The big picture: The world already knew that Musk's shoot-from-the-hip Twitter posting style recalled the former president's. Now it appears that Musk's management approach owes a debt to the former president as well.

Musk's first days as Twitter CEO have such a familiar feel because the world's richest man is leading his new company from the same playbook Trump used as he tried to change the U.S. government's direction after his 2017 inauguration:

•     Rely on an inner circle chosen for loyalty more than expertise;

•     seize and hold the public's attention by rolling out new proposals and ideas on Twitter first before they've been widely vetted internally;

•     Keep the existing organization in a state of uncertainty and fear.

Why it matters: Trump achieved only a fraction of the goals he set because it's nearly impossible to turn around a gigantic organization by posting tweets.

• Twitter, the company, is much smaller than the federal bureaucracy. But Musk is likely to get no further than Trump did if his leadership-by-tweet isn't accompanied by tenacious and coherent internal planning.

• That could be difficult for someone who changes his mind in public a lot and is also the CEO of four other companies.

Between the lines: For Trump-detesters, watching Musk's leadership approach in action can feel like a reopened wound or a recurring nightmare.

• Trump fans angered by Twitter's ban on the former president have been cheered by Musk's promises to restore "free speech." But Trump says he will continue to use his own network at Truth Social for his online communiques, and Musk says it will be weeks before anyone's ban is lifted.

Yes, but: There are big differences between Trump and Musk, too.

• Musk really is the world's richest man, whereas the extent of Trump's wealth has always been a subject of dispute, further clouded by his secrecy.

• Musk is a veteran leader in business with success running companies like Tesla and SpaceX, whereas Trump took the presidency with zero government experience.

• Musk has no limit on his term and no reelection schedule, nor does he face a threat of impeachment — though he still needs to win over users, court advertisers and show his investors positive results.

• Also unlike Trump, Musk actually replies to other users on Twitter and seems to enjoy the site.

Be smart: Tweets move at the speed of electrons but actual change to products and organizations happens more slowly.

• Musk fans who expected a Twitter free-for-all in the hours after Musk tweeted "the bird is freed" last Thursday have been disappointed in the new owner's announcement this week that he won't change content moderation policies until he has formed a new council.

• It's still unclear when Musk will spring an expected big round of layoffs.

• Since taking the Twitter helm, he has feverishly tweeted out new product ideas, from charging monthly fees for "verified user" checks to relaunching the Vine short-video service.

• Other ideas that have leaked include a "paywalled video" feature that would let users charge others to view videos (potentially including pornography) and give Twitter a cut.

• At the same time these ideas received public airing, communications from the top to workers inside Twitter have been minimal, sources inside the company have told Axios throughout the week.

Our thought bubble: Until Musk's team announces concrete plans or ships actual features, the rest of us are well advised to watch what they do, not what they say.

• In this, too, Musk's management resembles Trump's presidency, which was much better at announcing broad new policies than codifying and implementing them.

What to watch: Trump made a practice of ignoring his own advisers and cutting out experts. It remains to be seen who will have Musk's ear and how he will make decisions at Twitter.

• Trump also used the White House as an asset to leverage advantages for his other assets (most famously, hotels and golf courses). Musk is also known for blurring the boundaries between the companies he owns.

• He's already drafted Tesla engineers to assess the work of Twitter's developers. Critics have suggested that Tesla's business interests in China could sway Musk's decisions about content on Twitter.

The bottom line: Relying on tweets to try to impose change on a company or a country can control the narrative and speedily draw the public's attention to new ideas. It can also undermine the process of turning those ideas into reality.

• When the tweeting stops, a leader still needs managers and employees to go and make things happen. To do that, they have to understand and embrace a plan — and maniacal public improvising only gets in the way.

Source

The Asia-Africa-Europe-1 internet cable travels 15,500 miles along the seafloor, connecting Hong Kong to Marseille, France. As it snakes through the South China Sea and toward Europe, the cable helps provide internet connections to more than a dozen countries, from India to Greece. When the cable was cut on June 7, millions of people were plunged offline and faced temporary internet blackouts.

The cable, also known as AAE-1, was severed where it briefly passes across land through Egypt. One other cable was also damaged in the incident, with the cause of the damage unknown. However, the impact was immediate. “It affected about seven countries and a number of over-the-top services,” says Rosalind Thomas, the managing director of SAEx International Management, which plans to create a new undersea cable connecting Africa, Asia, and the US. “The worst was Ethiopia, that lost 90 percent of its connectivity, and Somalia thereafter also 85 percent.” Cloud services belonging to Google, Amazon, and Microsoft were all also disrupted, subsequent analysis revealed.

While connectivity was restored in a few hours, the disruption highlights the fragility of the world’s 550-plus subsea internet cables, plus the outsized role Egypt and the nearby Red Sea have in the internet’s infrastructure. The global network of underwater cables forms a large part of the internet’s backbone, carrying the majority of data around the world and eventually linking up to the networks that power cell towers and Wi-Fi connections. Subsea cables connect New York to London and Australia to Los Angeles.

Sixteen of these submarine cables—which are often no thicker than a hosepipe and are vulnerable to damage from ships’ anchors and earthquakes—pass 1,200 miles through the Red Sea before they hop over land in Egypt and get to the Mediterranean Sea, connecting Europe to Asia. The last two decades have seen the route emerge as one of the world’s largest internet chokepoints and, arguably, the internet’s most vulnerable place on Earth. (The region, which also includes the Suez Canal, is also a global choke point for shipping and the movement of goods. Chaos ensued when the container ship Ever Given got wedged in the canal in 2021.)

“Where there are chokepoints, there are single points of failure,” Nicole Starosielski, an associate professor of media, culture, and communication at New York University and an author on submarine cables. “Because it's a site of intense concentration of global movement, that does make it more vulnerable than many places around the world.”

The area has also recently gained attention from the European Parliament, which in a June report highlighted it as a risk for widespread internet disruption. “The most vital bottleneck for the EU concerns the passage between the Indian Ocean and the Mediterranean via the Red Sea because the core connectivity to Asia runs via this route,” the report says, flagging extremism and maritime terrorism are risks in the area.

Pyramid Scheme

Look at Egypt on a map of the world’s subsea internet cables and it immediately becomes clear why internet experts have been concerned about the area for years. The 16 cables in the area are concentrated through the Red Sea and touch land in Egypt, where they make a 100-mile journey across the country to reach the Mediterranean Sea. (Cable maps don’t show the exact locations of cables.)

It has been estimated that around 17 percent of the world’s internet traffic travels along these cables and passes through Egypt. Alan Mauldin, the research director of telecoms market research firm TeleGeography, says last year the region had 178 terabits of capacity, or 178,000,000 Mbps—the US has median home internet speeds of 167 Mbps.

Egypt has become one of the internet’s most prominent chokepoints for a few reasons, says Doug Madory, director of internet analysis at monitoring firm Kentik. Primarily, its geography contributes to the concentration of cables in the area. Passing through the Red Sea and across Egypt is the shortest (mostly) underwater route between Asia and Europe. While some intercontinental internet cables travel across land, it is generally safer for them to be placed at the bottom of the sea where it is harder for them to be disrupted or snooped upon.

Going through Egypt is one of the only practical routes available. To the south, cables that pass around Africa are longer; while to the north, only one cable (the Polar Express) travels above Russia. “Every time someone tries to draw up an alternative route, you end up going through Syria or Iraq or Iran or Afghanistan—all these places have a lot of issues,” Madory says. The JADI cable system that bypassed Egypt was shut down due to Syria’s civil war, Madory says, and it has not been reactivated. In March this year, another cable avoiding Egypt was severed as a consequence of Russia’s full-scale invasion of Ukraine.

Disruption also happens around the Red Sea itself. “The Red Sea is a fairly shallow body of water, and there's been historically a lot of cable cuts there as a result,” Madory says. In 2013, the Egyptian navy arrested three people who were allegedly cutting internet cables in the region. Other nearby cables also faced outages in the same year. The region isn’t the only cable choke point around the world. The UK, Singapore, and France are all key internet connection points, with the Strait of Malacca, near Singapore, being another chokepoint. “The Malacca Strait is also a problem area, but I don't think it's as bad as Egypt,” SAEx’s Thomas says.

Mauldin says the Egyptian region can be considered a single point of failure due to the number of cables in one place. However, there are reasons beyond costs to have multiple cables pass through the Red Sea. “There are values in concentration because you want networks to connect to each other,” Mauldin says. “At the same time, you have to balance that with the need to have diversity [in routes].”

When the submarine cables appear above land, at the very north of the Red Sea and the Gulf of Suez, Telecom Egypt, the country’s main internet provider, is involved. The company charges cable owners for running cables across the country. The cables travel across land among multiple different routes—and do not go in the Suez Canal—so there is variation in how they are spread out.

“It gives Egypt a lot of power in terms of telecommunications negotiations,” Starosielski says. A recent report from Data Center Dynamics, which covers Egypt’s “stranglehold” on the submarine cable industry, cites unnamed industry sources who claim Telecom Egypt charges “extortionate” fees for its services. (Neither Telecom Egypt, Egypt’s Ministry of Communications and Information Technology, nor the National Telecommunications Regulatory Authority replied to WIRED’s request for comment.)

Cable Ties

Subsea cables are relatively fragile and easily damaged. Every year, there are more than 100 incidents where the cables are cut or damaged. The majority of these are caused by shipping or environmental damage. However, in recent months, there have been growing concerns about sabotage. Following the Nord Stream gas pipeline explosions, governments around the world have pledged to better protect underwater infrastructure and subsea cables. The UK has also claimed Russian submarines have been monitoring cables landing in the country.

