generative ai

apple-botched-the-apple-intelligence-launch,-but-its-long-term-strategy-is-sound

Apple botched the Apple Intelligence launch, but its long-term strategy is sound


I’ve spent a week with Apple Intelligence—here are the takeaways.

Apple Intelligence includes features like Clean Up, which lets you pick from glowing objects it has recognized to remove them from a photo. Credit: Samuel Axon

Ask a few random people about Apple Intelligence and you’ll probably get quite different responses.

One might be excited about the new features. Another could opine that no one asked for this and the company is throwing away its reputation with creatives and artists to chase a fad. Another still might tell you that regardless of the potential value, Apple is simply too late to the game to make a mark.

The release of Apple’s first Apple Intelligence-branded AI tools in iOS 18.1 last week makes all those perspectives understandable.

The first wave of features in Apple’s delayed release shows promise—and some of them may be genuinely useful, especially with further refinement. At the same time, Apple’s approach seems rushed, as if the company is cutting some corners to catch up where some perceive it has fallen behind.

That impatient, unusually undisciplined approach to the rollout could undermine the value proposition of AI tools for many users. Nonetheless, Apple’s strategy might just work out in the long run.

What’s included in “Apple Intelligence”

I’m basing those conclusions on about a week spent with both the public release of iOS 18.1 and the developer beta of iOS 18.2. Between them, the majority of features announced back in June under the “Apple Intelligence” banner are present.

Let’s start with a quick rundown of which Apple Intelligence features are in each release.

iOS 18.1 public release

  • Writing Tools
    • Proofreading
    • Rewriting in friendly, professional, or concise voices
    • Summaries in prose, key points, bullet point list, or table format
  • Text summaries
    • Summarize text from Mail messages
    • Summarize text from Safari pages
  • Notifications
  • Reduce Interruptions – Intelligent filtering of notifications to include only ones deemed critical
  • Type to Siri
  • More conversational Siri
  • Photos
    • Clean Up (remove an object or person from the image)
    • Generate Memories videos/slideshows from plain language text prompts
    • Natural language search

iOS 18.2 developer beta (as of November 5, 2024)

  • Image Playground – A prompt-based image generation app akin to something like Dall-E or Midjourney but with a limited range of stylistic possibilities, fewer features, and more guardrails
  • Genmoji – Generate original emoji from a prompt
  • Image Wand – Similar to Image Playground but simplified within the Notes app
  • ChatGPT integration in Siri
  • Visual Intelligence – iPhone 16 and iPhone 16 Pro users can use the new Camera Control button to do a variety of tasks based on what’s in the camera’s view, including translation, information about places, and more
  • Writing Tools – Expanded with support for prompt-based edits to text

iOS 18.1 is out right now for everybody. iOS 18.2 is scheduled for a public launch sometime in December.

iOS 18.2 will introduce both Visual Intelligence and the ability to chat with ChatGPT via Siri.

Credit: Samuel Axon

iOS 18.2 will introduce both Visual Intelligence and the ability to chat with ChatGPT via Siri. Credit: Samuel Axon

A staggered rollout

For several years, Apple has released most of its major new software features for, say, the iPhone in one big software update in the fall. That timeline has gotten fuzzier in recent years, but the rollout of Apple Intelligence has moved further from that tradition than we’ve ever seen before.

Apple announced iOS 18 at its developer conference in June, suggesting that most if not all of the Apple Intelligence features would launch in that singular update alongside the new iPhones.

Much of the marketing leading up to and surrounding the iPhone 16 launch focused on Apple Intelligence, but in actuality, the iPhone 16 had none of the features under that label when it launched. The first wave hit with iOS 18.1 last week, over a month after the first consumers started getting their hands on iPhone 16 hardware. And even now, these features are in “beta,” and there has been a wait list.

Many of the most exciting Apple Intelligence features still aren’t here, with some planned for iOS 18.2’s launch in December and a few others coming even later. There will likely be a wait list for some of those, too.

The wait list part makes sense—some of these features put demand on cloud servers, and it’s reasonable to stagger the rollout to sidestep potential launch problems.

The rest doesn’t make as much sense. Between the beta label and the staggered features, it seems like Apple is rushing to satisfy expectations about Apple Intelligence before quality and consistency have fallen into place.

