An anonymous reader quotes a report from Reuters: Independent browser companies in the European Union are seeing a spike in users in the first month after EU legislation forced Alphabet's Google, Microsoft and Apple to make it easier for users to switch to rivals, according to data provided to Reuters by six companies. The early results come after the EU's sweeping Digital Markets Act, which aims to remove unfair competition, took effect on March 7, forcing big tech companies to offer mobile users the ability to select from a list of available web browsers from a "choice screen." [...]

Cyprus-based Aloha Browser said users in the EU jumped 250% in March -- one of the first companies to give monthly growth numbers since the new regulations came in. Founded in 2016, Aloha, which markets itself as a privacy focused alternative to browsers owned by big tech, has 10 million monthly average users and earns money through paid subscriptions, rather than selling ads by tracking users. "Before, EU was our number four market, right now it's number two," Aloha CEO Andrew Frost Moroz said in an interview. Norway's Vivaldi, Germany's Ecosia and U.S.-based Brave have also seen user numbers rise following the new regulation. U.S.-based DuckDuckGo, which has about 100 million users, and its bigger rival, Norway-based Opera (OPRA.O), opens new tab are also seeing growth in users, but said the choice screen rollout is still not complete. "We are experiencing record user numbers in the EU right now," said Jan Standal, vice president at Opera, which counts over 324 million global users.

Under the new EU rules, mobile software makers are required to show a choice screen where users can select a browser, search engine and virtual assistant as they set up their phones. Previously, tech companies such as Apple and Google loaded phones with default settings that included their preferred services, such as the voice assistant Siri for iPhones. Changing these settings required a more complicated process. Apple is now showing up to 11 browsers in addition to Safari in the choice screens curated for each of the 27 countries in the EU, and will update those screens once every year for each country. While DuckDuckGo and Opera are offered in Apple's list, opens new tab in all 27 countries, Aloha is in 26 countries, Ecosia is in 13 and Vivaldi in 8. Google is currently showing browser choices on devices made by the company and said new devices made by other companies running Android operating system will also display choice screen in the coming months. A Google spokesperson said they do not have data on choice screens to share yet.

Meta Platforms is planning to launch two small versions of its forthcoming Llama 3 large-language model next week, The Information has reported [non-paywalled link]. From the report: The models will serve as a precursor to the launch of the biggest version of Llama 3, expected this summer. Release of the two small models will likely help spark excitement for the forthcoming Llama 3, which will be coming out roughly a year after Llama 2 launched last July.

It comes as several companies, including Google, Elon Musk's xAI and Mistral, have released open-source LLMs. Meta hopes Llama 3 will catch up with OpenAI's GPT-4, which can answer questions based on images users upload to the chatbot. The biggest version will be multimodal, which means it will be capable of understanding and generating both texts and images. In contrast, the two small models to be released next week won't be multimodal, the employee said.

Gemini 1.5 Pro, Google's most capable generative AI model, is now available in public preview on Vertex AI, Google's enterprise-focused AI development platform. From a report: The company announced the news during its annual Cloud Next conference, which is taking place in Las Vegas this week. Gemini 1.5 Pro launched in February, joining Google's Gemini family of generative AI models. Undoubtedly its headlining feature is the amount of context that it can process: between 128,000 tokens to up to 1 million tokens, where "tokens" refers to subdivided bits of raw data (like the syllables "fan," "tas" and "tic" in the word "fantastic").

One million tokens is equivalent to around 700,000 words or around 30,000 lines of code. It's about four times the amount of data that Anthropic's flagship model, Claude 3, can take as input and about eight times as high as OpenAI's GPT-4 Turbo max context. A model's context, or context window, refers to the initial set of data (e.g. text) the model considers before generating output (e.g. additional text). A simple question -- "Who won the 2020 U.S. presidential election?" -- can serve as context, as can a movie script, email, essay or e-book.

