Words have meaning. Proper word selection is integral to strong communication, whether it's about relaying one’s feelings to another or explaining the terms of a deal, agreement, or transaction.

Language can be confusing, but typically when something is available to "buy," ownership of that good or access to that service is offered in exchange for money. That’s not really the case, though, when it comes to digital content.

Often, streaming services like Amazon Prime Video offer customers the options to “rent” digital content for a few days or to “buy” it. Some might think that picking "buy" means that they can view the content indefinitely. But these purchases are really just long-term licenses to watch the content for as long as the streaming service has the right to distribute it—which could be for years, months, or days after the transaction.

A lawsuit [PDF] recently filed against Prime Video challenges this practice and accuses the streaming service of misleading customers by labeling long-term rentals as purchases. The conclusion of the case could have implications for how streaming services frame digital content.

New lawsuit against Prime Video

On August 21, Lisa Reingold filed a proposed class-action lawsuit in the US District Court for the Eastern District of California against Amazon, alleging "false and misleading advertising." The complaint, citing Prime Video’s terms of use, reads:

On its website, Defendant tells consumers the option to 'buy' or 'purchase' digital copies of these audiovisual works. But when consumers 'buy' digital versions of audiovisual works through Amazon’s website, they do not obtain the full bundle of sticks of rights we traditionally think of as owning property. Instead, they receive 'non-exclusive, nontransferable, non-sublicensable, limited license' to access the digital audiovisual work, which is maintained at Defendant’s sole discretion.

The complaint compares buying a movie from Prime Video to buying one from a physical store. It notes that someone who buys a DVD can view the movie a decade later, but “the same cannot be said," necessarily, if they purchased the film on Prime Video. Prime Video may remove the content or replace it with a different version, such as a shorter theatrical cut.

Amazon has fought this fight before

In 2020, Amanda Caudel filed a similar complaint [PDF] in the same court as Reingold. Caudel's complaint argued that Amazon “secretly reserves the right to terminate the consumers’ access and use of the Video Content at any time, and has done so on numerous occasions, leaving the consumer without the ability to enjoy their already-bought Video Content.”

Caudel's case was dismissed [PDF] in October 2021 with the judge ruling that the plaintiff lacked “standing.”

California law reignites debate

Despite the dismissal of similar litigation years ago, Reingold's complaint stands a better chance due to a California law that took effect in January banning the selling of a "digital good to a purchaser with the terms ‘buy,’ ‘purchase,' or any other term which a reasonable person would understand to confer an unrestricted ownership interest in the digital good, or alongside an option for a time-limited rental.”

There are some instances where the law allows digital content providers to use words like "buy." One example is if, at the time of transaction, the seller receives acknowledgement from the customer that the customer is receiving a license to access the digital content; that they received a complete list of the license's conditions; and that they know that access to the digital content may be "unilaterally revoked."

A seller can also use words like "buy" if it provides to the customer ahead of the transaction a statement that "states in plain language that 'buying' or 'purchasing' the digital good is a license," as well as online access to terms and conditions, the law states.

The California legislation helps strengthen the lawsuit filed by Reingold, a California resident. The case is likely to hinge on whether or not fine print and lengthy terms of use are appropriate and sufficient communication.

Reingold's complaint acknowledges that Prime Video shows relevant fine print below its "buy" buttons but says that the notice is “far below the ‘buy movie’ button, buried at the very bottom” of the page and is not visible until “the very last stage of the transaction,” after a user has already clicked “buy.”

“This does not meet the standards set by the statute for a clear and conspicuous notice that the thing they are purchasing is a revocable license to access the digital good,” the complaint reads.

Speaking about the case with Ars Technica, Danny Karon, a consumer attorney who has taught class-action litigation at the University of Michigan Law School and Ohio State University Moritz College of Law, said:

… Amazon is sure to argue that, never mind its ‘buy’ or ‘purchase’ language, the rest of the story was described (or buried) in its terms and conditions. If plaintiff didn’t want to read her contract, including the small print, that’s on her.

The plaintiffs will be challenged to prove that Amazon's communication was ineffective and inaccurate.