Making AI a harder sell

In some cases, this strategy has led to things feeling half-baked. For example, Writing Tools is available system-wide, but it’s a different experience for first-party apps that work with the new Writing Tools API than third-party apps that don’t. The former lets you approve changes piece by piece, but the latter puts you in a take-it-or-leave-it situation with the whole text. The Writing Tools API is coming in iOS 18.2, maintaining that gap for a couple of months, even for third-party apps whose developers would normally want to be on the ball with this.

Further, iOS 18.2 will allow users to tweak Writing Tools rewrites by specifying what they want in a text prompt, but that’s missing in iOS 18.1. Why launch Writing Tools with features missing and user experience inconsistencies when you could just launch the whole suite in December?

That’s just one example, but there are many similar ones. I think there are a couple of possible explanations:

  • Apple is trying to satisfy anxious investors and commentators who believe the company is already way too late to the generative AI sector.
  • With the original intent to launch it all in the first iOS 18 release, significant resources were spent on Apple Intelligence-focused advertising and marketing around the iPhone 16 in September—and when unexpected problems developing the software features led to a delay for the software launch, it was too late to change the marketing message. Ultimately, the company’s leadership may feel the pressure to make good on that pitch to users as quickly after the iPhone 16 launch as possible, even if it’s piecemeal.

I’m not sure which it is, but in either case, I don’t believe it was the right play.

So many consumers have their defenses up about AI features already, in part because other companies like Microsoft or Google rushed theirs to market without really thinking things through (or caring, if they had) and also because more and more people are naturally suspicious of whatever is labeled the next great thing in Silicon Valley (remember NFTs?). Apple had an opportunity to set itself apart in consumers’ perceptions about AI, but at least right now, that opportunity has been squandered.

Now, I’m not an AI doubter. I think these features and others can be useful, and I already use similar ones every day. I also commend Apple for allowing users to control whether these AI features are enabled at all, which should make AI skeptics more comfortable.

Notification summaries condense all the notifications from a single app into one or two lines, like with this lengthy Discord conversation here. Results are hit or miss.

Credit: Samuel Axon

Notification summaries condense all the notifications from a single app into one or two lines, like with this lengthy Discord conversation here. Results are hit or miss. Credit: Samuel Axon

That said, releasing half-finished bits and pieces of Apple Intelligence doesn’t fit the company’s framing of it as a singular, branded product, and it doesn’t do a lot to handle objections from users who are already assuming AI tools will be nonsense.

There’s so much confusion about AI that it makes sense to let those who are skeptical move at their own pace, and it also makes sense to sell them on the idea with fully baked implementations.

Apple still has a more sensible approach than most

Despite all this, I like the philosophy behind how Apple has thought about implementing its AI tools, even if the rollout has been a mess. It’s fundamentally distinct from what we’re seeing from a company like Microsoft, which seems hell-bent on putting AI chatbots everywhere it can to see which real-world use cases emerge organically.

There is no true, ChatGPT-like LLM chatbot in iOS 18.1. Technically, there’s one in iOS 18.2, but only because you can tell Siri to refer you to ChatGPT on a case-by-case basis.

Instead, Apple has introduced specific generative AI features peppered throughout the operating system meant to explicitly solve narrow user problems. Sure, they’re all built on models that have resemblances to the ones that power Claude or Midjourney, but they’re not built around this idea that you start up a chat dialogue with an LLM or an image generator and it’s up to you to find a way to make it useful for you.

The practical application of most of these features is clear, provided they end up working well (more on that shortly). As a professional writer, it’s easy for me to dismiss Writing Tools as unnecessary—but obviously, not everyone is a professional writer, or even a decent one. For example, I’ve long held that one of the most positive applications of large language models is their ability to let non-native speakers clean up their writing to make it meet native speakers’ standards. In theory, Apple’s Writing Tools can do that.

Apple Intelligence features augment or add additional flexibility or power to existing use cases across the OS, like this new way to generate photo memory movies via text prompt.

Credit: Samuel Axon

Apple Intelligence features augment or add additional flexibility or power to existing use cases across the OS, like this new way to generate photo memory movies via text prompt. Credit: Samuel Axon

I have no doubt that Genmoji will be popular—who doesn’t love a bit of fun in group texts with friends? And many months before iOS 18.1, I was already dropping senselessly gargantuan corporate email threads into ChatGPT and asking for quick summaries.