Despite the dangers, the internet is built on resilience. It isn’t easy to take down large parts of the internet. Companies that send data through subsea internet cables don’t just use one cable and will have space on multiple cables. If one cable fails, traffic is eventually rerouted through others. (In some areas, such as Tonga, where there is only one cable, cuts can have devastating impacts.) The need for redundancy is why Google, Facebook, and Microsoft have been spending hundreds of millions on their own subsea internet cables in recent years.

When it comes to Egypt and the Red Sea, there are limited options, and more cables are often the answer. While Elon Musk’s Starlink has popularized satellite internet, this kind of system doesn’t offer a replacement for underwater cables. Satellites are used for providing connectivity in rural locations or as emergency backups, but they can’t replace physical infrastructure entirely. “They aren't going to handle carrying hundreds of terabits between continents. That's only cables,” Mauldin says. (Satellite systems also rely on wired connections to connect to the internet.)

That’s all the more reason to further protect routes around Egypt. Mauldin says extra landing sites are being built along Egypt’s shore, such as at Ras Ghareb, to allow cables to dock in different locations. Egyptian telecom authorities are also building a new land-based route for cables alongside the Suez Canal—it is believed cables will be housed in concrete ducts to protect them.

However, the biggest effort to bypass Egypt comes from Google. In July 2021, the company announced it is creating the Blue-Raman subsea cable that will connect India to France. The cable travels through the Red Sea, but instead of crossing land in Egypt, it reaches the Mediterranean via Israel. Google did not respond to a request for an interview, but the cable likely comes with its own geopolitical challenges. Google has split the cable into two separate projects: Blue runs through Israel and into Europe, while Raman connects to Saudi Arabia before passing along to India. (Israel and Saudi Arabia have a complex relationship.)

Mauldin says the new route, which is expected to be ready in 2024, is likely to set a precedent for more cables to travel through Israel over time. Once one cable is built, others will come. “It's hard to turn a proposal or just a good idea into a reality,” Madory adds. “Unless you're Google and you have limitless funds to do these things.”

Elsewhere, Thomas says the proposed SAEx cable, which she overseas, plans to bypass Europe and connect Africa to the Americas and Singapore. Thomas says the route will be an “all wet” network and claims it manages to avoid many of the current risks. “Look at all this piracy, you've got all of your anchors, and you've got high-risk countries and war zones,” Thomas says. “Our cable and Blue Ramen are unlikely to replace Egypt, we only provide alternatives.”

Ultimately, Egypt is always going to be at the center of Europe and Asia’s internet connections. Geography can’t be changed. However, Mauldin says, more should be done to protect the world’s underwater internet cables, as everyone relies on them. “It's super important for national security, for the economy, to keep this stuff up and running.”

The Most Vulnerable Place on the Internet

(May require free registration to view)

ASRock

 

Adding a secondary monitor to your setup is an easy office hack for boosting productivity. But not every desk or office has room for another display, even a small portable monitor. That's why we're intrigued by a new ASRock kit, as niche as it may be.

 

ASRock's 13.3-inch Side Panel Kit, spotted by Tom's Hardware on Tuesday, includes a 13.3-inch IPS display meant to attach to the inside of a desktop PC's side panel. This is a convenient design for people who are low on space but still keep their tower on a desk. The kit provides a 1920×1080 display with a 60 Hz refresh rate, 300 nits of brightness, and an 800:1 contrast ratio, making its specs comparable to dedicated portable monitors in the $200 range, such as Lenovo's ThinkVision M14 (although, when we tested the M14, we recorded a notably higher contrast ratio, 1,064:1, than ASRock's panel claims).

 

The installation kit comes with two long and two short brackets, plus double-sided tape strips and five cable clips.

ASRock

 

There are some major caveats, though.

 

First, ASRock's product page says you have to attach the screen to a transparent glass side panel.

 

The Taiwan-based company, which is known for selling DIY PC parts, also says the 13.3-inch Side Panel Kit will work on "most" ATX, Mini-ITX, and Micro-ATX PC cases. As any PC builder will assure you, compatibility with "most" PC cases does not mean compatibility with them all.

 

Further, for power and data, ASRock's display must connect to a motherboard via Embedded DisplayPort (eDP), a rarity among desktops, considering it was made for embedded displays, such as those in laptops. According to Tom's Hardware, this limits the display to people with an ASRock motherboard, and there are fewer than 10 ASRock motherboards with eDP. And because the connector includes power as well, there are no readily available adapters, either.

 

Those are enough caveats to make this product impractical for most users, but it's still a neat idea. The kit provides a secondary monitor without requiring additional space, providing a convenient place to display things like a Slack chat or to stream video while you work. On its product page, ASRock says the display could be used to watch video guides or streaming for gamers.

 

The kit also provides flexibility. Theoretically, you could use it on multiple monitors, assuming everything's still compatible. That's different from iBuyPower's Snowblind S PC case, which has a 19-inch, 1280x1024 transparent LCD display on the side. However, the transparent panel doesn't appear to have the strongest image quality and is also meant to showcase PC components, which are encouraged to be white or silver, to help images on the display pop.

 

Today, the options for desktop side-panel monitors are limited, and available products carry caveats galore. ASRock's 13.3-inch Panel Kit isn't suitable for most. But it's nice to see companies exploring creative ways to bring more pixels to cramped spaces.

 

Laptop users already have more options. Take the NexPad, for instance. Announced in October, it's a 12-inch IPS display that's meant to sit on top of a laptop display for more vertical screen space. While ASRock's kit isn't as simple as the NexPad, perhaps it will help pave the way for more versatile solutions for desktop users.

 

We've reached out to ASRock about availability and pricing.

 

No room for another monitor? Use your desktop’s side panel instead

Musk calls Twitter verification system “bullsh—,” announces $8 monthly charge

AMD’s profits have cratered as the PC and crypto miner markets slow down

Audio-Technica's 2022 Sound Burger.

Audio-Technica

 

Vinyl is far from dead. Just ask your most nostalgic audiophile friend, old-school uncle, or even Taylor Swift. And for those who prefer their vinyl with some modern-day comforts, there's the revival of the Sound Burger, Audio-Technica's portable record player.

 

According to the Recording Industry Association of America's Mid-Year 2022 report (PDF), vinyl is continuing a trend initiated in 2020, when vinyl revenue surpassed CD revenue—a state of affairs we haven't seen since 1986, according to the RIAA [PDF]. The RIAA reported that vinyl revenue grew 22 percent in the first half of this year to $570 million, outpacing CDs ($200 million) and representing physical music's largest revenue share.

 

Japanese audio brand Audio-Technica has seemingly taken note of this trend and has decided to rerelease its Sound Burger portable record player. The product is one of several that the company is releasing to celebrate its 60th birthday. Based on a plate on the player's side, the company is producing just 7,000 units.

 

Debuting in 1980, the original Sound Burger AT727 offered a way for people to listen to their 33-1/3 and 45 vinyl records outside of their homes. There were some caveats, though; records stuck precariously out of the unit and, as noted by SlashGear, the player needed to sit on a flat surface during operation.

 

The new Sound Burger AT-SB2022 looks much like the old one, but Audio-Technica added Bluetooth 5.2 support, so you can listen to your vinyl through wireless headphones. There's also a 3.5 mm jack this time around instead of stereo RCA outputs, but Audio-Technica is also including a dual RCA adapter.

 

Additionally, there's no longer a need for clunky batteries, as the refreshed portable record player charges over USB-C. Audio-Technica claims the record play will last for about 12 hours before needing a charge, which it says will take around 12 hours to complete.

 

As noted by Digital Trends, the original Sound Burger, also known as Mister Disc in some geographies, required user maintenance to make sure records continued spinning at the proper speed. But Audio-Technica's 2022 record player has a DC servo motor driving its belt-drive system for "stable rotation," the record player's product page says.

 

"Designed for both portability and stability, the tonearm employs a dynamic balance system in which stylus pressure is applied through a spring," the page says. Audio-Technica says its ATN3600L stylus works with the new record player.

 

Other specs for the Sound Burger include a 50 dB output and a 20–20,000 Hz transmission band.

 

The 2022 Sound Burger is a niche product for enthusiasts who want to embrace their vinyl passion on the go. The original product has maintained somewhat of a cult following, with YouTubers and the like still touting the unique premise and surprisingly impressive sound quality.

 

We wouldn't be surprised to see the 2022 Sound Burger ending up on sites like eBay after stock runs out. You can currently find 1980s Sound Burgers being resold in the $400–$550 range.

 

Audio-Technica is selling the 2022 Sound Burger for $200.

 

Sound Burger portable record player returns from the ’80s with Bluetooth, USB-C

The multi-billionaire will be its chief executive after buying the company last week, ending months of back and forth over the $44bn (£38.3bn) deal.

He has moved quickly to put his mark on the firm, which is used by politicians and journalists around the world.

The reforms he is contemplating include changes for how Twitter verifies accounts, as well as job cuts.

The Washington Post has reported that a first round of cuts is under discussion that could affect 25% of the company's staff.

Twitter did not respond to a request for comment from the BBC on the report.

Top executives have already been removed, as Mr Musk brings in high profile allies to the company.

The latest move will mean that he is now chief executive of three companies. Along with taking the top role at Twitter, Mr Musk is chief executive of electric car maker Tesla and rocket company SpaceX.