Google Cloud on Tuesday joined AWS and Azure in announcing its first custom-built Arm processor, dubbed Axion. From a report: Based on Arm's Neoverse 2 designs, Google says its Axion instances offer 30% better performance than other Arm-based instances from competitors like AWS and Microsoft and up to 50% better performance and 60% better energy efficiency than comparable X86-based instances. [...] "Technical documentation, including benchmarking and architecture details, will be available later this year," Google spokesperson Amanda Lam said. Maybe the chips aren't even ready yet? After all, it took Google a while to announce Arm-chips in the cloud, especially considering that Google has long built its in-house TPU AI chips and, more recently, custom Arm-based mobile chips for its Pixel phones. AWS launched its Graviton chips back in 2018.

For decades, work has revolved around documents, spreadsheets, and slide decks. Word, Excel, PowerPoint; Pages, Numbers, Keynote; Docs, Sheets, Slides. Now Google is proposing to add another to that triumvirate: an app called Vids that aims to help companies and consumers make collaborative, shareable video more easily than ever. From a report: Google Vids is very much not an app for making beautiful movies... or even not-that-beautiful movies. It's meant more for the sorts of things people do at work: make a pitch, update the team, explain a complicated concept. The main goal is to make everything as easy as possible, says Kristina Behr, Google's VP of product management for the Workspace collaboration apps. "The ethos that we have is, if you can make a slide, you can make a video in Vids," she says. "No video production is required."

Based on what I've seen of Vids so far, it appears to be roughly what you'd get if you transformed Google Slides into a video app. You collect assets from Drive and elsewhere and assemble them in order -- but unlike the column of slides in the Slides sidebar, you're putting together a left-to-right timeline for a video. Then, you can add voiceover or film yourself and edit it all into a finished video. A lot of those finished videos, I suspect, will look like recorded PowerPoint presentations or Meet calls or those now-ubiquitous training videos where a person talks to you from a small circle in the bottom corner while graphics play on the screen. There will be lots of clip art-heavy product promos, I'm sure. But in theory, you can make almost anything in Vids. ou can either do all this by yourself or prompt Google's Gemini AI to make a first draft of the video for you. Gemini can build a storyboard; it can write a script; it can read your script aloud with text-to-speech; it can create images for you to use in the video. The app has a library of stock video and audio that users can add to their own Vids, too.

The capability of new AI models will surpass human intelligence by the end of next year [non-paywalled link], so long as the supply of electricity and hardware can satisfy the demands of the increasingly powerful technology, according to Elon Musk. From a report: "My guess is that we'll have AI that is smarter than any one human probably around the end of next year," said the billionaire entrepreneur, who runs Tesla, X and SpaceX. Within the next five years, the capabilities of AI will probably exceed that of all humans, Musk predicted on Monday during an interview on X with Nicolai Tangen, the chief executive of Norges Bank Investment Management.

Musk has been consistently bullish on the development of so-called artificial general intelligence, AI tools so powerful they can beat the most capable individuals in any domain. But Monday's prediction is ahead of schedules he and others have previously forecast. Last year, he predicted "full" AGI would be achieved by 2029. Some of Musk's boldest predictions, such as rolling out self-driving Teslas and landing a rocket on Mars, have not yet been fulfilled. A number of AI breakthroughs over the past 18 months, including the launch of video generation tools and more capable chatbots, have pushed the frontier of AI forward faster than expected. Demis Hassabis, the co-founder of Google's DeepMind, predicted earlier this year that AGI could be achieved by 2030.

The pace of development has been slowed by a bottleneck in the supply of microchips, particularly those produced by Nvidia, which are essential for training and running AI models. Those constraints were easing, Musk said, but new models are now testing other data centre equipment and the electricity grid. "Last year it was chip constrained ... people could not get enough Nvidia chips. This year it's transitioning to a voltage transformer supply. In a year or two [the constraint is] just electricity supply," he said.

An anonymous reader shares a report: Microsoft is getting ready to fully unveil its vision for "AI PCs" next month at an event in Seattle. Sources familiar with Microsoft's plans tell The Verge that Microsoft is confident that a round of new Arm-powered Windows laptops will beat Apple's M3-powered MacBook Air both in CPU performance and AI-accelerated tasks. After years of failed promises from Qualcomm, Microsoft believes the upcoming Snapdragon X Elite processors will finally offer the performance it has been looking for to push Windows on Arm much more aggressively. Microsoft is now betting big on Qualcomm's upcoming Snapdragon X Elite processors, which will ship in a variety of Windows laptops this year and Microsoft's latest consumer-focused Surface hardware.