"One has to prove that when they bought the movie, they were given indication that they 'owned it' in perpetuity and that they were harmed when the movie was taken away," Rahul Telang, a professor of information systems at Carnegie Mellon University’s Heinz College who has researched the digitization of copyrighted industries, told Ars.

"I think it is well established that digital products are licensed and they cannot be used as physical products. One has to prove that Amazon provided false information that they were buying a product ..."

Karon, pointing to Amazon’s choice of words, told Ars: “I like plaintiff’s chances.” He explained:

A normal consumer, after whom the California statute at issue is fashioned, would consider ‘buy’ or ‘purchase’ to involve a permanent transaction, not a mere rental, as is occurring here. If the facts are as plaintiff alleges, Amazon’s behavior would likely constitute a breach of contract or statutory fraud.

The attorney added that such complaints could be more viable if the content purchased from the plaintiffs was removed from the service.

The complaint requests a jury trial and is seeking unspecified damages, including the disgorgement of relevant profits.

A break in communication

The two lawsuits against Prime Video highlight consequential gaps in understanding among people who spend money on digital content. Both cases could contribute to potential changes in how the streaming industry describes digital content transcations moving forward. Prime Video is the target of both suits discussed, but other streaming companies, such as YouTube, also use the word “buy” when selling long-term licenses for digital content.

Streaming is now the most popular way to watch TV, yet many are unaware of what they’re buying. As Reingold's lawsuit points out, paying for content in the streaming era is different from buying media from physical stores. Physical media nets control over your ability to watch stuff for years. But you also had to have retrieved the media from a store (or website) and to have maintained that physical copy, as well the necessary hardware and/or software for playing it. Streaming services can rip purchased content from customers in bulk, but they also offer access to a much broader library that's instantly watchable with technology that most already have (like a TV and Internet).

We can debate the best approach to distributing media. What's clearer is the failure of digital content providers to ensure that customers fully understand they’re paying for access to content, and that this access could be revoked at any time.

There’s an argument to be made for thoughtful word choices that communicate clearly to customers. Ars asked Amazon if it might change its word choices on Prime Video to avoid the potential confusion embodied by the aforementioned lawsuits. We didn’t receive a response.

Read full article

Comments

Within days of joining Meta, Shengjia Zhao, co-creator of OpenAI’s ChatGPT, had threatened to quit and return to his former employer, in a blow to Mark Zuckerberg’s multibillion-dollar push to build “personal superintelligence.”

Zhao went as far as to sign employment paperwork to go back to OpenAI. Shortly afterwards, according to four people familiar with the matter, he was given the title of Meta’s new “chief AI scientist.”

The incident underscores Zuckerberg’s turbulent effort to direct the most dramatic reorganisation of Meta’s senior leadership in the group’s 20-year history.

One of the few remaining Big Tech founder-CEOs, Zuckerberg has relied on longtime acolytes such as Chief Product Officer Chris Cox to head up his favored departments and build out his upper ranks.

But in the battle to dominate AI, the billionaire is shifting towards a new and recently hired generation of executives, including Zhao, former Scale AI CEO Alexandr Wang, and former GitHub chief Nat Friedman.

Current staff are adapting to the reinvention of Meta’s AI efforts as the newcomers seek to flex their power while adjusting to the idiosyncrasies of working within a sprawling $1.95 trillion giant with a hands-on chief executive.

“There’s a lot of big men on campus,” said one investor who is close with some of Meta’s new AI leaders.

Adding to the tumult, a handful of new AI staff have already decided to leave after brief tenures, according to people familiar with the matter.

This includes Ethan Knight, a machine-learning scientist who joined the company weeks ago. Another, Avi Verma, a former OpenAI researcher, went through Meta’s onboarding process but never showed up for his first day, according to a person familiar with the matter.

In a tweet on X on Wednesday, Rishabh Agarwal, a research scientist who started at Meta in April, announced his departure. He said that while Zuckerberg and Wang’s pitch was “incredibly compelling,” he “felt the pull to take on a different kind of risk,” without giving more detail.