However, he has indicated that his position at the social media company may be temporary.

'Singular solution'

Twitter co-founder Jack Dorsey has rolled his entire stake of 18m shares, worth almost $978m at the buyout price of $54.20, into the new private company, according to a filing with the US Securities and Exchange Commission on Monday.

Mr Dorsey, who left Twitter's board in May, supported Mr Musk's purchase of the firm.

"Elon is the singular solution I trust. I trust his mission to extend the light of consciousness," he said in a tweet after the takeover was approved by the company's former management team.

Meanwhile, technology investor Jason Calacanis who changed his Twitter bio to "chief meme officer", said he was "hanging out at Twitter a bit ... during the transition".

He took to the social media site to solicit opinions on a range of topics from advertising to video.

Mr Calacanis also asked how much people would pay to be verified, amid reports that Twitter could charge people $20 per month to keep the blue ticks that indicate verified accounts.

Later on Tuesday, responding to a Twitter user calling for the platform to purge accounts that had been inactive for more than a year, Mr Musk said "definitely".

Mr Musk, who also leads electric car maker Tesla, changed his profile details on the social media site to "Chief Twit" after the takeover. It has since been updated to "Twitter Complaint Hotline Operator".

He is now the sole director of Twitter, another filing on Monday showed.

The nine ousted directors include former chairman of the board Bret Taylor and former chief executive Parag Agrawal.

Baroness Martha Lane Fox, currently president of the British Chambers of Commerce, who is among the board members to be removed under the merger agreement, declined to comment when approached by the BBC.

'National-security issue'

Mr Musk's takeover has drawn widespread scrutiny, as he signals plans to overhaul how Twitter has moderated the spread of information on its platform, including from sources such as state media, politicians and celebrities.

Mr Musk said the company would create a new council to govern those decisions and that no changes would occur yet.

On Monday, Senator Chris Murphy, a Democrat, said that he had asked the government to review the national security implications of the deal, given the large stake in the company held by firms tied to Saudi Arabia, which has an increasingly tense relationship with the US.

"We should be concerned that the Saudis, who have a clear interest in repressing political speech and impacting US politics, are now the second-largest owner of a major social media platform," Mr Murphy wrote on Twitter.

"There is a clear national-security issue at stake and CFIUS [the Committee on Foreign Investment in the United States] should do a review."

Mr Musk financed his takeover with his own money, a group of other investors and roughly $13bn in debt financing.

Analysts say that the increase in debt is likely to constrain the firm, which has struggled to expand its user base and not turned a profit in years.

Together, Prince Alwaleed bin Talal bin Abdulaziz Al Saud and the Kingdom Holding Company are the second largest investor in the newly private company, according to a filing with the US government.

Prince Alwaleed, who was a major Twitter investor prior to the Musk deal, has tweeted the stake aligns with Kingdom Holding Company's "long-term investment strategy".

The BBC is not responsible for the content of external sites.

View original tweet on Twitter

Source

What's your favorite ice cream flavor? You might say vanilla or chocolate, and if I asked why, you’d probably say it’s because it tastes good. But why does it taste good, and why do you still want to try other flavors sometimes? Rarely do we ever question the basic decisions we make in our everyday lives, but if we did, we might realize that we can’t pinpoint the exact reasons for our preferences, emotions, and desires at any given moment.

There's a similar problem in artificial intelligence: The people who develop AI are increasingly having problems explaining how it works and determining why it has the outputs it has. Deep neural networks (DNN)—made up of layers and layers of processing systems trained on human-created data to mimic the neural networks of our brains—often seem to mirror not just human intelligence but also human inexplicability.  

Most AI systems are black box models, which are systems that are viewed only in terms of their inputs and outputs. Scientists do not attempt to decipher the “black box,” or the opaque processes that the system undertakes, as long as they receive the outputs they are looking for. For example, if I gave a black box AI model data about every single ice cream flavor, and demographic data about economic, social, and lifestyle factors for millions of people, it could probably guess what your favorite ice cream flavor is or where your favorite ice cream store is, even if it wasn’t programmed with that intention. 

These types of AI systems notoriously have issues because the data they are trained on are often inherently biased, mimicking the racial and gender biases that exist within our society. The haphazard deployment of them leads to situations where, to use just one example, Black people are disproportionately misidentified by facial recognition technology. It becomes difficult to fix these systems in part because their developers often cannot fully explain how they work, which makes accountability difficult. As AI systems become more complex and humans become less able to understand them, AI experts and researchers are warning developers to take a step back and focus more on how and why a system produces certain results than the fact that the system can accurately and rapidly produce them.

“If all we have is a ‘black box’ it is impossible to understand causes of failure and improve system safety,” Roman V. Yampolskiy, a professor of Computer Science at the University of Louisville, wrote in his paper titled “Unexplainability and Incomprehensibility of Artificial Intelligence.” “Additionally, if we grow accustomed to accepting AI’s answers without an explanation, essentially treating it as an Oracle system, we would not be able to tell if it begins providing wrong or manipulative answers.”

Black box models can be extremely powerful, which is how many scientists and companies justify sacrificing explainability for accuracy. AI systems have been used for autonomous cars, customer service chatbots, and diagnosing disease, and have the power to perform some tasks better than humans can. For example, a machine that is capable of remembering one trillion items, such as digits, letters, and words, versus humans, who on average remember seven in their short-term memory would be able to process and compute information at a much faster and improved rate than humans. Among the different deep learning models include generative adversarial networks (GANs), which are most often used to train generative AI models, such as text-to-image generator MidJourney AI. GANs essentially pit AI models against each other to do a specific task; the "winner" of each interaction is then pitted against another model, allowing the model to iterate itself until it becomes very good at doing that task. The issue is that this creates models that their developers simply can't explain.  

“I think in a lot of cases, people look to these black box models as a response to lack of resources. It would be very convenient to have an automated system that can produce the kinds of outputs they're looking for from the kind of inputs they have,” Emily M. Bender, a Professor of Linguistics at the University of Washington told Motherboard. “If you have a dataset consisting of such inputs and outputs, it's always possible to train a black box system that can produce outputs of the right type—but often much, much harder to evaluate whether they are correct.

Furthermore, there are lots of cases where it's impossible to make a system where the outputs would be reliably correct because the inputs just don't contain enough information.”

When we put our trust in a system simply because it gives us answers that fit what we are looking for, we fail to ask key questions: Are these responses reliable, or do they just tell us what we want to hear? Whom do the results ultimately benefit? And who is responsible if it causes harm?

“If business leaders and data scientists don’t understand why and how AI calculates the outputs it does, that creates potential risk for the business. A lack of explainability limits AI’s potential value, by inhibiting the development and trust in the AI tools that companies deploy,” Beena Ammanath, Executive Director of the Deloitte AI Institute, told Motherboard.

These people do not exist. Image: Arxiv

“The risks are that the system may be making decisions using values we disagree with, such as biased (e.g. racist or sexist) decisions. Another risk is that the system may be making a very bad decision, but we cannot intervene because we do not understand its reasoning,” Jeff Clune, an Associate Professor of Computer Science at the University of British Columbia, told Motherboard.

AI systems are already deeply entrenched with bias and are constantly reproducing such bias in their output without developers understanding how. In a groundbreaking 2018 study called “Gender Shades,” researchers Joy Buolamwini and Timnit Gebru found that popular facial recognition systems most accurately detected males with lighter skin and had the highest errors detecting females with darker skin. Facial recognition systems, which are skewed against people of color and have been used for everything from housing to policing, deepen pre-existing racial biases by determining who is more likely to get a house or be identified as a criminal, for example. Predictive AI systems can also guess a person’s race based on X-rays and CT scans, but scientists have no idea why or how this is the case. Black and female patients are less likely to receive an accurate diagnosis from automated systems that analyze medical images, and we're not sure why. These are just a few examples of how viewing AI-generated results as concrete data without understanding the system’s potential biases creates rippling societal consequences.

At the same time, some experts argue that simply shifting to open and interpretable AI models—while allowing greater visibility of these processes—would result in systems that are less effective.

“There are many tasks right now where black box approaches are far and away better than interpretable models,” Clune said. “The gap can be large enough that black box models are the only option, assuming that an application is not feasible unless the capabilities of the model are good enough. Even in cases where the interpretable models are more competitive, there is usually a tradeoff between capability and interpretability. People are working on closing that gap, but I suspect it will remain for the foreseeable future, and potentially will always exist.”

Though there is already a subset of AI known as Explainable AI (XAI), the general techniques it promotes are often diminutive and inaccurate in encompassing the true breadth of the processes, and AI developers are not incentivized to follow this model. The issue with explainability has to do with the fact that because AI systems have become so complex, blanket explanations only increase the power differential between AI systems and their creators, and AI developers and their users. In other words, seeking to add explainability after an AI system is already in place makes it more difficult to approach than if you start with it.   

“Maybe the answer is to abandon the illusion of explanation, and instead focus on more rigorously testing the reliability, biases, and performance of models, as we try to do with humans,” Clune said.

In recent years there's been a small but real push by some in the industry to develop "white-box models," which are more transparent and whose outputs can be better explained (it's worth mentioning that the white-box / black-box terminology is in itself part of a long history of racially coded terms in science; researchers have pushed to change "blacklist" to "blocklist," for example.) White-box models, nonetheless, are a relatively new branch of AI research that are seeking to make AI more explainable.

AI researchers say giving users who are impacted by a certain system a bigger role in participating in the development process is an important first step in making AI systems that are more transparent and adequately represent user needs.