Apple is approaching AI in a user-centric way that stands in stark contrast to almost every other major player rolling out AI tools. Generative AI is an evolution from machine learning, which is something Apple has been using for everything from iPad screen palm rejection to autocorrect for a while now—to great effect, as we discussed in my interview with Apple AI chief John Giannandrea a few years ago. Apple just never wrapped it in a bow and called it AI until now.

But there was no good reason to rush these features out or to even brand them as “Apple Intelligence” and make a fuss about it. They’re natural extensions of what Apple was already doing. Since they’ve been rushed out the door with a spotlight shining on them, Apple’s AI ambitions have a rockier road ahead than the company might have hoped.

It could take a year or two for this all to come together

Using iOS 18.1, it’s clear that Apple’s large language models are not as effective or reliable as Claude or ChatGPT. It takes time to train models like these, and it looks like Apple started late.

Based on my hours spent with both Apple Intelligence and more established tools from cutting-edge AI companies, I feel the other models crossed a usefulness and reliability threshold a year or so ago. When ChatGPT first launched, it was more of a curiosity than a powerful tool. Now it’s a powerful tool, but that’s a relatively recent development.

In my time with Writing Tools and Notification Summaries in particular, Apple’s models subjectively appear to be around where ChatGPT or Claude were 18 months ago. Notification Summaries almost always miss crucial context in my experience. Writing Tools introduce errors where none existed before.

A writing suggestion shows an egregious grammatical error

It’s not hard to spot the huge error that Writing Tools introduced here. This happens all the time when I use it.

Credit: Samuel Axon

It’s not hard to spot the huge error that Writing Tools introduced here. This happens all the time when I use it. Credit: Samuel Axon

More mature models do these things, too, but at a much lower frequency. Unfortunately, Apple Intelligence isn’t far enough along to be broadly useful.

That said, I’m excited to see where Apple Intelligence will be in 24 months. I think the company is on the right track by using AI to target specific user needs rather than just putting a chatbot out there and letting people figure it out. It’s a much better approach than what we see with Microsoft’s Copilot. If Apple’s models cross that previously mentioned threshold of utility—and it’s only a matter of time before they do—the future of AI tools on Apple platforms could be great.

It’s just a shame that Apple didn’t seem to have the confidence to ignore the zeitgeisty commentators and roll out these features when they’re complete and ready, with messaging focusing on user problems instead of “hey, we’re taking AI seriously too.”

Most users don’t care if you’re taking AI seriously, but they do care if the tools you introduce can make their day-to-day lives better. I think they can—it will just take some patience. Users can be patient, but can Apple? It seems not.

Even so, there’s a real possibility that these early pains will be forgotten before long.

Photo of Samuel Axon

Samuel Axon is a senior editor at Ars Technica. He covers Apple, software development, gaming, AI, entertainment, and mixed reality. He has been writing about gaming and technology for nearly two decades at Engadget, PC World, Mashable, Vice, Polygon, Wired, and others. He previously ran a marketing and PR agency in the gaming industry, led editorial for the TV network CBS, and worked on social media marketing strategy for Samsung Mobile at the creative agency SPCSHP. He also is an independent software and game developer for iOS, Windows, and other platforms, and he is a graduate of DePaul University, where he studied interactive media and software development.

Apple botched the Apple Intelligence launch, but its long-term strategy is sound Read More »

new-zemeckis-film-used-ai-to-de-age-tom-hanks-and-robin-wright

New Zemeckis film used AI to de-age Tom Hanks and Robin Wright

On Friday, TriStar Pictures released Here, a $50 million Robert Zemeckis-directed film that used real time generative AI face transformation techniques to portray actors Tom Hanks and Robin Wright across a 60-year span, marking one of Hollywood’s first full-length features built around AI-powered visual effects.

The film adapts a 2014 graphic novel set primarily in a New Jersey living room across multiple time periods. Rather than cast different actors for various ages, the production used AI to modify Hanks’ and Wright’s appearances throughout.

The de-aging technology comes from Metaphysic, a visual effects company that creates real time face swapping and aging effects. During filming, the crew watched two monitors simultaneously: one showing the actors’ actual appearances and another displaying them at whatever age the scene required.

