‘Avatar’ sequel’s cutting-edge tech crashed some movie projectors in Japan

Despite being the widest release of all time in Japan, Avatar: The Way of the Waterfailed to claim the top ranking last weekend as it was topped by an anime basketball picture called The First Slam Dunk. On top of that, multiple theaters in the nation reported technical problems, with one in central Japan forced to reduce the 48 fps frame rate down to the traditional 24 fps, Bloomberg reported. 

Fans were reportedly turned away from other screenings and issued refunds. Some of the theater chains cited by fans as having issues, including United Cinemas Co., Toho Col, and Tokyu Corp., declined to comment on the problem.

Not many movie theaters support high frame rate (HFR) 48 fps playback, as it requires the latest projectors or upgrades to existing ones. Normally, movie theaters would be aware of which formats they can play and plan accordingly. But HFR has been used so little that it would be understandable if errors cropped up. 

Avatar: The Way of the Water is available in multiple formats, including 2D 48 fps, 3D 48 fps and regular 24 fps. If you see the 48 fps version, it only uses the HFR tech for action sequences, while dialog and slower scenes are dialed down to 24 fps (by duplicating frames). Engadget’s Devindra Hardawar saw the film at 48 fps and liked it, but added that the technology remains divisive.

Other notable films using HFR were Ang Lee’s Gemini Man and The Hobbit trilogy. When the latter came out in 2012, I argued that high frame rates work best with the 3D format as it helps eliminate potential eye strain and even nausea. That’s not an issue in 2D, so 24 fps gives the most natural look with none of the video game/soap opera effect that many people dislike.

Elon Musk says he’ll step down as Twitter CEO, but won’t sell the company

Elon Musk has said that he will step down as CEO of Twitter once a suitable replacement can be found. On Sunday he ran a poll asking if he should leave the role, and Twitter users overwhelmingly told him to go. He didn’t immediately respond to the results, but by Tuesday he seemed to have accepted the will of the people, after originally suggesting that he might instead change it so that only paying users could vote in Twitter polls.

Of course there’s no timeline for Musk to hand over the reins yet. And as his tweet says it will require finding “someone foolish enough to take the job.” Following a chaoticstring of badpress and thegutting of its staff it might be hard to attract high quality candidates to the position. But, the bar has been set exceedingly low under Musk’s tenure and a simple change of face at the top could do a lot to improve Twitter’s image in the short term.

Of course, there is a chance that the change could be largely cosmetic. Musk tweeted that he planned to stay on and run the software and server teams. And there’s no indication that he plans to sell the company, which he bought in October for $44 billion and quickly took private. So it’s clear that he wants to still be involved in the company’s operation, but it’s unclear just how much. 

It’s hard to imagine the hard driving businessman who made a major spectacle of firing Twitter’s entire executive team and dissolved the board when he took over simply stepping aside at this point. But he has also in his short reign allowed the whim of polls to make several major decisions. 

Regardless of who eventually takes over as CEO, they’ll be taking over a company that is almost unrecognizable compared to just a few months ago. Roughly 70 percent of the staff have either been laid off or quit. Advertisers have fled the platform as hate speech and trolls have thrived. And the content moderation and safety efforts have completely stagnated. The company is also now facing increasing pressure from the government on a number of fronts. 

FTX founder Sam Bankman-Fried agrees to extradition to the US

When the Bahamas Attorney General’s office announced that it had arrested former FTX CEO Sam Bankman-Fried, it noted that the former FTX CEO was likely to be extradited at the request of the United States. Just over a week later, that prediction has co…

The EU is investigating Broadcom’s $61 billion deal to buy VMware

The European Union plans to carry out a full-scale investigation of Broadcom’s $61 billion bid to buy VMware. Following a preliminary probe, the European Commission, the bloc’s executive branch, announced on Tuesday it believes the proposed acquisition may allow Broadcom to “restrict competition” in the markets for network interface cards, fiber channel host-bus adapters and storage adapters.