 “A lot of the explanations that people treat as explanations really aren't. They are reductive, they are written to the interests of the developers and what the developers think are important to explain, rather than what the user needs,” Os Keyes, a PhD Candidate at the University of Washington's Department of Human Centered Design & Engineering, told Motherboard. “Arguably, I'd say there are two big changes to AI that would be necessary to change this state of affairs, and the first is that, ultimately, this is in part a problem of the massive gap in practice between developers and actual users.”

“Broader participation in not just building the system, but also asking, what questions are interesting, what things need to be possible for this to really be explainable,” Keyes added. “That would make a massive difference.”

Ammanath agrees that some of the best practices in fostering explainability include tailoring explanations and reporting to the people who will engage with or be impacted by the automated systems. Along the same line, she said, developers need to first identify the needs and priorities of the people who will be most affected.

A more challenging problem is that many AI systems are designed for the concept of universalism—the idea that “[a] system is good if it works everywhere for everyone at all times” Keyes explained. “But the problem is that that's not how reality works, different people are going to need different explanations of different things. If we really want AI to be more explainable, we actually have to really, fundamentally change how we imagine and how developers imagine AI.”

In other words, if you build explainable AI with a one-size-fits-all design process, “you end up with something where it has explanations that only make sense to one group of people who are involved in the system in practice,” said Keyes. “The internal change is a [much] broader set of involvement in deciding explainable to whom, and what does explainable mean.”

Keyes’ call for more localized AI systems and their concern about the universality of AI models is what researchers have been warning us about in the past few years. In a 2021 paper co-authored by Bender and Gebru, who was terminated from Google for publishing this research, the authors argue that training AI models with big data make it difficult to audit for embedded biases. They wrote that big data also fails to represent populations that have less access to the internet and “overrepresents younger users and those from developed countries.”

“If we orient knowledge and AI around big data, then we're always going to bias towards those who have the resources to spin up a thousand servers, or those who have the resources to, you know, get a billion images and train them,”  said Keyes. “There's something fundamentally, I'd say undemocratic, but I'd also say just badly incentivized in that.”

“The question first is, what are the conditions under which AI is developed? Who gets to decide when it's deployed? And with what reasoning? Because if we can't answer that, then all good intentions in the world around how do we live with that [AI] are all screwed,” they added. “[I]f we're not participating in those conversations, then it's a losing game. All you can do is have something that works for people with power, and silences the people who don't.”

Debiasing the datasets that AI systems are trained on is near impossible in a society whose Internet reflects inherent, continuous human bias. Besides using smaller datasets, in which developers can have more control in deciding what appears in them, experts say a solution is to design with bias in mind, rather than feign impartiality.

“The approach I currently think is the best is to have the system learn to do what we want it to,” Clune said. “That means it tries to do what we ask it to (for example, generate pictures of CEOs) and if it does not do what we like (e.g. generating all white males), we give it negative feedback and it learns and tries again. We repeat that process until it is doing something we approve of (e.g. returning a set of pictures of CEOs that represents the diversity in the world we want to reflect). This is called ‘reinforcement learning through human feedback’, because the system is effectively using trial and error learning to bring its outputs in line with our values. It is far from perfect, and much more research and innovation is required to improve things.”

As we continue to negotiate where and how AI should be used, there are many things to consider before we start letting AI hire people or decide who to give loans to.

“I think it is absolutely critical to start by keeping in mind that what gets called ‘AI’ isn't any kind of autonomous agent, or intelligence, or thinking entity,” Bender said. “These are tools, which can serve specific purposes. As with any other tools, we should be asking: How well do they work? How suited are they to the task at hand? Who are they designed to be used by? And: How can their use reinforce or disrupt systems of oppression?”

Source

Soraton Charehkphunpol, a Thai cook scraping by in a restaurant in Bangkok, couldn’t resist the offer to earn more than double his $470 monthly salary.

Late one evening last year, after another grueling shift, Soraton answered the ad. Within hours, a man arrived at his door to take him to the Cambodian border city of Poipet.

It wasn’t long before Soraton realized he had made a mistake. Dumped in a high-rise building above a casino, he was turned over to mobsters who seized his passport and put him to work bilking gamblers on a sham sports betting app.

“The bosses said if I tried to leave they would just sell me to another gang,” said Soraton, a rail-thin 20-year-old from the central Thai city of Nakhon Pathom. “That’s when I realized I was a slave.”

In a dystopian nightmare come to life, the Cambodian government has given Chinese crime syndicates free rein to bring in tens of thousands of foreign men and women who — according to human rights organizations and their own accounts — are held captive to work in crowded cyber scam mills.

Lured by the promise of legitimate employment, they are instead forced to run online and telephone rackets targeting people around the world with gambling, money lending and romance schemes, to name a few. Other scams have included fake real estate developments and bogus initial coin offerings.

“The criminals are limited only by their imagination,” said Jason Tower, an executive at the United States Institute of Peace and an expert on transnational Chinese crime syndicates. “These are sophisticated schemes.”

Workers who meet their targets are rewarded. Those who fail are tortured, abused and sold like chattel to other gangs on private messaging apps such as Telegram. Reports of murder, depression and suicidal ideation are rampant.

The syndicates run their operations not unlike private companies trying to motivate a sales force. The big difference: Employees aren’t allowed to leave.

“Instead of getting fired for poor performance, you get physical punishments — forced push-ups and squats, tased, beaten, deprived of food, locked up in dark rooms or worse,” said Jacob Sims, country director for International Justice Mission Cambodia, a rights group that has helped rescue more than 100 victims. “On the other hand, those who consistently meet or surpass their targets are rewarded with more freedoms, food, money and control over other victims.”

The duped workers come from across Asia, including China, Vietnam, Malaysia, Taiwan and Hong Kong. Many cite the economic fallout from the pandemic as a motivating factor for taking a job in Cambodia.

More details about the harrowing conditions inside the mills have emerged in the last few months from rights activists, escapees and independent reporting from the Cambodian media outlet VOD.

In August, the Vietnamese digital media outlet VnExpress obtained dramatic footage of about 40 Vietnamese workers fleeing a converted casino in Cambodia’s Kandal province. Guards wielding sticks tried to block the escapees, most of whom plunged into the Binh Di River to swim across to Vietnam.

Other abuse victims who spoke to The Times and have returned home said they’ll never be the same.

“I’m traumatized. I’m too ashamed to even leave my house,” said Dedek Mulyana, 25, who was tortured with an electric baton and forced to run a scam in which he pretended to be a woman selling pornographic images of herself.

He fled Cambodia in August and returned to his native Indonesia. “I don’t want anyone to have to go through what I did,” he said.

Cambodian officials estimate up to 100,000 people could be involved in the industry, but have denied that it involves widespread human trafficking.

The government — which is led by Prime Minister Hun Sen, a former military commander who has been in power since 1985 — maintains that the vast majority of foreign nationals are in Cambodia willingly and that most conflicts between workers and their employers are contract disputes.

But as more details of the mob operations have come to light, international pressure has mounted on the government to crack down on the industry and free the captive foreign workers.

In July, the U.S. State Department downgraded Cambodia to the lowest tier on its human trafficking index, placing it alongside Afghanistan, Syria and North Korea — an embarrassment to a government struggling for international legitimacy even as it continues to stifle political dissent and independent journalism.

Finally, in September, Cambodian law enforcement raided dozens of compounds in the capital, Phnom Penh, and the seedy resort city of Sihanoukville. Thousands of foreigners were sent home.

Experts said the raids were mostly for show and that many workers were bused to new facilities in less visible areas of Cambodia or to neighboring Laos and Myanmar.

“It was only after incontrovertible video evidence in media reports emerged combined with a chorus of diplomatic demands for nationals to be returned that the Cambodian authorities realized something needed to be done,” said Sophal Ear, a political scientist and Cambodia specialist at Arizona State University. “Cambodia’s international standing has fallen several rungs as a result of this.”

The coziness between the Cambodian government and the criminal syndicates has focused attention not only on endemic corruption in a country still recovering from decades of civil war but also on its relationship with China.

No country in Southeast Asia boasts closer links to Beijing, which covets Cambodia’s strategic position along the South China Sea and views it as a counterbalance to Vietnam, a historical adversary.

The Chinese government made Cambodia one of the first beneficiaries of its most important foreign policy project, the Belt and Road Initiative, which was launched in 2013 and has dramatically expanded China’s global influence by loaning countries money to build ports, roads and other infrastructure.

Chinese investment has since transformed communities in Cambodia — and not always for the best.

Chinese crime figures seized on the initiative and the clamor for Chinese money. Several garnered Cambodian citizenship and got rich opening casinos targeting visitors from China, which prohibits gambling everywhere except Macau.

Some of the gangsters are wanted in China for running illegal gambling operations there. But as long as Cambodia remains a devoted ally, Chinese authorities seem unlikely to demand a crackdown that would upset the Cambodian elites profiting from their association with the Mafias.

When the COVID-19 pandemic struck in 2019, the casino owners switched their focus to cyber and telecom scams. All they needed was exploitable labor.

Rezky Rizaldi fit the bill. The 32-year-old Indonesian chef, who previously worked in hotels in Qatar and the Maldives, had been unemployed for months because of the pandemic when he came across an ad on Facebook in February seeking customer service workers for an investment company in Cambodia.

The social media platform said in a statement that it has “long prohibited” such content and has “invested significant resources” in keeping it off the site. It is not clear how often it fails to do so.