Meanwhile, Chaya Nayak and Loredana Crisan, generative AI staffers who had worked at Meta for nine and 10 years respectively, are among the more than half a dozen veteran employees to announce they are leaving in recent days. Wired first reported some details of recent exits, including Zhao’s threatened departure.

Meta said: “We appreciate that there’s outsized interest in seemingly every minute detail of our AI efforts, no matter how inconsequential or mundane, but we’re just focused on doing the work to deliver personal superintelligence.”

A spokesperson said Zhao had been scientific lead of the Meta superintelligence effort from the outset, and the company had waited until the team was in place before formalising his chief scientist title.

“Some attrition is normal for any organisation of this size. Most of these employees had been with the company for years, and we wish them the best,” they added.

Over the summer, Zuckerberg went on a hiring spree to coax AI researchers from rivals such as OpenAI and Apple with the promise of nine-figure sign-on bonuses and access to vast computing resources in a bid to catch up with rival labs.

This month, Meta announced it was restructuring its AI group—recently renamed Meta Superintelligence Lab (MSL)—into four distinct teams. It is the fourth overhaul of its AI efforts in six months.

“One more reorg and everything will be fixed,” joked Meta research scientist Mimansa Jaiswal on X last week. “Just one more.”

Overseeing all of Meta’s AI efforts is Wang, a well-connected and commercially minded Silicon Valley entrepreneur, who was poached by Zuckerberg as part of a $14 billion investment in his Scale data labeling group.

The 28-year-old is heading Zuckerberg’s most secretive new department known as “TBD”—shorthand for “to be determined”—which is filled with marquee hires.

In one of the new team’s first moves, Meta is no longer actively working on releasing its flagship Llama Behemoth model to the public, after it failed to perform as hoped, according to people familiar with the matter. Instead, TBD is focused on building newer cutting-edge models.

Multiple company insiders describe Zuckerberg as deeply invested and involved in the TBD team, while others criticize him for “micromanaging.”

Wang and Zuckerberg have struggled to align on a timeline to achieve the chief executive’s goal of reaching superintelligence, or AI that surpasses human capabilities, according to another person familiar with the matter. The person said Zuckerberg has urged the team to move faster.

Meta said this allegation was “manufactured tension without basis in fact that’s clearly being pushed by dramatic, navel-gazing busybodies.”

Wang’s leadership style has chafed with some, according to people familiar with the matter, who noted he does not have previous experience managing teams across a Big Tech corporation.

One former insider said some new AI recruits have felt frustrated by the company’s bureaucracy and internal competition for resources that they were promised, such as access to computing power.

“While TBD Labs is still relatively new, we believe it has the greatest compute-per-researcher in the industry, and that will only increase,” Meta said.

Wang and other former Scale staffers have struggled with some of the idiosyncratic ways of working at Meta, according to someone familiar with his thinking, for example having to adjust to not having revenue goals as they once did as a startup.

Despite teething problems, some have celebrated the leadership shift, including the appointment of popular entrepreneur and venture capitalist Friedman as head of Products and Applied Research, the team tasked with integrating the models into Meta’s own apps.

The hiring of Zhao, a top technical expert, has also been regarded as a coup by some at Meta and in the industry, who feel he has the decisiveness to propel the company’s AI development.

The shake-up has partially sidelined other Meta leaders. Yann LeCun, Meta’s chief AI scientist, has remained in the role but is now reporting into Wang.

Ahmad Al-Dahle, who led Meta’s Llama and generative AI efforts earlier in the year, has not been named as head of any teams. Cox remains chief product officer, but Wang reports directly into Zuckerberg—cutting Cox out of overseeing generative AI, an area that was previously under his purview.

Meta said that Cox “remains heavily involved” in its broader AI efforts, including overseeing its recommendation systems.

Going forward, Meta is weighing potential cuts to the AI team, one person said. In a memo shared with managers last week, seen by the Financial Times, Meta said that it was “temporarily pausing hiring across all [Meta Superintelligence Labs] teams, with the exception of business critical roles.”

Wang’s staff would evaluate requested hires on a case-by-case basis, but the freeze “will allow leadership to thoughtfully plan our 2026 headcount growth as we work through our strategy,” the memo said.

