License, allowing for a wide range of uses, including commercial applications, modifications, and redistribution, provided the copyright notice is preserved.

This makes Pyramid Flow an attractive option for developers and companies looking to integrate the model into proprietary systems, and could challenge Luma AI and Runway as both look to offer paid application programming interfaces for developers seeking to integrate their proprietary AI video generation technology into customer or employee-facing apps.

Yet those proprietary models already exist as inferences suitable for developers, while Pyramid Flow has a demo inference on Hugging Face, it is not suitable for building full applications atop it and users would need to host their own version of an inference, which could also be costly, despite the model itself being “free.”

In addition, Pyramid Flow may prove to be enticing to film studios looking to leverage AI to gain efficiencies, cut costs, and explore new creative tools. One major film studio, Lionsgate — owner of the John Wick and Twilight films franchises, among many other tiles — recently inked a deal for an unspecified sum with Runway to train a custom AI video generation model. Furthermore, Titanic and Terminator director James Cameron joined the board of AI video and image model provider Stability (the latter also subject to the same class-action lawsuit from artists as Runway).

Using Pyramid Flow, Lionsgate or any other film studio could fine-tune the open source version without paying a third party company. However, they would still need to have on hand or contract out the developer talent and computing resources necessary to do so, which may make partnering with established AI providers such as Runway more appealing, since that company and others like it already have the AI engineering talent at their disposal in house.

The research team behind Pyramidal Flow Matching has also made a commitment to openness and accessibility. All code and model weights will be made freely available to the public through their official project page, ensuring that researchers and developers around the world can utilize and build upon this work.

Despite its strengths, Pyramid Flow does have some limitations. For now, it lacks some of the advanced fine-tuning capabilities found in models like Runway Gen-3 Alpha, which offers precise control over cinematic elements like camera angles, keyframes, and human gestures. Similarly, Luma’s Dream Machine provides advanced camera control options that Pyramid Flow is still catching up to.

Moreover, the relatively recent launch of Pyramid Flow means its ecosystem—while robust—isn’t as mature as those of its competitors.

Looking ahead: AI video race shows no signs of slowing

As the AI video generation market continues to evolve, Pyramid Flow’s launch signals a shift toward more accessible, open-source solutions that can compete with proprietary offerings such as Runway and Luma.

For now, it offers a solid alternative for those looking to avoid the cost and limitations of closed models, while providing impressive video quality on par with its more commercial counterparts.

In the coming months, developers and creators will likely keep a close eye on Pyramid Flow’s growth. With the potential for further improvements and optimizations, it could very well become a go-to tool in the arsenal of video content creators everywhere. All the companies and researchers are currently battling both for technological supremacy and users.

Meanwhile, OpenAI’s Sora, first shown off in February 2024, remains nowhere to be seen — outside of its collaborations with a handful of small early alpha users.

How to preorder the standard PlayStation 5 Pro

$699

The “pro” version of Sony’s PlayStation 5 has a bigger GPU, new PSSR AI-based upscaling, advanced ray tracing, and 2TB SSD (with expandable storage). The new features allow for better graphics at higher frame rates, allowing for the blending of “fidelity” and “performance” modes in some games.

Where to order a PS5 Pro disc drive

$80

Sony’s official disc drive for PS5 Digital Edition consoles allows you to retrofit both the PS5 Digital Edition and PS5 Pro with a bay for inserting physical games.

Huawei has announced the seventh generation of its all-flash OceanStor Dorado array, claiming it will meet enterprise data storage needs for the AI era.

The new storage solution is specifically designed for mission-critical applications, delivering what the company describes as “extreme performance and resilience” for complex AI workloads.

Earlier this week, Variety published a list of the 100 best horror movies ever. Sitting at the top, like an exhumed corpse festering in the brilliant midday sun, was The Texas Chain Saw Massacre. This was not a controversial choice on the publication’s part, not in the year of our unholy lord of darkness 2024. Tobe Hooper’s deranged thriller, which roared into theaters 50 years ago, has been rising in critical esteem for decades, its reputation as a truly great movie — rather than merely a deeply upsetting and effective one — steadily cementing over the last half-century. Time, in other words, has been very kind to a savage, scandalous act of grindhouse exploitation once considered so shocking, it was banned in multiple countries. Yesterday’s outrage machine has become today’s lionized classic.