Rezky had all the prerequisite skills: proficiency in both English and Microsoft Office. Desperate for a paycheck, he took the job in Cambodia in May.

It marked the start of a four-month ordeal in which he would be trafficked three separate times between ethnic Chinese gangs in Sihanoukville and beaten and shocked with an electric prod by his captors on several occasions for failing to generate enough money from victims.

His first job was a catfishing scam. Rezky was told to create Facebook and Instagram profiles using pictures of male models. He passed the profiles to other workers trapped in the operation who tricked women into online romances and persuaded them to invest in fake cryptocurrency schemes.

Rezky said scammers targeted women in Vietnam, Thailand and Indonesia because they have emerging middle classes. The aim was to bilk each victim of at least $30,000, at which point her suitor’s profile would simply disappear.

Rezky ran afoul of his bosses when they learned he had contacted the Indonesian Embassy seeking help. They punished him by trying to extort money from his friends and associates, concocting a story that he had accumulated gambling debts.

When no one responded, they tortured Rezky for three days and put him up for sale on a messaging app. Rezky, Rizaldi, who was devoid of emotion as he flatly recounted his horror, said he saw photos of other captives on the app, including one bloodied man with several fingers severed.

One rainy July evening when security guards were distracted, Rezky fled the compound where he had been operating another scam — selling fake World Cup tickets — and made his way to the Indonesian Embassy in Phnom Penh. It took him another month to get home.

Soraton, the Thai captive, was spared violent reprisals from his various minders, who wore luxury brands head to toe like team uniforms.

Supervisors at one scam dressed in Gucci, while those at another were outfitted in Louis Vuitton.

It was after he was trafficked a second time that Soraton suffered his most significant trauma. He was working at a fictitious loan company that tricked customers into sending money by telling them it was the only way to access a larger loan. One victim, a middle-aged man from eastern Thailand who needed cash to pay for his mother’s medical care, called the office on video to plead for his life savings back.

When a supervisor refused, the man picked up a handgun, pointed it at his head and pulled the trigger.

“There was just silence after we heard the gunshot,” said Soraton, who was stationed next to the monitor and watched the entire incident. “The boss just walked away. These people have no feelings because they’re human traffickers.”

Soraton would spend several more weeks living in abject terror. Fellow captives were regularly assaulted and one was taken away with a black plastic bag over his head never to return.

Soraton hatched a plan to escape by hiding a cellphone in the ceiling of his sleeping quarters. That allowed him to call his mother.

“She was hysterical,” Soraton said. “It was the first time she heard from me in six months.”

She reached out to an activist outside Bangkok named Ekkaphop Luangprasert who was organizing rescue operations, and with the help of a local Cambodian contact, Soraton was able to flee. He walked for two days through rice paddies, sugar cane plantations and jungle with a guide who steered him away from landmines placed more than 40 years ago during Cambodia’s civil war.

In the early morning of Jan. 16, he emerged from the thicket with a television news crew and an ambulance waiting for him on the Thai side of the border.

Soraton is now employed in Bangkok as an electrician. He’s recently overcome months of insomnia, a condition he blames on his captivity.

Occasionally he’ll receive a spam call offering the latest scam.

“I tell them, ‘Don’t bother, I just made it back home,’” Soraton said. “Then they always ask me, ‘How did you escape?’”

Source

YouTube has also introduced ambient mode on the web. The ambient mode works using dynamic color sampling, where it introduces a subtle effect so that the video player background matches the video itself. YouTube says they wanted to inspired by the light that a screencasts out in a dark room, and wanted to recreate the same.

YouTube has also formatted the video description as well as the buttons there. The subscribe button has also been changed from red to white, although YouTube says the new shape and color is way more accessible, and will make is really stand out.

Google has been slowly transitioning towards Material You design across all of its major apps. Last year, Google rolled out Material You to its Workspace apps as well.

YouTube's Material You redesign is now rolling out to all users

Instagram fixes outage that told millions their accounts were suspended

While your mind may jump to app stores when you think about digital gatekeepers, the legislation actually targets are a wider range of services including online search engines, social networking services, certain messaging services, video sharing platform services, virtual assistants, web browsers, cloud computing services, operating systems, online marketplaces, and advertising services.

To help give a clearer understanding on what the legislation means for gatekeepers, the Commission outlined several “do’s” and “don’ts”, the do’s are as follows:

• allow third parties to inter-operate with the gatekeeper’s own services in certain specific situations
• allow their business users to access the data that they generate in their use of the gatekeeper’s platform
• provide companies advertising on their platform with the tools and information necessary for advertisers and publishers to carry out their own independent verification of their advertisements hosted by the gatekeeper
• allow their business users to promote their offer and conclude contracts with their customers outside the gatekeeper’s platform

Some example of don’ts include:

• treat services and products offered by the gatekeeper itself more favourably in ranking than similar services or products offered by third parties on the gatekeeper's platform
• prevent consumers from linking up to businesses outside their platforms
• prevent users from un-installing any pre-installed software or app if they wish so
• track end users outside of the gatekeepers' core platform service for the purpose of targeted advertising, without effective consent having been granted

For a company to be affected by the DMA legislation, it will need to meet certain criteria including having a certain annual turnover, having a certain number of users and business users, and be in an entrenched and durable position where they met the first two criteria for the last three years. From tomorrow, it will enter what is known as the implementation phase. Only on May 2 will the rules be applied.

Source: European Commission

EU's Digital Markets Act enters into force on November 1

For those unware of the issue, a lot of Nvidia RTX 4090 12-pin connectors are breaking and in worst case scenarios, they are catching fire.

Thankfully for those too afraid to proceed with a 4090 purchase, AMD had already confirmed it is not going to be using this type of power connector due to the many hazards associated with it. The confirmation came via SVP and GM of AMD Radeon, Scott Herkelman:

Hence with the launch nearly upon us and with official confirmation that no 12-pin will be present on next gen AMD cards, it was only a matter of time for the reference card designs to pop up. We have them now courtesy of Twitter user HXL. It looks like there are two card designs in question here with one of them using the typical cooler shroud with the "R" representing Radeon. The other one has an interesting logo though both of them seem to feature the same axial fan design.

Of course the more interesting thing is the apparent use of the more common dual 8-pin power connectors which is currently also present on RX 6000 series, just like Herkelman had said. This is the tried and tested method that works and it may be wise of AMD to be re-using this design. The dual 8-pins will be providing the AMD RDNA 3 reference GPU with a power headroom of 375W. This is far less than the 450W TGP (typical graphics power) of the RTX 4090.

In terms of horsepower, RDNA 3 is expected to have a big jump in typical raster performance, though Nvidia is expected to be ahead in terms of ray tracing performance.

Source and images: HXL (Twitter)

Alleged AMD RX 7000 (RDNA 3) reference card leaks and it's already better than RTX 4090

There are also reports of Instagram outage on Twitter as well as on Downdetector. Numerous people are reporting that they are losing access to their Instagram accounts and are also receiving suspension notices.

Instagram has responded on Twitter acknowledging the issue, saying that it is looking into it.

Instagram is prompting users with an account suspension notice, while also giving users an option to disagree with the decision. Instagram is asking users for their email and phone number to initiate the verification process.

More than 7,500 users reported the incident on DownDetector although it is believed that the outage has impacted a much larger number of users.

Instagram is facing multiple glitches worldwide, company "looking into it"

As you may expect after years of dormancy, the Vine code requires a “lot of work” to get back up and running – this is according to one of the sources speaking to Axios. Should the service launch around the new year, Elon’s $44 billion buyout of Twitter will essentially leave him with two social media platforms instead of one – this will undoubtedly irk certain sections of society, who have already voiced their opposition to him taking over Twitter.

Relaunching Vine is not Elon’s first big idea. He has also told employees to implement a pay-for-verification system, according to reports. Under this system, Twitter users that want a verification tick will have to pay $19.99 per month for a Twitter Blue subscription.

It will be interesting going forward to see what other ideas Elon comes up with to make Twitter better. He will also want to find new ways to monetize the service, so it makes it worth his time buying it in the first place.

Source: Axios

Twitter is now looking to bring back the video-looping service Vine

Commenting on the findings, Anuroopa Nataraj, senior research analyst with IDC’s Mobility and Consumer Device Trackers, said:

“After massive growth in 2020 and 2021, a decline in the tablet market was expected in 2022. The market is now experiencing not only a slowdown in demand, but also some strong macro-economic headwinds. Even though most tablets (Android) and Chromebooks are lower cost, we're now seeing buyer concerns even at the low end. This is largely driven by these rising economic concerns. However, tablets have found many more use cases since the pandemic, from their role in the workspace to both in-person and remote learning, entertainment, and even digital transformations across various verticals.”

The top five tablet manufacturers and their respective performance is as follows (shipments in millions):

The top five Chromebook manufacturers and their respective performance is as follows (shipments in millions):

The effects of interest rate rises from central banks tend to be lagging, so they are only just now starting to be felt for normal people. It’ll be surprising if analysts like IDC report good numbers for products like Chromebooks and tablets in the next few quarters, as people’s budgets remain squeezed.

Source: IDC

Bad economic conditions knock tablet and Chromebook shipment figures

Chief Twit Elon Musk has taken control of Twitter and shown he is well and truly up to the job title he gave himself after paying $44 billion for the micro-blogging platform.

Musk began his tenure with a few quips, before offering the following announcement detailing his plans to address content moderation on Twitter:

That tweet represents a departure from his previous statements suggesting he favors maximalist interpretations of the right to free speech, but is consistent with a tweet he emitted the day before his takeover was completed that addressed the advertising community and promised the service would not become a "free-for-all hellscape."