Specifically, the EU is concerned Broadcom may harm competition in those markets by limiting interoperability between rival hardware and VMware’s server virtualization software. It also worries the company could either prevent or degrade access to VMware’s software. The European Commission warns those actions “could lead to higher prices, lower quality and less innovation for business customers, and ultimately consumers.”

The Commission will also investigate whether Broadcom could hinder rivals like NVIDIA and Intel from developing their own smart network interface cards. Here it points to VMWare’s involvement in Project Monterey, an industry-wide effort the company announced in 2020. “Broadcom may decrease VMware’s involvement in Project Monterey to protect its own NICs revenues,” the Commission notes. “This could hamper innovation to the detriment of customers.” Another concern is that Broadcom could start bundling VMware’s virtualization software with its own mainframe and security tools, a move that would reduce choice in the market.

As one of the larger tech acquisitions of 2022, Broadcom’s bid to buy VMware was certain to draw scrutiny. The European Commission won’t necessarily block the deal, but the investigation could significantly delay the transaction and force concessions out of Broadcom. With today’s announcement, the Commission has 90 working days or until May 11th, 2023, to make a decision. If the deal were to fall through, it would be a bitter repeat of Broadcom’s 2018 attempt to buy chipmaker Qualcomm. While the circumstances and concerns were different, the company was forced to abandon the takeover after the Trump administration blocked the transaction.

Lionel Messi’s World Cup celebration is now the most-liked post on Instagram

Soccer megastar Lionel Messi finally secured the one prize that had eluded him during his illustrious career this weekend when Argentina won the World Cup. Afterward, the best player of the 21st century (yeah, I said it) added another record to his resume, as he now has the most-liked post on Instagram.

Messi posted a slideshow of him and his teammates celebrating after winning the World Cup. At the time of writing, the post has more than 65.8 million likes. The previous record-holder, a stock photo of an egg, claimed the top spot in early 2019 and currently has north of 57.3 million likes.

It’s not super surprising that Messi broke the Instagram record. As The Verge points out, the majority of the most-liked posts on the platform are from footballers, with the Paris Saint-Germain forward holding eight of the top 20 spots. 

Messi is also the second most-followed person on the platform with 404 million followers. Instagram’s own official account has the most followers overall. In second place, one spot ahead of Messi, is his old rival Cristiano Ronaldo with 520 million followers. Bet he’d rather have a World Cup, though.

OpenAI releases Point-E, which is like DALL-E but for 3D modeling

OpenAI, the Elon Musk-founded artificial intelligence startup behind popular DALL-E text-to-image generator, announced on Tuesday the release of its newest picture-making machine POINT-E, which can produce 3D point clouds directly from text prompts. Whereas existing systems like Google’s DreamFusion typically require multiple hours — and GPUs — to generate their images, Point-E only needs one GPU and a minute or two.

There's a corgi in a santa hat, an
OpenAI

3D modeling is used across a variety industries and applications. The CGI effects of modern movie blockbusters, video games, VR and AR, NASA’s moon crater mapping missions, Google’s heritage site preservation projects, and Meta’s vision for the Metaverse all hinge on 3D modeling capabilities. However, creating photorealistic 3D images is still a resource and time consuming process, despite NVIDIA’s work to automate object generation and Epic Game’s RealityCapture mobile app, which allows anyone with an iOS phone to scan real-world objects as 3D images. 

Text-to-Image systems like OpenAI’s DALL-E 2 and Craiyon, DeepAI, Prisma Lab’s Lensa, or HuggingFace’s Stable Diffusion, have rapidly gained popularity, notoriety and infamy in recent years. Text-to-3D is an offshoot of that research. Point-E, unlike similar systems, “leverages a large corpus of (text, image) pairs, allowing it to follow diverse and complex prompts, while our image-to-3D model is trained on a smaller dataset of (image, 3D) pairs,” the OpenAI research team led by Alex Nichol wrote in Point·E: A System for Generating 3D Point Clouds from Complex Prompts, published last week. “To produce a 3D object from a text prompt, we first sample an image using the text-to-image model, and then sample a 3D object conditioned on the sampled image. Both of these steps can be performed in a number of seconds, and do not require expensive optimization procedures.”