Microsoft is so confident in these new Qualcomm chips that it's planning a number of demos that will show how these processors will be faster than an M3 MacBook Air for CPU tasks, AI acceleration, and even app emulation. Microsoft claims, in internal documents seen by The Verge, that these new Windows AI PCs will have "faster app emulation than Rosetta 2" -- the application compatibility layer that Apple uses on its Apple Silicon Macs to translate apps compiled for 64-bit Intel processors to Apple's own processors.

App emulation has been a big problem for Windows on Arm over the past decade, but Microsoft did deliver x64 app emulation for Windows 11 more than two years ago. This helps ensure apps can run on Windows on Arm devices when there isn't a native ARM64 version. Native Arm apps are key for improved performance on upcoming Windows on Arm laptops, and Google has just recently released its own ARM64 version of Chrome ready for these upcoming devices.

It's the world's most widely used vulnerability database, reports SC Magazine, offering standards-based data on CVSS severity scores, impacted software and platforms, contributing weaknesses, and links to patches and additional resources.

But "there is a growing backlog of vulnerabilities" submitted to America's National Vulnerability Database and "requiring analysis", according to a new announcement from the U.S. Commerce Department's National Institute of Standards. "This is based on a variety of factors, including an increase in software and, therefore, vulnerabilities, as well as a change in interagency support." From SC Magazine: According to NIST's website, the institute analyzed only 199 of 3370 CVEs it received last month. [And this month another 677 came in — of which 24 have been analyzed.]

Other than a short notice advising it was working to establish a new consortium to improve the NVD, NIST had not provided a public explanation for the problems prior to a statement published [April 2]... "Currently, we are prioritizing analysis of the most significant vulnerabilities. In addition, we are working with our agency partners to bring on more support for analyzing vulnerabilities and have reassigned additional NIST staff to this task as well."

NIST, which had its budget cut by almost 12% this year by lawmakers, said it was committed to continuing to support and manage the NVD, which it described as "a key piece of the nation's cybersecurity infrastructure... We are also looking into longer-term solutions to this challenge, including the establishment of a consortium of industry, government and other stakeholder organizations that can collaborate on research to improve the NVD," the statement said. "We will provide more information as these plans develop..."

A group of cybersecurity professionals have signed an open letter to Congress and Commerce Secretary Gina Raimondo in which they say the enrichment issue is the result of a recent 20% cut in NVD funding.
The article also cites remarks from NVD program manager Tanya Brewer (reported by Infosecurity Magazine) from last week's VulnCon conference on plans to establish a NVD consortium. "We're not going to shut down the NVD; we're in the process of fixing the current problem. And then, we're going to make the NVD robust again and we'll make it grow."

Thanks to Slashdot reader spatwei for sharing the article.

On Mozilla's blog, engineer Martin Thomson explores Google's "Privacy Sandbox" initiative (which proposes sharing a subset of private user information — but without third-party cookies).

The blog post concludes that Google's Protected Audience "protects advertisers (and Google) more than it protects you." But it's not all bad — in theory: The idea behind Protected Audience is that it creates something like an alternative information dimension inside of your (Chrome) browser... Any website can push information into that dimension. While we normally avoid mixing data from multiple sites, those rules are changed to allow that. Sites can then process that data in order to select advertisements. However, no one can see into this dimension, except you. Sites can only open a window for you to peek into that dimension, but only to see the ads they chose...

Protected Audience might be flawed, but it demonstrates real potential. If this is possible, that might give people more of a say in how their data is used. Rather than just have someone spy on your every action then use that information as they like, you might be able to specify what they can and cannot do. The technology could guarantee that your choice is respected. Maybe advertising is not the first thing you would do with this newfound power, but maybe if the advertising industry is willing to fund investments in new technology that others could eventually use, that could be a good thing.
But here's some of the blog post's key criticisms:

• "[E]ntities like Google who operate large sites, might rely less on information from other sites. Losing the information that comes from tracking people might affect them far less when they can use information they gather from their many services... [W]e have a company that dominates both the advertising and browser markets, proposing a change that comes with clear privacy benefits, but it will also further entrench its own dominance in the massively profitable online advertising market..."