Over the weekend Musk clarified that the creation of the content moderation council does not mean Twitter has changed its content moderation policies.

He later japed about the issue of free speech.

A reminder: Twitter permanently suspended the account of former US president Donald Trump, thrusting the service into a roiling partisan debate about accountability, censorship, and free speech.

Musk then demonstrated the kind of behavior that makes those debates so hard to resolve by responding to a tweet from former US first lady, senator, secretary of state and Democratic presidential candidate Hilary Clinton. In the tweet, she expressed concern about how conspiracy theories have poisoned political debate, as typified by the attack on Paul Pelosi, the husband of speaker of the house Nancy Pelosi.

Musk replied with a link to a site often rated as a dubious source of news for purveying conspiracy theories, deleted it, then mocked The New York Times reporting of his actions.

While all that was going on, Twitter's head of safety Yoel Roth shared news of a spike in hateful tweeting:

Musk later surfaced what appears to be a Slack message from Roth that he alleged was kept from him by Twitter's lawyers and board during the litigation over his acquisition of the company.

Twitter's new owner also found time to complain about the company's automated onboarding process after it insisted he complete a basic management course.

Aside from his online activity, Musk has reportedly fired several senior staff, forced developers to submit code for a quality review conducted by coders from Tesla, and prepared to fire either 75 percent of Twitter staff, or 50 percent, or … maybe some other quantity. Other rumors suggest Musk will force verified Twitter users – so called "Blue Ticks" – to adopt the service's $4.99/month Twitter Blue premium service. As that service only operates in the US, Canada, Australia, and New Zealand, that does not seem to be an immediately achievable aim (and FWIW your correspondent thinks the service is not very good).

All of which brings you up to date with Musk's first weekend as chief twit, during which he certainly appears to have lived up to the title and provided entertainment and exasperation in equal measure. Meanwhile, the world waits to see what his reign will do to a site that – for better or worse – remains the de facto real-time vox-pop of global events and opinions. So totally something that will benefit from becoming a plaything for a billionaire who's also trying to run an electric car company and a space launch outfit. ®

Source

Now that Elon Musk has completed his $44 billion deal for Twitter, let’s walk through the math of owning the social media company.

It’s not just an academic exercise. Mr. Musk’s ability to make the numbers work will have a direct impact on the health of a service called the world’s town square, determining what direction he takes the business and the platform. If the financial case is grim, any plans to invest in Twitter may have to take a back seat to simply paying the bills.

The numbers are already daunting. The $44 billion acquisition was the largest leveraged buyout of a technology company in history. To do the deal, Mr. Musk, the world’s richest man, loaded about $13 billion in debt on the company, which had not turned a profit for eight of the past 10 years. The deal was inked before the global economy looked to be headed toward a recession as interest rates surged higher. And digital advertising, which makes up 90 percent of Twitter’s revenue, has been falling at social media companies.

“Technology companies should be left with enough flexibility to pay for new research and development and new levels of innovation,” said Drew Pascarella, a senior lecturer of finance at Cornell University who previously worked as a banker advising on acquisitions. This buyout “completely eliminates flexibility.”

Twitter and a representative for Mr. Musk did not respond to requests for comment.

Last year, Twitter’s interest expense was about $50 million. With the new debt taken on in the deal, that will now balloon to about $1 billion a year. Yet the company’s operations last year generated about $630 million in cash flow to meet its financial obligations.

That means that Twitter is generating less money per year than what it owes its lenders. The company also does not appear to have a lot of extra cash on hand. While it had about $6 billion in cash before Mr. Musk’s buyout, a large portion of that probably went into the cost of closing the acquisition.

That gives Mr. Musk little wiggle room, Mr. Pascarella said. “They are essentially going to take all the financial resources of the company and just pour it into servicing the debt,” he said.

To make ends meet, Mr. Musk probably has to slash costs — by a lot. Over the weekend, he was said to be already moving to do so by ordering job cuts across Twitter. One investor who put less than $1 million in the buyout of the company said he was told by the head of Mr. Musk’s family office to expect that around 50 percent of Twitter’s 7,500 employees would be laid off.

Mr. Musk could target many areas of Twitter for layoffs. The company has about $1.2 billion in annual sales and marketing expenses, a large portion of which goes to employees’ salaries, benefits and other compensation. But by cutting those costs, he risks getting rid of employees who have relationships with advertisers that would be hard to replace.

Then there is the $1.2 billion that Twitter spends annually on research and development, which also goes mainly to employee compensation. Mr. Musk could cut jobs there, too. But he has said he has grand plans for the site, like combating fake accounts and creating new ways to manage content, which require people to develop those tools. The kinds of engineers Mr. Musk has said he wants to hire are expensive.

There are other paths to slashing costs — like money spent on rent, data centers and additional expenses, which collectively cost the company more than $1 billion a year but may be harder to quickly shed. Unlike traditional targets of leveraged buyouts, Twitter does not clearly have specific businesses to shed or downsize, like a struggling division.

“This has always been the biggest challenge of this particular acquisition,” said Eric Talley, a professor of corporate law at Columbia Law School. “The last thing you want to do is sell off some integral part of what you need to run Twitter on a profitable basis. And then you’ve essentially tied your hands behind your back.”

If even cost cuts do not help, Mr. Musk may need to raise more money from outside investors within a year, Mr. Talley said.

Mr. Musk already has about $13 billion in debt from lenders, while other investors, like the venture capital firms Sequoia Capital and Andreessen Horowitz, chipped in about $7.1 billion in cash. Mr. Musk was personally responsible for the buyout’s remaining roughly $25 billion, and it remains unclear whether he gathered more investors to help lighten that load.

If Twitter needs more money in a year, finding new investors could be a lot to ask given the economics of the company. Even Mr. Musk has conceded that his initial investors in the deal valuing Twitter at $44 billion were “obviously overpaying.” The stocks of many social media companies have tumbled this year as they navigate the same problems as the rest of the economy.

Given his net worth of more than $200 billion, Mr. Musk himself could theoretically help cover Twitter’s extra cash needs. He could also try to buy out some of Twitter’s lenders and reduce its debt load.

But most of his wealth is tied up in shares of his electric vehicle company, Tesla, and its stock has plunged about 40 percent this year. At one point, Mr. Musk tried backing away from buying Twitter, and he may opt not to funnel more money into what would be at least his fifth company.

Putting more money into a leveraged, slow-growth company like Twitter is also not the same as investing in a rapidly growing venture-backed start-up like his rocket-making company, SpaceX. The risks are greater at Twitter because the banks doing the lending care only about getting paid their interest on the day it is owed. Unlike, say, a real estate company, Twitter does not have a large amount of assets to offer lenders as collateral to keep them at bay.

Still, billionaires have sought to prop up beleaguered deals before. The hedge fund manager Eddie Lampert sought to rethink the retail industry and spent billions of his own fortune keeping Sears alive after its failed merger with Kmart in the 2000s. Sears filed for bankruptcy in 2018.

And Mr. Musk has gone into businesses before that naysayers had said were doomed and proved them wrong, like manufacturing electric cars. Twitter has suffered years of mismanagement and may benefit from fixing its business out of the glare of the public markets. Mr. Musk could bring new product ideas and hire engineering experts who might not have wanted to work for Twitter before.

Mr. Musk is “a phenomenal capital allocator, and I think he’ll make a lot of money in Twitter,” said Chamath Palihapitiya, a venture capitalist who was an early Facebook executive. “It doesn’t fit my risk profile. But I think he’s going to be very successful.”

Others caution against the ebullience that initially drove investors to Mr. Musk’s deal, warning that the lure of tech visionaries can fade with market fortunes, especially as global economic fears have mushroomed in recent months.

“At the height of a market boom, those appeals work more easily than they do in times like we are presently entering,” said Robert Bruner, a professor at the Darden School of Business at the University of Virginia and author of the book “Deals From Hell.”

Mr. Bruner said the worst deals are typically struck at the peak of a market — as with Mr. Musk’s purchase of Twitter. He offered what he thought could be a worst-case scenario for the company. In that future, Mr. Musk would not be able to “get the expenses down to the level necessary to cover the debt burden.” That would “slowly erode the company’s equity, and he’s unable to find more equity investors.”

The final outcome? “Slowly, Twitter implodes,” Mr. Bruner said.

Maureen Farrell contributed reporting.

Source

[Note:  Registration or email address is required to view articles at NY Times.]

Elon Musk wastes no time changing Twitter

Apple’s 14- and 16-inch MacBook Pros might not arrive until next March

Windows experiments

We'll kick off this edition of Microsoft Weekly by talking about some new, hidden, and experimental features in development for Windows.

For starters, Microsoft is working on an expedited Windows updates mechanism for commercial customers. As the name implies, it enables IT admins to configure policies which will ensure that security updates are installed on an urgent basis. Work is also underway to give the same treatment to non-security updates.

In the same vein, Microsoft has kicked off a public preview of Unified Update Platform (UUP) on premises, which should enable faster Windows updates. UUP on premises basically acts as the connective tissue between Windows Server Update Services (WSUS) and Microsoft Configuration Manager to ensure a quick and seamless update process. Following general availability next year, all Windows updates will be delivered through UUP.