© 2025 The Financial Times Ltd. All rights reserved. Not to be redistributed, copied, or modified in any way.

Read full article

Comments

Microsoft has introduced AI models that it trained internally and says it will begin using them in some products. This announcement may represent an effort to move away from dependence on OpenAI, despite Microsoft's substantial investment in that company. It comes more than a year after insider reports revealed that Microsoft was beginning work on its own foundational models.

A post on the Microsoft AI blog describes two models. MAI-Voice-1 is a natural speech-generation model meant to deliver "high-fidelity, expressive audio across both single and multi-speaker scenarios." The idea is that voice will be one of the main ways users interact with AI tools in the future, though we haven't really seen that come to fruition so far.

The second model is called MAI-1-preview, and it's a foundational large language model specifically trained to drive Copilot, Microsoft's AI chatbot tool. It was trained on around 15,000 Nvidia H100 GPUs, and runs inference on a single GPU. As reported last year, this model is significantly larger than the models seen in Microsoft's earlier experiments, which focused on smaller models meant to run locally, like Phi-3.

To date, Copilot has primarily depended on OpenAI's models. Microsoft has invested enormous amounts of money in OpenAI, and it's unlikely the two companies will fully divorce any time soon. That said, there have been some tensions in recent months when their incentives or objectives have strayed out of alignment.

Since it's hard to predict where this is all going, it's likely to Microsoft's longterm advantage to develop its own models.

It's also possible Microsoft has introduced these models to address use cases or queries that OpenAI isn't focused on. We're seeing a gradual shift in the AI landscape toward models that are more specialized for certain tasks, rather than general, all-purpose models that are meant to be all things to all people.

These new models follow that somewhat, as Microsoft AI lead Mustafa Suleyman said in a podcast with The Verge that the goal here is "to create something that works extremely well for the consumer... my focus is on building models that really work for the consumer companion."

As such, it makes sense that we're going to see these models rolling out in Copilot, which is Microsoft's consumer-oriented AI chat bot product. Of MAI-1-preview, the Microsoft AI blog post specifies, "this model is designed to provide powerful capabilities to consumers seeking to benefit from models that specialize in following instructions and providing helpful responses to everyday queries."

So yes, MAI-1-preview has a target audience in mind, but it's still a general-purpose model since Copilot is a general-purpose tool.

MAI-Voice-1 is already being used in Microsoft's Copilot Daily and Podcasts features. There's also a Copilot Labs interface that you can visit right now to play around with it, giving it prompts or scripts and customizing what kind of voice or delivery you want to hear.

MA1-1-preview is in public testing on LMArena and will be rolled out to "certain text use cases within Copilot over the coming weeks."

Read full article

Comments

Blogger, soon after its launch in August 1999.

On January 26, 1999, Cameron Barrett — who ran a website called Camworld — pondered the meaning of a new term he’d recently discovered:

“A few months back, I heard the term weblog for the first time. I'm not sure who coined it or where it came from, so I can't properly credit it. Typically, a weblog is a small web site, usually maintained by one person that is updated on a regular basis and has a high concentration of repeat visitors. Weblogs often are highly focused around a singular subject, an underlying theme or unifying concept.”

Camworld 'anatomy of a weblog' post, January 26, 1999 (screenshot April 1999).

Thus began a multi-year process of people trying to define what a weblog is. One of the better efforts was Rebecca Blood’s article, Weblogs: A History and Perspective, published in September 2000 — when the birth of blogging in 1999 was still fresh in her mind. She noted that in 1998, there were “just a handful of sites of the type that are now identified as weblogs.”

By the start of 1999, her friend Jesse James Garrett’s "page of only weblogs" was listing out “the 23 known to be in existence at the beginning of 1999.” Also in early 1999, Brigitte Eaton compiled a list of weblogs and called it the Eatonweb Portal. According to Blood, Eaton “evaluated all submissions by a simple criterion: that the site consist of dated entries.”

Crucially, at this point weblogs tended to be run by people who knew how to wrangle code. As Blood put it:

"The original weblogs were link-driven sites. Each was a mixture in unique proportions of links, commentary, and personal thoughts and essays. Weblogs could only be created by people who already knew how to make a website."