• "[T]he proposal fails to meet its own privacy goals. The technical privacy measures in Protected Audience fail to prevent sites from abusing the API to learn about what you did on other sites.... Google loosened privacy protections in a number of places to make it easier to use. Of course, by weakening protections, the current proposal provides no privacy. In other words, to help make Protected Audience easier to use, they made the design even leakier..."

• "A lot of these leaks are temporary. Google has a plan and even a timeline for closing most of the holes that were added to make Protected Audience easier to use for advertisers. The problem is that there is no credible fix for some of the information leaks embedded in Protected Audience's architecture... In failing to achieve its own privacy goals, Protected Audience is not now — and maybe not ever — a good addition to the Web."

While the European Parliament passed a wide-ranging "AI Act" in March, "Leaders from Microsoft, Google, and OpenAI have all called for AI regulations in the U.S.," writes CIO magazine. Even the Chamber of Commerce, "often opposed to business regulation, has called on Congress to protect human rights and national security as AI use expands," according to the article, while the White House has released a blueprint for an AI bill of rights.

But even though the U.S. Congress hasn't passed AI legislation — 16 different U.S. states have, "and state legislatures have already introduced more than 400 AI bills across the U.S. this year, six times the number introduced in 2023." Many of the bills are targeted both at the developers of AI technologies and the organizations putting AI tools to use, says Goli Mahdavi, a lawyer with global law firm BCLP, which has established an AI working group. And with populous states such as California, New York, Texas, and Florida either passing or considering AI legislation, companies doing business across the US won't be able to avoid the regulations. Enterprises developing and using AI should be ready to answer questions about how their AI tools work, even when deploying automated tools as simple as spam filtering, Mahdavi says. "Those questions will come from consumers, and they will come from regulators," she adds. "There's obviously going to be heightened scrutiny here across the board."
There's sector-specific bills, and bills that demand transparency (of both development and output), according to the article. "The third category of AI bills covers broad AI bills, often focused on transparency, preventing bias, requiring impact assessment, providing for consumer opt-outs, and other issues."

One example the article notes is Senate Bill 1047, introduced in the California State Legislature in February, "would require safety testing of AI products before they're released, and would require AI developers to prevent others from creating derivative models of their products that are used to cause critical harms."

Adrienne Fischer, a lawyer with Basecamp Legal, a Denver law firm monitoring state AI bills, tells CIO that many of the bills promote best practices in privacy and data security, but said the fragmented regulatory environment "underscores the call for national standards or laws to provide a coherent framework for AI usage."

Thanks to Slashdot reader snydeq for sharing the article.

In Communication of the ACM, Google's VP of Education notes how calculators impacted math education — and wonders whether generative AI will have the same impact on CS education: Teachers had to find the right amount of long-hand arithmetic and mathematical problem solving for students to do, in order for them to have the "number sense" to be successful later in algebra and calculus. Too much focus on calculators diminished number sense. We have a similar situation in determining the 'code sense' required for students to be successful in this new realm of automated software engineering. It will take a few iterations to understand exactly what kind of praxis students need in this new era of LLMs to develop sufficient code sense, but now is the time to experiment."
Long-time Slashdot reader theodp notes it's not the first time the Google executive has had to consider "iterating" curriculum: The CACM article echoes earlier comments Google's Education VP made in a featured talk called The Future of Computational Thinking at last year's Blockly Summit. (Blockly is the Google technology that powers drag-and-drop coding IDE's used for K-12 CS education, including Scratch and Code.org). Envisioning a world where AI generates code and humans proofread it, Johnson explained: "One can imagine a future where these generative coding systems become so reliable, so capable, and so secure that the amount of time doing low-level coding really decreases for both students and for professionals. So, we see a shift with students to focus more on reading and understanding and assessing generated code and less about actually writing it. [...] I don't anticipate that the need for understanding code is going to go away entirely right away [...] I think there will still be at least in the near term a need to understand read and understand code so that you can assess the reliabilities, the correctness of generated code. So, I think in the near term there's still going to be a need for that." In the following Q&A, Johnson is caught by surprise when asked whether there will even be a need for Blockly at all in the AI-driven world as described — and the Google VP concedes there may not be.