Point-E
OpenAI

If you were to input a text prompt, say, “A cat eating a burrito,” Point-E will first generate a synthetic view 3D rendering of said burrito-eating cat. It will then run that generated image through a series of diffusion models to create the 3D, RGB point cloud of the initial image — first producing a coarse 1,024-point cloud model, then a finer 4,096-point. “In practice, we assume that the image contains the relevant information from the text, and do not explicitly condition the point clouds on the text,” the research team points out. 

These diffusion models were each trained on “millions” of 3d models, all converted into a standardized format. “While our method performs worse on this evaluation than state-of-the-art techniques,” the team concedes, “it produces samples in a small fraction of the time.” If you’d like to try it out for yourself, OpenAI has posted the projects open-source code on Github.

Apple adds M1 Mac desktops and Studio Display to the Self Service Repair program

Apple has expanded its self-repair program once again. As noted by Six Colors and The Verge, folks in the US can now try to fix issues with the M1 iMac, M1 Mac mini, Mac Studio and Apple Studio Display themselves with genuine parts, repair manuals and tools.

The self-repair program is designed for those who have the time, patience, skills and confidence to carry out fixes at home, rather than taking their busted device to an Apple Store or third-party repair shop, or shipping it to Apple. You can buy all the parts and rent the tools you need, but at checkout you’ll need to enter a code from the relevant manual to show that you’ve actually read the document.

Apple debuted the Self Service Repair program in April by offering manuals and parts for select iPhone models in the US. Since then, it has expanded the program to Mac laptops and more territories. 

Apple introduced the program ahead of right-to-repair rules likely coming into force in the US and Europe. In 2021, President Joe Biden signed an executive order focused on bolstering competition in the US economy, partly in the tech sector. Among other things, it urged the Federal Trade Commission to ban “anticompetitive restrictions on using independent repair shops or doing DIY repairs of your own devices and equipment.” 

The agency has taken a stronger stance on such issues. In July, it announced settlements with three companies (including Weber and Harley-Davidson), which it accused of threatening to unlawfully void warranties if consumers used third-party repair parts or independent repair shops.

Congress attempts to ban TikTok on government devices as part of $1.7 trillion spending bill

After obtaining Senate approval last week, the No TikTok on Government Devices Act could become law thanks to the $1.7 trillion spending bill federal lawmakers unveiled early Tuesday morning. In addition to allocating more funding for Ukraine and earmarking $40 billion for natural disaster recovery efforts across the US, the sprawling 4,155-page bill includes provisions that would prohibit the use of TikTok on government-owned phones and other devices.

While some Republican lawmakers are pushing for a country-wide ban on TikTok, the spending bill stops short of prohibiting all government use of TikTok. If passed, the legislation would order the General Services Administration and Office of Management and Budget to create guidelines for staff at executive agencies to remove TikTok from government-owned devices by mid-February. The draft legislation allows congressional staff and elected officials to continue using the app. It also carves out some exceptions for law enforcement agents and officials.

“We’re disappointed that Congress has moved to ban TikTok on government devices — a political gesture that will do nothing to advance national security interests — rather than encouraging the Administration to conclude its national security review,” TikTok spokesperson Brooke Oberwetter told Engadget.

“The agreement under review by [The Committee on Foreign Investment in the United States] will meaningfully address any security concerns that have been raised at both the federal and state level. These plans have been developed under the oversight of our country’s top national security agencies — plans that we are well underway in implementing — to further secure our platform in the United States, and we will continue to brief lawmakers on them.”

The proposed ban comes after at least 11 states, including Georgia and South Dakota, prohibited TikTok on government-owned devices. Political concerns over TikTok hit a high earlier this month after FBI Director Chris Wray said China could use the app to collect user data. TikTok has tried to address those concerns. As of June, the app began routing all domestic traffic through Oracle servers in the US. At the same time, TikTok and parent company ByteDance pledged to delete US user data from their own data servers in the US and Singapore.