Rebecca Blood's weblog.

Blood started her own weblog in April 1999; and it was shortly after that the word “blog” was coined. Sometime in late April or early May, San Francisco web designer Peter Merholz added a note to his sidebar:

“I've decided to pronounce the word "weblog" as wee'- blog. Or "blog" for short.”

A screenshot of Peter Merholz's site on 28 April 1999 (before he'd added the blog definition).

Around the same time, Jorn Barger — another early weblogger with his site Robot Wisdom — was pondering the utility of the format. On May 5, 1999, he wrote in a "meta" note:

"My ideal for weblogs is that everybody should keep one-- publicly or privately-- as the most efficient way of archiving good bookmarks. (Since I started keeping mine, I've hardly ever lost an URL!) If this means you copy 90% of my links, I don't mind at all if you also: 1) write your own comments rather than copying mine, and 2) include a link to me from time to time that will let your readers choose whether they want to follow RWWL here, directly."

Again, the focus is on links and a dash of commentary — but there are hints that weblogs can also help form communities. By linking to him, Barger suggests, you might share readers.

In the following paragraph, Barger uses the term "loggers"; so at this point there is no set term for people who run a weblog.

Robot Wisdom homepage.

Later that month, an article in Salon blew open the doors for blogging. On May 28, 1999, Scott Rosenberg wrote that “a phenomenon known as the weblog is one of the fastest-growing and most fertile creative areas on the Web today.” He then offered a detailed definition, emphasizing dated entries and multiple links:

“Weblogs, typically, are personal Web sites operated by individuals who compile chronological lists of links to stuff that interests them, interspersed with information, editorializing and personal asides. A good weblog is updated often, in a kind of real-time improvisation, with pointers to interesting events, pages, stories and happenings elsewhere on the Web. New stuff piles on top of the page; older stuff sinks to the bottom.”

According to Rosenberg, a weblog wasn’t just a straight “linkalist” (a term that one of his journalist colleagues derisively used) and it was also different from the often painfully raw online journals that people like Justin Hall ran. A weblog was more like an ongoing record of personal discovery on the web. Or as Rosenberg nicely put it, “a good weblog is also a window onto the mind and daily life of its creator.”

Salon's article on weblogs on May 28, 1999.

One of the early bloggers Rosenberg pointed to was Dave Winer, who not only ran his own weblog — Scripting News, started in 1997 — but was a developer of web publishing software. In his post commenting on the Salon piece, Winer pointed out that weblogs made it easier for people to publish to the web:

“The web is maturing and growing. It's becoming easier to create websites. As time goes by it will become much easier.”

Blogger Launches

By now, there were web publishing products that made it relatively simple to start a weblog. Winer’s was called Frontier (which soon spawned Radio UserLand, which would become my first blogging tool). But in the second half of 1999, it was a new product called Blogger that brought this activity to the mainstream web.

Pyra Labs, a company co-founded by twenty-something entrepreneurs Evan Williams and Meg Hourihan, launched Blogger on August 23, 1999.

Pyra's website on October 12, 1999, less than two months after the launch of Blogger.

Hourihan later confirmed that the name 'Blogger' derived from Merholz:

"The use of "blog" came from Peter. We were friends with him in San Francisco and so know about his joke to change the pronunciation of "weblog." At the time everyone said "weblog" and "weblogger" so picking up his pronounciation, Evan Williams came up with the name "Blogger" for our product."

By early November, Blogger had added a "Blog this!" bookmarklet for IE browsers on Windows.

The key to Blogger's success was that users needed no technical knowledge — all you had to do was fill in a form on the Blogger website ("No muss. No fuss."). Specifically, you didn't have to mess around with web servers. Behind the scenes, Blogger used FTP (File Transfer Protocol) to publish the contents of the form to the user's weblog.

Subscribing to Blogs via RSS

One thing early Blogger users didn't have was an RSS feed. On December 21st, Blogger developer Paul Bausch suggested a manual approach. "Blogger doesn't have automatic RSS file generation yet," he wrote, "but you can still use Blogger to easily maintain an RSS file."