For as much as the movie deserves every drop of overdue recognition it’s increasingly earned, it’s still a little unusual seeing The Texas Chain Saw Massacre canonized by mainstream arbiters of taste. The next two films on Variety’s list, The Exorcist and Psycho, have more commonly wrestled for the nebulous title of horror’s pinnacle. Both of those movies were plenty shocking and controversial in their time, of course. But like most historic hair- and knuckle-whiteners, they’ve lost a little of their transgressive power over the decades since, as the standard of what gets under the skin of the average moviegoer evolves. Generally speaking, they don’t traumatize like they used to. They’re safer — which, in a sense, makes them easier to enshrine as the Mona Lisas of their video store aisle.

The Texas Chain Saw Massacre is different. This is not a safe movie. Though it’s now widely recognized as a masterpiece, it’s not something you appreciate from a respectful distance, admiring its historic qualities like an anthropologist of B-movie artifacts. It’s an experience, undiminished by everything that’s come after it. The primal immediacy of Hooper’s achievement — the sheer demented intensity of his 83-minute assault on the senses — has not waned. The Texas Chain Saw Massacre will still mess you up. If it doesn’t, there might actually be something off about you.

John Larroquette provides the faux-true-crime narration of the opening scene — a dryly ominous introduction that frames the events to come as reenactments of real unspeakable crimes, which the wildly successful marketing campaign exploited. This is not a true story in any literal sense, though Hooper did base some of the violence on the very real mayhem of the serial killer Ed Gein. In a much more general sense, few films have felt more in touch with the madness of American culture, the evil lurking in our country’s heart and its forgotten corners.

Part of what remains so unnerving about the film, five decades later, is how it seems to straddle the line between a harsh, scraggly, almost documentary realism and something more hallucinatory. Even as Hooper rubs our noses in forensic evidence (retrospective glimpses of the horror to come — a devious device for triggering our dread early), he also begins to pull at the fabric of the reality he’s establishing, washing out the images, drowning out the audio with droning, atonal music. It’s as if the insanity of the Sawyer clan was already polluting the movie’s style, minutes before we’ve met a single one of them.

Rewatching the film, it’s striking how much its opening act functions as one long premonition of doom. Over and over again, Hooper seems to erect an existential “Turn back now” sign, giving his van of unlucky city slickers numerous warnings that something awful waits down the road they’re traversing. Hell, one of the abominable killers himself tries to warn them, in his own way, in a gas station scene that would become a cliché of the 1980s slasher movies Texas Chain Saw helped inspire. The early stretch of the film is littered with bad omens: roadkill, reports of violence on the radio, the roar of what just might be a chainsaw in the distance. Even the daily horoscope seems to be beaming in a red alert from the universe: “There are moments where we cannot believe what is happening is really true,” one of the kids reads aloud. She’ll grasp the meaning of those words when she’s hanging from a meat hook later in the afternoon.

So much has been written over the years about Leatherface’s iconic first appearance, that moment when he stumbles into a doorway without prelude or fanfare, and clocks someone dead with a hammer, before slamming the metal door behind him. It happens so fast, you can almost miss it. It’s like the “We’re gonna need a bigger boat” scene in Jaws, the multiplex phenomenon that opened a year later: a jump scare so unexpected and so off rhythm — a moment you can’t possibly anticipate — that it scrambles your sense of security. Nearly half a century earlier, Boris Karloff got a star’s entrance as Frankenstein’s monster, slowly turning to face the camera and reveal his hideous face. The Texas Chain Saw Massacre seemed to announce a scary new era of sudden lunacy, when the monster simply blips into our line of sight, too obscene for formal introduction.