Matthew Connatser reports via The Register: A study has concluded that Apple's privacy practices aren't particularly effective, because default apps on the iPhone and Mac have limited privacy settings and confusing configuration options. The research was conducted by Amel Bourdoucen and Janne Lindqvist of Aalto University in Finland. The pair noted that while many studies had examined privacy issues with third-party apps for Apple devices, very little literature investigates the issue in first-party apps -- like Safari and Siri. The aims of the study [PDF] were to investigate how much data Apple's own apps collect and where it's sent, and to see if users could figure out how to navigate the landscape of Apple's privacy settings.

The lengths to which Apple goes to secure its ecosystem -- as described in its Platform Security Guide [PDF] -- has earned it kudos from the information security world. Cupertino uses its hard-earned reputation as a selling point and as a bludgeon against Google. Bourdoucen and Janne Lindqvist don't dispute Apple's technical prowess, but argue that it is undermined by confusing user interfaces. "Our work shows that users may disable default apps, only to discover later that the settings do not match their initial preference," the paper states. "Our results demonstrate users are not correctly able to configure the desired privacy settings of default apps. In addition, we discovered that some default app configurations can even reduce trust in family relationships."

The researchers criticize data collection by Apple apps like Safari and Siri, where that data is sent, how users can (and can't) disable that data tracking, and how Apple presents privacy options to users. The paper illustrates these issues in a discussion of Apple's Siri voice assistant. While users can ostensibly choose not to enable Siri in the initial setup on macOS-powered devices, it still collects data from other apps to provide suggestions. To fully disable Siri, Apple users must find privacy-related options across five different submenus in the Settings app. Apple's own documentation for how its privacy settings work isn't good either. It doesn't mention every privacy option, explain what is done with user data, or highlight whether settings are enabled or disabled. Also, it's written in legalese, which almost guarantees no normal user will ever read it. "We discovered that the features are not clearly documented," the paper concludes. "Specifically, we discovered that steps required to disable features of default apps are largely undocumented and the data handling practices are not completely disclosed."

An anonymous reader quotes a report from Ars Technica: Will Google ever launch its "Find My" network? The Android ecosystem was supposed to have its own version of Apple's AirTags by now. Google has had a crowd-sourced device-tracking network sitting dormant on 3 billion Android phones since December 2022. Partners have been ready to go with Bluetooth tag hardware since May 2023! This was all supposed to launch a year ago, but Google has been in a holding pattern. The good news is we're finally seeing some progress after a year of silence. The reason for Google's lengthy delay is actually Apple. A week before Google's partners announced their Android network Bluetooth tags, Google and Apple jointly announced a standard to detect "unknown" Bluetooth trackers and show users alerts if their phone thinks they're being stalked. Since you can constantly see an AirTag's location, they can be used for stalking by just covertly slipping one into a bag or car; nobody wants that, so everyone's favorite mobile duopoly is teaming up.

Google did its half of this partnership and rolled out AirTag detection in July 2023. At the same time, Google also announced: "We've made the decision to hold the rollout of the Find My Device network until Apple has implemented protections for iOS." Surely Apple would be burning the midnight oil to launch iOS Android tag detection as soon as possible so that Google could start competing with AirTags. It looks like iOS 17.5 is the magic version Google is waiting for. The first beta was released to testers recently, and 9to5Mac recently spotted strings for detecting "unwanted" non-Apple tracking devices that were suddenly following you around. This 17.5 update still needs to ship, and the expectation is sometime in May. That would be 11 months after Google's release. [...]

With the impending iOS release, Google seems to be getting its ducks in a row as well. 9to5Google has a screenshot of the new Find My Device settings page that is appearing for some users, which gives them a chance to opt out of the anonymous tracking network. That report also mentions that some users received an email Thursday of an impending tracking network launch, saying: "You'll get a notification on your Android devices when this feature is turned on in 3 days. Until then, you can opt out of the network through Find My Device on the web." The vast majority of Android users have not gotten this email, though, suggesting maybe it was a mistake. It's very weird to announce a launch in "days remaining" rather than just saying what date something will launch, and this email went out Thursday, which would mean a bizarre Sunday launch when everyone is off for the weekend.