Bausch linked to a separate post demonstrating the method. But since it required users to mess with template code, it wasn't very user friendly and didn't seem to fit the 'no muss, no fuss' ethos of Blogger.

Paul Bausch demonstrating how to create an RSS feed, using his own blog at onfocus.com.

A few days later, Ev Williams proudly announced that the company had created an RSS feed for "the 10 most recently updated Blogger-powered blogs." He noted that they were registering the feed "as a channel on My.Netscape.Com and My.Userland.Com (as soon as we figure out the error there)."

Part of the reason for the error was because web syndication formats were still new — and the protocol far from settled.

Netscape had launched the first RSS specification on March 15, 1999: RSS 0.90, which stood for "RDF Site Summary." The company recommended using the .rdf suffix, although this wasn't a strict requirement. RSS 0.90 was a way for publishers to add their website as a "channel" to My.Netscape.com, a new customizable version of Netcenter (Netscape's portal).

Elements of an RSS 0.90 feed, via Netscape.

The most notable thing about RSS 0.90, from an historical perspective, is that it was designed to list out linked item titles only — since that was how portals worked at the time.

Dave Winer had been experimenting with creating XML files as a way to syndicate content since December 1997. But he hadn't specced it out yet, so in February 1999 he was curious to test out Netscape's new syndication format.

Dave Winer testing RSS 0.90 in Netcenter in February 1999.

In March, he started a new syndication server (My.UserLand.Com) that followed Netscape's specification. By early June, things seemed to be going well:

"Along with Netscape's My.Netscape.Com, we are operating compatible servers that support a new web content syndication format called RSS. Our servers have been operating for three months, aggregating popular news channels and weblogs to each of our respective audiences."

But just twelve days later, Winer claimed there was "a format split". He now said that Netscape's version of RSS was "woefully inadequate."

Start of the RSS Format Wars

The primary bone of contention was that Winer wanted to syndicate much more of his content than the Netscape RSS format allowed. As he put it:

"A channel is not a series of links pointing to articles, it's a set of paragraphs that point to one or more articles per paragraph. We told Netscape this, and they said they would work with us on the next rev of the spec."

At this point, Winer wasn't calling his version "RSS" — he called it the <scriptingNews> format.

Winer demonstrated the difference in formats by showing two screenshots: Netscape's "skinny" RSS with just headline links, and his "fat" syndication format with expanded content.

Netscape's "skinny" RSS on the left, Winer's "fat" format on the right.

Netscape listened to Winer's concerns, and in July Dan Libby produced a new version called RSS 0.91. This version removed the RDF elements, made it "100% XML valid," and "included several tags from the popular <scriptingNews> format." Libby also re-parsed the acronym to mean "Rich Site Summary."

There were character limits on the new tags and My Netscape continued to only use item headings, but RSS 0.91 was a concession that other syndication hubs — like Winer's — wanted more content.

Example RSS 0.91 code; note the <description> tag within <item>, directly inspired by Winer, which allowed up to 500 characters.

In Winer's view, RSS 0.91 was still too limited. As he put it later:

"In RSS 0.91 various elements are restricted to 500 or 100 characters. There can be no more than 15 <items> in a 0.91 <channel>."

In retrospect, Winer's position was entirely correct: there was no technical reason why web syndication feeds shouldn't be full-text. But perhaps because it was tied to its portal strategy at this time, Netscape wasn't inclined to make RSS a first-class citizen of the web.

Netscape seemed to lose interest in RSS after this (the company had been acquired by AOL at the end of 1998, so things were in flux). So Winer continued to build up his own syndication format, which by the end of 1999 was called <scriptingNews> 2.0b1.

Conclusion

Needless to say, the RSS format wars were far from over — this was just the beginning, in fact. But the unsettled nature of the formats in the second half of 1999 probably explains why Blogger didn't provide RSS feeds to its users.

By the end of 1999, then, blogging was off to the races and fast becoming a mainstream web activity. RSS was slow out of the gates in comparison, but at least it had made a start. I'll cover more history of blogging and RSS in upcoming posts here on Cybercultural (and you can subscribe to our RSS feed to be notified!).