The scene almost plays like a hole has been torn in the fabric of time and space, depositing something horrible and beyond reason. Who in their right mind thought we needed an origin story (nay, two!) for this brute? Leatherface is so much scarier as an evil that’s just suddenly there. Not a single one of the sequels, prequels, and remakes is essential. They all give us more than we need of this awful place, these unknowable monsters. They try to bring psychology into the matter, when the Sawyers — bloodthirsty emblems of American derangement — exist beyond the purview of diagnosis. The Texas Chain Saw Massacre is a complete thing. To continue what passes for its story is futile, because its power can’t be replicated. Hooper seemed to realize that, and took his own over-the-top sequel, the best one, in the only direction that made sense: black comedy.

All the mythology those mostly dire follow-ups would build upon is largely implied in the original. No one ever even identifies the Sawyers as cannibals; a couple ominous close-ups of cooking meat at the BBQ pitstop says it all, though, doesn’t it? That goes for the movie’s politics, too. The Texas Chain Saw Massacre is one of the most witheringly resonant movies ever made about the violent spirit of America, but it doesn’t do the interpretative work for you, the way so many modern festival-feted creepshows do. There’s a world of sociopolitical meaning in the one-line revelation that the Sawyers used to be factory workers before automation put them out to pasture. British censors certainly got the message, warning that the film might inspire something in the working class. Fearmongering? Of course. But you can’t say this isn’t a political movie. It just keeps its ideas draped in nightmare logic.

Another thing that’s easy to forget, if you haven’t seen The Texas Chain Saw Massacre in a while, is how elegant it is — especially for a film made on a shoestring budget and with such savage, unpretentious aims. The slaughterhouse ambiance of the film looms large in the imagination, but there’s nothing artless about how it’s put together; you could go shot for shot through Chain Saw Massacre, like Roger Ebert used to do with Citizen Kane and other art-house milestones in lecture halls, and find something to admire after every cut. It’s such a carefully, brilliantly assembled movie — maybe the ultimate example of how the best horror movies work your nerves with the how as well as the what of their dark vision.

It’s in the last half hour that the movie starts to feel truly wrong, like something you shouldn’t be watching, like something that short-circuits that old “it’s only a movie” rationale we use to get us through rough sledding. It’s not the violence, which never gets all that explicit (to the point where Hooper hilariously reasoned that he might be able to secure a PG rating for the film, can you even imagine?). It’s the way The Texas Chain Saw Massacre devolves into pure, primal emotion, as Sally runs and screams and pleads for what feels like a grueling eternity, all while her tormentors buffoonishly giggle like Hee Haw bit players. 

In the pantheon of scream queens, there is Marilyn Burns and then there is everyone else. No one has seemed as believably destroyed by terror as she does here. The film runs less than an hour and a half, but that dinner table scene — all bulging eyes in extreme close up, all slapstick near-death — seems to go on forever. That’s because Hooper has locked us into Sally’s crucible, and offered a vision of insanity that feels realer than what movies ever offer. It’s hard to think, too, of a more iconic ending to a horror movie — that frustrated chainsaw ballet in the daylight, Sally laughing hysterically with a relief that tells you she’ll never be OK ever again, an abrupt cut to credits denying us the creature comfort of a denouement. 

Yes, fear is subjective — one person’s phobia fuel is another’s sleeping pill, blah blah blah. You may have your own personal, idiosyncratic choice for scariest movie ever. (For this writer, no single moment in cinema is more irrationally petrifying than the Winkie’s diner scene in Mulholland Drive, which isn’t even a horror movie by the strictest definition.) But if we’re talking about consensus power to unsettle, there’s still nothing like The Texas Chain Saw Massacre. It simply doesn’t operate like a normal movie. It feels inherently touched by death from its opening frames. And it eventually cracks into a madness beyond plot or suspense — a total immersion into blinding panic and fear. It’s possible no movie has ever felt more like a nightmare. You wake up, but it’s still there, twirling like Leatherface in your head.

The Texas Chain Saw Massacre is currently streaming on Peacock, Tubi, Plex, and other streaming services. For more of A.A. Dowd’s writing, visit his Authory page.