An anonymous reader quotes a report from Kotaku: A new patent recently filed by TV and streaming device manufacturer Roku hints toward a possible future where televisions could display ads when you pause a movie or game. For Roku, the time in which the TV is on but users aren't doing anything is valuable. The company has started leasing out ad space in its popular Roku City screensaver -- which appears when your TV is idle -- to companies like McDonald's and movies like Barbie. As tech newsletter Lowpass points out, Roku finds this idle time and its screensaver so valuable that it forbids app developers from overriding the screensaver with their own. But, if you plug in an Xbox or DVD player into the HDMI port on a Roku TV, you bypass the company's screensaver and other ads. And so, Roku has been figuring out a way to not let that happen.

As reported by Lowpass on April 4, Roku recently filed a patent for a technology that would let it inject ads into third-party content -- like an Xbox game or Netflix movie -- using an HDMI connection. The patent describes a situation where you are playing a video game and hit pause to go check your phone or grab some food. At this point, Roku would identify that you have paused the content and display a relevant ad until you unpaused the game. Roku's tech isn't designed to randomly inject ads as you are playing a game or watching a movie, it knows that would be going too far and anger people. Instead, the patent suggests several ways that Roku could spot when your TV is paused, like comparing frames, to make sure the user has actually paused the content. Roku might also use the HDMI's audio feed to search for extended moments of silence. The company also proposes using HDMI CEC -- a protocol designed to help devices communicate better -- to figure out when you pause and unpause content. Similarly, Roku's patent explains that it will use various methods to detect what people are playing or watching and try to display relevant ads. So if it sees you have an Xbox plugged in, it might try to serve you ads that it thinks an Xbox owner would be interested in.

Google has introduced a new JPEG library called Jpegli, which reduces noise and improves image quality over traditional JPEGs. Proponents of the technology said it has the potential to make the Internet faster and more beautiful. InfoWorld reports: Announced April 3 and accessible from GitHub, Jpegli maintains high backward compatibility while offering enhanced capabilities and a 35% compression ratio at high-quality compression settings, Google said. Jpegli works by using new techniques to reduce noise and improve image quality. New or improved features include adaptive quantization heuristics from the JPEG XL reference implementation, improved quantization matrix selection, calculation of intermediate results, and the possibility to use more advanced colorspace.

The library provides an interoperable encoder and decoder complying with the original JPEG standard and its most convenient 8-bit formalism and API/ABI compatibility with libjeg-turbo and MozJPEG. When images are compressed or decompressed through Jpegli, more precise and psycho-visually effective computations are also performed; images will look clearer and have fewer observable artifacts. While improving on the density ratio of image quality and compression, Jpegli's coding speed is comparable to traditional approaches such as MozJPEG, according to Google. Web developers can thus integrate Jpegli into existing workflows without sacrificing coding speed, performance, or memory use.

Jpegli can be encoded with 10-plus bits per component. The 10-bit encoding happens in the original 8-bit formalism and the resulting images are interoperable with 8-bit viewers. The 10-bit dynamics are available as an API extension and application code changes are necessary to apply it. Also, Jpegli compresses images more efficiently than traditional JPEG codecs; this can save bandwidth and storage space and make web pages faster, Google said.

Many investors are lining up to bet on the collapse of former President Donald J. Trump's social media company, Trump Media & Technology Group Corp., which made its stock market debut last week under the ticker "DJT." The stock has been called the "mother of all meme stocks" since it is highly volatile and there are no fundamental underpinnings. It's being valued at roughly 1,600 times its annual revenue, at Wednesday's closing price. "By comparison, the stock of Facebook's owner trades at about eight times revenues, and Google's owner trades at six times," notes Fast Company. The New York Times reports: Trump Media is the most "shorted" special purpose acquisition vehicle in the country, according to the financial data company S3 Partners. Short-sellers bet that the price of a stock will fall. They do that by borrowing shares of a company and selling them into the market, hoping to buy them back later at a lower price, before returning the shares to the lender and pocketing the difference as profit. The demand to short Trump Media, the parent company of the social media platform Truth Social, is so great that stock lenders can charge enormous fees, making it hard for short-sellers to turn a profit unless the shares fall significantly. Still, there is a lot of interest in taking the bet. "They are looking for this stock to crater and crater very quickly," said Ihor Dusaniwsky, managing director of predictive analytics at S3. Last month, traders lost $126 million betting against Trump Media, according to S3.