Talking about commercial customers and developers, Microsoft has also launched Project Volterra - also known as Windows Dev Kit 2023 -, which is the first ARM-powered mini PC from the company. It aims to encourage Windows developers to bring their apps to ARM devices and help populate the category among consumers. Powered by a Snapdragon processor, 32GB of RAM, and 512GB of SSD, it costs $599.

In terms of platform support, we have learned that Microsoft plans to upgrade Windows Subsystem for Android (WSA) to Android 13 and add file transfer, shortcuts, picture-in-picture, and local network access support. However, there are no concrete timelines for this update publicly available. In the same vein, we also found out how you can force-enable more designs of Widgets in Windows 11, if that's what floats your boat. In addition, you might also want to know that Microsoft has released a new update for the Sysinternals toolset, which now offers a dark mode for Process Explorer.

Microsoft is seemingly also readying another Moment update. The company gave us a brief early look during its latest Windows Insider Program webcast, but the feature-set is still unconfirmed.

Wrapping up this section with some Windows updates related news, Windows 11 version 22H2 received an optional cumulative update this week in the form of KB5018496 (build 22621.755). It brought a lot of new features including the right-click Task Manager option and an enhanced backup experience using Microsoft Account (MSA). Mostly the same improvements also made their way to the Beta Channel with build 22623.875.

The optional KB5018496 cumulative update also packs some hidden features such as the tablet-optimized taskbar, redesigned notification area, and improved File Explorer search. You can find out how to enable them here.

Finally, we were also treated to build 25231 in the Dev Channel. It's mostly just bug fixes but there is a welcome new feature for Phone Link which allows Samsung phone users to immediately access a hotspot.

Bugs infestation

As has been the case for the past few weeks, quite a few Windows bugs were recently discovered. Microsoft has confirmed that the latest Patch Tuesday update has broken OneDrive on Windows 10 version 22H2. The app might crash unexpectedly and you might be greeted with some error messages when you perform certain activities. Microsoft is working on a resolution.

And while the merits of this being a bug are debatable, it seems like Windows 11 won't allow you to "safely eject" a USB device if Task Manager is open. It also seems like readings for Nvidia hardware are a bit wonky in Task Manager. We've heard reports of trend visualizations being inverted with one instance of 3D usage for an Nvidia GPU being shown at 96% even though the card itself is clearly idling at a pleasantly cool temperature of 39°C. The readings appear to be inverted, which means the actual usage in this capture instance may be 4%.

Another issue you might face while being on the Windows Insider Program is that the dedicated Settings app won't work as you expect, but the good news is that you can read our step-by-step guide to fix the issue here.

But we'd like to end this section on a pleasant note so let's shift our attention to some bug squashing too. AMD has released a new driver version that fixes the infamous black screen issue on Radeon GPUs on Windows 11 version 22H2. Meanwhile, KB5018482 for Windows 10 is now out with fixes for gaming, updates, time zones, and Microsoft Edge.

OneNote and other Office updates

We have a lot of updates to share regarding Office and Microsoft 365 apps in this edition of Microsoft Weekly again. We'll start off with OneNote, which is now available in the Microsoft Store. If you're thinking "hey, it's always been available in the Microsoft Store", you're absolutely correct but that's the old "OneNote for Windows 10" that won't be getting any new features and is in support mode until October 2025. The Microsoft Store hosts both versions of OneNote but will only surface the "recommended" OneNote for Windows version, which will continue receiving new functionalities. As if the OneNote branding wasn't confusing enough on its own.

In addition, Microsoft rolled out some OneNote enhancements that will allow you to quickly take notes with fewer distractions, check out the details here.

Meanwhile, Word for the web and PowerPoint for the web now allow you to pick photos from your phone. There are a few caveats though, you need Phone Link, a Microsoft 365 subscription, and version 104 of Firefox (if you're a Firefox user, that is).

Speaking of browsers, Microsoft Edge 107 is now out. The highlight of this release is an improved Sidebar and a new policy for flexibility on the browser's startup.

Finally, Microsoft also recapped all the capabilities that it added to Excel in October 2022. The list isn't very lengthy but includes some very welcome additions such as a shortcut menu, new DAX functions, and expanded availability of Office Scripts automation.

Git gud

In its latest earnings call, Microsoft revealed that Xbox Cloud Gaming now has over 20 million customers, which is double the number that the company touted six months ago. That said, Xbox chief Phil Spencer has noted that despite being profitable, Game Pass - which offers cloud gaming through its Ultimate tier - might see a price hike in the future.

Coming over to actual games, Microsoft-owned Bethesda has announced a next-gen update for Fallout 4 on Xbox Series X|S, PlayStation 5, and PC. Meanwhile, real-time strategy fans will be plenty pleased with some recent gaming news too. The Anniversary Update for Age of Empires IV - now also slated for an Xbox release in 2023 - is now live with new civilizations, cheats, and more. In the same vein, Age of Mythology Retold has been announced as a remaster of the RTS classic, but without a launch window.

Capping this section off with our deals roundup, we have a whopping four titles in the Xbox Free Play Days promotion, along with Praetorians HD Remaster and Dead End Job as the selection for November's Games with Gold. And as always, for all you PC purists out there, don't miss out on this Weekend's PC Game Deals, personally curated by our News Editor Pulasthi Ariyasinghe.

Dev Channel

• A new typosquatting campaign is delivering Windows and Android malware

• Google Chrome is dropping support for Windows 7 and 8.1 in a few months

• Android 12L for Surface Duo and Duo 2 is now available

• Microsoft has quietly updated its list of supported AMD and Intel CPUs for Windows 11 version 22H2

   

Under the spotlight

A few days ago, I wrote an editorial highlighting Microsoft's secrecy around the changes made in the Windows 10 2022 Update (version 22H2). After much digging around, I also managed to find a changelog so if you're curious about what the update actually packs, do have a look at my piece here.

Meanwhile, News Reporter Taras Buria got three guides out of the door. The first is about how to enable experimental Windows Widgets designs through a third-party tool.

The second guide details the process to force-disable File Explorer tabs in Windows 11. It's quite strange that Microsoft doesn't offer a dedicated configuration section to control its behavior, so you'll have to rely on a third-party tool again.

Finally, the third guide explains how you can enable the tablet-optimized Taskbar in Windows 11 desktop PCs if you are not a fan of the current thicc variant.

Neowin co-founder Steven Parker wasn't too far behind with his guides either. His first guide gives our readers a walkthrough of how to revert to the Windows 10 File Explorer if you are not a fan of the Windows 11 variant.

Steven's second guide is a very handy piece of writing that describes what to do if the Windows 11 Insider Settings page is not working for you.

Steven also took a trip down memory lane by discovering a "Windows 12" design concept that merges Windows 11's Fluent design with Windows Vista. Of course, not every design is everyone's cup of tea, but to find out if this tickles your fancy, do check out the screenshots.

Finally, in his latest edition of Warwagon's Tech Tip Tuesday, forum member Adam Bottjen detailed a useful way to transfer photos from your smartphone to a PC, read it here.

Logging off

Our most interesting news item this week is about former Windows chief Steven Sinofsky talking about the origins and development process of Windows 8, over a decade after the OS' release. Some might find Sinofsky's point of views a bit controversial as the former executive argues that Windows 8's drastic UX changes came into being because he believes traditional Windows had run its course. He still thinks that the desktop computing market is on decline since most people do their computing via touch-based interfaces and a grid of apps. Finally, he believes that Windows 8 failed because the market just couldn't handle the innovation that it brought to desktop computing. It's a really interesting interview and you can find out more details here.

Microsoft Weekly: Windows experiments, bugs, and OneNote confusions

“There is nothing that can replace it,” says LaBelle. “No one has what Twitter has.” For her, Twitter served as a place for trans people to meet, to date, and to post with anonymity if they are fearful of harassment or still exploring their identity and don’t want to share photos or videos. With Musk’s takeover, LaBelle isn’t leaving—she needs the platform for her work—but she is rebuilding a smaller version of her community on a forum that feels safer.

“Discord isn’t really going to be Twitter,” says LaBelle. “It’s just going to be me grabbing all of the people that I have in my circle right now and running in the other direction as fast as possible because we’re being chased by fascists.”

LaBelle is one of many who fear Twitter could be plunged into chaos under Musk’s rule. There’s already harassment on the platform, but one of Musk’s stated rationales for buying Twitter was to roll back moderation rules. For people like LaBelle who come from marginalized groups, that feels like an invitation for trolls to spread more hatred and harassment. 

Musk has said he would reinstate former US president Donald Trump and allow any content that doesn’t break the law, although he tweeted today that no major decisions would be made until he had convened a council of people with “widely diverse viewpoints” on moderation. And some Twitter users are planning to flee in protest because they don’t want to feed free content to a platform owned by the world’s richest person.

Rather than rushing to the exits and deleting their accounts, many with reservations about Musk’s takeover are making plans similar to LaBelle’s. They’re giving out information for followers to find them on Discord or Mastodon, a decentralized microblogging platform that most closely resembles Twitter.

Mastodon is already benefiting from speculation about Twitter’s new owner. Some 18,000 people signed up for Mastodon accounts between October 20 and 27, says Eugen Rochko, the platform’s CEO. As of October 28, it had 381,113 active users. Mastodon’s Twitter handle is also being used a lot across Twitter by people announcing new Mastodon accounts, Rochko says.

Many people tweeting Friday morning under #TwitterMigration said they weren’t ready to abandon Twitter entirely but had set up Mastodon accounts in anticipation of sweeping changes. Some in academic or tech communities on Twitter put their new Mastodon profiles in their bios or Twitter names. “Looks like #Mastodon is trending on Twitter as more and more people are announcing their new profiles,” the company wrote Thursday.