On Monday, Trump Media published updated financial information, revealing little revenue, large losses and a statement from the company's independent auditor expressing "substantial doubt" about its financial viability. This appeared to galvanize investors betting against the company, as the stock slipped from its highs. But short-sellers are finding it difficult and costly to trade in Trump Media. There are roughly 137 million shares in the company, and only around five million of those are available to short-sellers. Mr. Trump owns about 60 percent of shares, and company executives also hold a chunk of the stock. Company insiders tend not to lend their shares to short-sellers. Big asset managers like BlackRock, Vanguard and State Street, which regularly lend out shares, are not major holders of Trump Media, further crimping the supply.

According to S3, 4.9 million of the roughly five million available shares are already on loan. As with any loan, when share owners lend their stock to a short-seller, they charge a fee, usually expressed as an annual interest rate on the stock's current value. Typically, the fee for borrowing stock is a fraction of a percentage point. For Trump Media, it has risen to 550 percent, Mr. Dusaniwsky said. Trump Media's stock currently trades at around $50. That means that shorting it for a month would cost more than $20 per share. For a short-seller to break even, the stock price would have to fall by almost half by early May.

There is another wrinkle, too. One large broker said much of the short trading was not an outright bet against Trump Media. Since the advent of meme-stock trading and the vilification of short-sellers that win only if popular companies lose, large investors are wary of making such trades. Instead, the current trade driving demand is designed to capture the difference between DJT's stock price and outstanding "warrants," which will give the owners the right to new stock at a fixed price as long as regulators approve the new shares. Partly because of that uncertainty, those warrants currently trade below $19, with a list of hedge funds as recent holders. Even after the high cost to borrow stock is accounted for, they are still able to profit from the $30 difference between existing stock and what the warrants are worth, assuming the warrants become registered as shares.

An anonymous reader quotes a report from 404 Media: Last week, Ernie Smith, the publisher of the website Tedium, got a "copyright infringement notice" from a law firm called Commonwealth Legal: "We're reaching out on behalf of the Intellectual Property division of a notable entity, in relation to an image connected to our client," it read. [...] In this case, though, the email didn't demand that the photo be taken down or specifically threaten a lawsuit. Instead, it demanded that Smith place a "visible and clickable link" beneath the photo in question to a website called "tech4gods" or the law firm would "take action." Smith began looking into the law firm. And he found that Commonwealth Legal is not real, and that the images of its "lawyers" are AI generated.

The threat to "activate the case No. 86342" is obviously nonsense. Beyond that, Commonwealth Legal's website looks generic and is full of stock photos, though I've seen a lot of generic template websites for real law firms. All of its lawyers have vacant, thousand-yard stares that are commonly generated by websites like This Person Does Not Exist, none of them come up in any attorney or LinkedIn searches, and the only reverse image search results for them are for a now-broken website called Generated.Photos, which offered a service to "use AI to generate people online that don't exist, change clothing and modify face and body traits. Download generated people in different postures." "All of the faces scanned were likely AI generated, most likely by a Generative Adversarial Network (GAN) model," Ali Shahriyari, cofounder and CTO of the AI detection startup Reality Defender told 404 Media. Commonwealth Legal's listed address is the fourth floor of a one-story building that looks nothing like the image on its website, and both of its phone numbers are disconnected. No one responded to the contact form that I filled out. Smith realized that what's happening here isn't a copyright enforcement or copyright trolling attempt at all. Instead, it's a backlink SEO scam, where a website owner tries to improve their Google ranking by asking, paying, or threatening someone to link to their website.

Tech4Gods.com is a gadget review website run by a man named Daniel Barczak, whose content is "complemented by AI writing assistants." In this case, the photo that Smith had "infringed" was a photo downloaded from the royalty free, free-to-use website Unsplash, which 404 Media also sometimes uses. The image was not taken by Barczak, and has nothing to do with him, he told me in an email: "I certainly don't own any images on the web," he said. The original photographer did not respond to a request for comment sent through Unsplash. Barczak told me that he had been previously buying backlinks to his website for SEO, but said he wasn't aware of who was doing this or why. "I have no idea; it certainly has nothing to do with me," he said. "However, recently, someone has been building spammy links against my site that I have been dealing with." "I have mastered on-page SEO, but unfortunately, I buy links due to a lack of time," he added. "In the past, I had a bad link builder. I wonder if it's him going mad at me for letting him go It's hard to say the web is massive, and everyone can link whenever they want." Link building is an SEO strategy devised to get outside websites to link to your website. He added that "bad links may damage [the site's] profile in Google's eyes." In this case, however, the "lawyers" were threatening a well-established tech blogger, and a link from Tedium would likely be treated as a positive in the search algorithm's eyes.

Google Books is indexing low quality, AI-generated books that will turn up in search results, and could possibly impact Google Ngram viewer, an important tool used by researchers to track language use throughout history. From a report: I was able to find the AI-generated books with the same method we've previously used to find AI-generated Amazon product reviews, papers published in academic journals, and online articles. Searching Google Books for the term "As of my last knowledge update," which is associated with ChatGPT-generated answers, returns dozens of books that include that phrase. Some of the books are about ChatGPT, machine learning, AI, and other related subjects and include the phrase because they are discussing ChatGPT and its outputs. These books appear to be written by humans. However, most of the books in the first eight pages of results turned up by the search appear to be AI-generated and are not about AI.

For example, the 2024 book Bears, Bulls, and Wolves: Stock Trading for the Twenty-Year-Old by Tristin McIver, bills itself as "a transformative journey into the world of stock trading" and "a comprehensive guide designed for beginners eager to unlock the mysteries of financial markets." In reality, it reads like ChatGPT-generated text with surface, Wikipedia-level analysis of complex financial events like Facebook's initial public offering or the 2008 financial crisis summed up in a few short paragraphs. [...] Other books appear to be outdated to the point of being useless at the time they are published because they are generated with a version of ChatGPT with an old "knowledge update."

Google parent Alphabet has been talking to its advisers about the possibility of making an offer for HubSpot, an online marketing software company with a market value of $32 billion, Reuters reported Thursday, citing people familiar with the matter. From the report: If Alphabet moves ahead with an offer, it would be a rare example of a major technology company attempting a mega deal amid heightened regulatory scrutiny of the sector under U.S. President Joe Biden's administration.

The potential deal would also allow Alphabet to put some of its cash pile, which reached $110.9 billion as of the end of December, to work. Alphabet has met with Morgan Stanley investment bankers in recent days about a potential offer for HubSpot, the sources said. It has been discussing how much it should offer and whether antitrust regulators would clear such a tie-up, the sources added.

An anonymous reader quotes a report from the Financial Times: Google is considering charging for new "premium" features powered by generative artificial intelligence, in what would be the biggest ever shake-up of its search business. The proposed revamp to its cash cow search engine would mark the first time the company has put any of its core product behind a paywall, and shows it is still grappling with a technology that threatens its advertising business, almost a year and a half after the debut of ChatGPT. Google is looking at options including adding certain AI-powered search features to its premium subscription services, which already offer access to its new Gemini AI assistant in Gmail and Docs, according to three people with knowledge of its plans. Engineers are developing the technology needed to deploy the service but executives have not yet made a final decision on whether or when to launch it, one of the people said. Google's traditional search engine would remain free of charge, while ads would continue to appear alongside search results even for subscribers. But charging would represent the first time that Google -- which for many years offered free consumer services funded entirely by advertising -- has made people pay for enhancements to its core search product. "For years, we've been reinventing Search to help people access information in the way that's most natural to them," said Google. "With our generative AI experiments in Search, we've already served billions of queries, and we're seeing positive Search query growth in all of our major markets. We're continuing to rapidly improve the product to serve new user needs."

It added: "We don't have anything to announce right now."