But a fledgling network can’t match one as ubiquitous as Twitter. Danny Groner, who works as a marketing director at venture group Forecast Labs, says he uses Twitter professionally to connect with people. He isn’t quitting yet but says he would if the platform changed and inappropriate content took over. Still, after seeing plenty of chatter about Mastodon, he created his own account on Thursday. Groner says he doesn’t know much about the platform yet, but he wants to have a contingency plan if trolls and harassers become more prominent on Twitter.

“It would be a real shame for me professionally to lose this resource,” he says. “There is a breaking point for me personally and morally. If I had to decide it’s time to close up my account and move on, I would have to be creative about how to replicate that success.”

Musk isn’t the only reason people are exploring other platforms. The app has steadily lost its heaviest users since the Covid-19 pandemic began, Reuters reported. That decline is likely a result of the often abrasive discourse on Twitter but also a response to the platform’s struggle to innovate. The different features and atmosphere offered by Discord and Mastodon, which operate across various servers or instances to organize users, could set them apart from Twitter.

For others, quitting is a principled move. Matt Haughey, who founded the community blogging platform MetaFilter, had plans to say goodbye to Twitter as soon as the deal closed. He says he doesn’t trust Musk’s judgment and doesn’t want to be part of a platform he owns. “I do not want to create value for him,” he says.

Instead, Haughey says he is thinking about going back to blogging. “It’s 15 years, 30,000 followers, but I don’t make any money from it,” Haughey says of his time on Twitter. “I don’t get paid to do it. It is a waste of my time if I’m not having fun with it.” By Friday, Haughey had deleted all his tweets and updated his bio: “I won't tweet on any musk-owned network,” it read.

Twitter Users Flock to Other Platforms as the Elon Era Begins

(May require free registration to view)

Will Elon Musk keep funding Twitter’s most interesting side project?

The power adapter for the RTX 4090 feeds four 8-pin power connectors into a single 12VHPWR connector.

Sam Machkovech

 

Earlier this week, a couple of Reddit users reported that the power connectors for their expensive new Nvidia GeForce RTX 4090 GPUs had partially melted and that Nvidia was looking into the issue. Since then, at least nine additional Reddit users have posted about the same problem with the 16-pin power connector (a thread collecting all information on the problem is here).

 

Igor Wallossek of the German-language hardware site Igor's Lab has also performed additional testing, and said that the power adapter cable (rather than the GPUs or the 12VHPWR connector) may be to blame for the problems. The adapter, which is apparently manufactured by a company called Astron and was provided by Nvidia to all of its board partners, uses "a total of four thick 14AWG wires distributed over a total of six contacts," with a thin solder base that Wallossek says can be damaged easily when the cables are moved or bent.

 

"If, in the worst case, the two outer wires break off, the entire current in the middle flows through the remaining two wires," Wallossek wrote. "The fact that this then becomes really hot does not have to be explained separately."

 

Contrast Nvidia's adapter cable to the 12VHPWR cable provided with a Be Quiet! Dark Power Pro 13 power supply, one of a few newer ATX 3.0 PSU models with a 12VHPWR cable that connects right to the PSU rather than relying on an adapter. The cable is thicker than the one in Nvidia's adapter because every pin has its own wire, but it also runs cool and delivers reliable power to the GPU.

 

Wallossek also said that Nvidia told its partners to send all affected cards back to Nvidia for further investigation and testing. We asked Nvidia about this problem and the investigation earlier this week and will update if we receive new information.

 

Testing suggests faulty cable may be to blame for melting RTX 4090 connectors

AMD released its Ryzen 7000 series with much fanfare. It was expected to compete with Intel’s then upcoming Raptor Lake series of processors.

While the AMD Ryzen 7000 series processors are themselves quite powerful, they have their own set of problems.

Problems With New AMD Ryzen 7000 Processors

The biggest problem currently is that AMD Ryzen 7000 processor series is based on a new Zen 4 platform. This new platform requires a new motherboard CPU socket called AM5. It also requires DDR5 RAM, unlike Intel Raptor Lake, which supports both DDR4 and DDR5 RAM. In addition, it also provides support for PCIe 5.0. All this might seem good, however, it increases the price of making an AMD computer tremendously.

Compare that to Intel Raptor Lake. In it, you don’t need a new motherboard or latest DDR5 RAM. One can easily replace the Alder Lake CPU they have with the latest Raptor Lake one and enjoy all the new speed boost with a simple upgrade. What’s funny is that this unique situation is usually the other way around for Intel and AMD. Either way, this means that going AMD is not a valuable option.

The second problem AMD Ryzen 7000 processors are having is the chiplet design. In simple words, making big sized CPUs with a single chip is not easy or cheap. So what AMD does is that in expensive processors, it makes multiple, smaller sized chiplets and connects them together to make a single powerful CPU. Chiplet design is not necessarily bad, just that sometimes it creates problems.

In recent Ryzen 7000 benchmarks, many found out that the lesser powerful Ryzen 7 7700X was performing better than more powerful Ryzen 9 7900X and 7950X processors in gaming. What’s the difference between the both. Ryzen 7 7700X contains a single chiplet whereas Ryzen 7900X and 7950X has two chiplets.

There are two possible reasons here. Some believe that the second chiplet in both 7900X and 7950X are slower than the first one, some say the latest operating systems are not optimized enough and are moving processing power to the second chiplet instead of the first one, which is said to be faster one. Another possibility is that there’s some latency in communication between both the chiplets.

Overall, the AMD Ryzen 9 7900X and 7950X processors are not as attractive as they should be. This where 7800X comes in.

Ryzen 7800X Benchmark Leaks

AMD so far has not released or detailed anything about the Ryzen 7 7800X processor. This is because only Ryzen 7600X, 7700X, 7900X and 7950X were released by AMD. Missing from the list are the usual processors like Ryzen 7500X and 7800X. So everyone was wondering when they will be announced, leading to many speculations.

A couple of days ago, a Twitter handle @BenchLeaks which tracks Geekbench results revealed some interesting information. It revealed that someone has tested two processors on Geekbench, namely, AMD Ryzen 3 7300X with 4 Cores and 8 Threads and AMD Ryzen 7 7800X with 10 Cores and 20 Threads.

It was shared all over the internet. Many wondered how can AMD release a X300X processor model, which usually comes at least a year after the main ones. Even more surprising is how can a CPU have 10 cores, like the leaked 7 7800X suggests. Also, it shows 7800X with a dual chiplet design, which is not to be found in AMD Ryzen processors below the X900X series. So did AMD use the chiplets in Ryzen 9 7900X processors and disable a core in each of the chiplets for 7800X. Many speculated about it.

Still, everyone was happy that AMD is finally providing a cheaper alternative to the Ryzen 9 7900X and 7950X processors, which are getting too expensive due to the new platform. Except, that’s not true. The benchmarks are fake.

Ryzen 7800X Benchmark Is Fake

Chips and Cheese have revealed that they had faked the whole thing. They even published an article about it, titled Why you can’t trust CPUID. They revealed the reason behind their act.

In the article, they reveal how easy it is to fool the computer with CPU model numbers. Usually, CPU model names are provided to the OS at boot. Their CPUID is accessed with a software like Geekbench asks for it.

What they did was that they took an AMD Ryzen 9 7900X or 7950X processor and then modified it’s CPUID with a software named MsrUtil, originally made by them to find out about CPU bottlenecks. In it, they also modified the processor by decreasing the PBO (Precision Boost Overdrive) and cut the processor cores that were being reported.

The outcome was that many benchmarking software were fooled into believing that it was a true Ryzen 7 7800X processor. This includes Geekbench, Cinebench, HWMonitor, Blender Benchmark, and many more. The only software which were not fooled were HWiNFO and BenchMate, which in-turn uses the same HWiNFO. This because HWiNFO seems to use a different technique to get the processor model information.

Chips and Cheese reveal that these instructions about the AMD model numbers are writable even in models produced a decade ago. Possibly up with same type of manipulation.

As both the Ryzen 7300X and 7800X benchmarks were posted by the same user TitanicFreak on Geekbench, it’s likely that AMD Ryzen 3 7300X results were completely fake too.

Outcome Of Fake Benchmarks

It goes without saying that one should not trust anything they see on the internet. It can be fake too. Especially when it comes to rumors, that is.

In fact, the @BenchLeaks Twitter handle seemed embarrassed by it, too. They joked about how they have activated an “AntiAmdHax0rPatchV69b1337”.

BenchLeaks accepting their innocence is welcomed. However, there were many sites which originally reported the news and did not acknowledge that this is fake. Some kept the articles online without any clarification, while some outright deleted their article to hide that they jumped onto the conclusion too early. This is not about small news sites like ours, but major tech news sites out there. Whatever it maybe, it’s a good lesson for all of us.

As far as Ryzen 7300X and 7800X is concerned, they are much required. AMD does not seem to care enough about budget builds and hence release the cheaper processors in the lineup with previous outdated architecture. Instead, we would like to see a Ryzen 3 7300X processor based on the latest Zen 4 architecture when the platform gets more affordable. For Ryzen 7 7800X, we feel it would be 3D-V Cache based one, just like how AMD Ryzen 5800X3D is. So it’s likely that Ryzen 7 7800X will be named 7800X3D instead. But so far it’s all speculation. So it’s best not be fooled by all this and best wait for the official confirmation about it.

The 10-Core AMD Ryzen 7800X CPU Leak Is Fake