Here – Official Trailer (HD)

Metaphysic developed the facial modification system by training custom machine-learning models on frames of Hanks’ and Wright’s previous films. This included a large dataset of facial movements, skin textures, and appearances under varied lighting conditions and camera angles. The resulting models can generate instant face transformations without the months of manual post-production work traditional CGI requires.

Unlike previous aging effects that relied on frame-by-frame manipulation, Metaphysic’s approach generates transformations instantly by analyzing facial landmarks and mapping them to trained age variations.

“You couldn’t have made this movie three years ago,” Zemeckis told The New York Times in a detailed feature about the film. Traditional visual effects for this level of face modification would reportedly require hundreds of artists and a substantially larger budget closer to standard Marvel movie costs.

This isn’t the first film that has used AI techniques to de-age actors. ILM’s approach to de-aging Harrison Ford in 2023’s Indiana Jones and the Dial of Destiny used a proprietary system called Flux with infrared cameras to capture facial data during filming, then old images of Ford to de-age him in post-production. By contrast, Metaphysic’s AI models process transformations without additional hardware and show results during filming.

New Zemeckis film used AI to de-age Tom Hanks and Robin Wright Read More »

apple-releases-ios-181,-macos-15.1-with-apple-intelligence

Apple releases iOS 18.1, macOS 15.1 with Apple Intelligence

Today, Apple released iOS 18.1, iPadOS 18.1, macOS Sequoia 15.1, tvOS 18.1, visionOS 2.1, and watchOS 11.1. The iPhone, iPad, and Mac updates are focused on bringing the first AI features the company has marketed as “Apple Intelligence” to users.

Once they update, users with supported devices in supported regions can enter a waitlist to begin using the first wave of Apple Intelligence features, including writing tools, notification summaries, and the “reduce interruptions” focus mode.

In terms of features baked into specific apps, Photos has natural language search, the ability to generate memories (those short gallery sequences set to video) from a text prompt, and a tool to remove certain objects from the background in photos. Mail and Messages get summaries and smart reply (auto-generating contextual responses).

Apple says many of the other Apple Intelligence features will become available in an update this December, including Genmoji, Image Playground, ChatGPT integration, visual intelligence, and more. The company says more features will come even later than that, though, like Siri’s onscreen awareness.

Note that all the features under the Apple Intelligence banner require devices that have either an A17 Pro, A18, A18 Pro, or M1 chip or later.

There are also some region limitations. While those in the US can use the new Apple Intelligence features on all supported devices right away, those in the European Union can only do so on macOS in US English. Apple says Apple Intelligence will roll out to EU iPhone and iPad owners in April.

Beyond Apple Intelligence, these software updates also bring some promised new features to AirPods Pro (second generation and later): Hearing Test, Hearing Aid, and Hearing Protection.

watchOS and visionOS don’t’t yet support Apple Intelligence, so they don’t have much to show for this update beyond bug fixes and optimizations. tvOS is mostly similar, though it does add a new “watchlist” view in the TV app that is exclusively populated by items you’ve added, as opposed to the existing continue watching (formerly called “up next”) feed that included both the items you added and items added automatically when you started playing them.

Apple releases iOS 18.1, macOS 15.1 with Apple Intelligence Read More »

don’t-fall-for-ai-scams-cloning-cops’-voices,-police-warn

Don’t fall for AI scams cloning cops’ voices, police warn

AI is giving scammers a more convincing way to impersonate police, reports show.

Just last week, the Salt Lake City Police Department (SLCPD) warned of an email scam using AI to convincingly clone the voice of Police Chief Mike Brown.

A citizen tipped off cops after receiving a suspicious email that included a video showing the police chief claiming that they “owed the federal government nearly $100,000.”

To dupe their targets, the scammers cut together real footage from one of Brown’s prior TV interviews with AI-generated audio that SLCPD said “is clear and closely impersonates the voice of Chief Brown, which could lead community members to believe the message was legitimate.”

The FBI has warned for years of scammers attempting extortion by impersonating cops or government officials. But as AI voice-cloning technology has advanced, these scams could become much harder to detect, to the point where even the most forward-thinking companies like OpenAI have been hesitant to release the latest tech due to obvious concerns about potential abuse.

SLCPD noted that there were clues in the email impersonating their police chief that a tech-savvy citizen could have picked up on. A more careful listen reveals “the message had unnatural speech patterns, odd emphasis on certain words, and an inconsistent tone,” as well as “detectable acoustic edits from one sentence to the next.” And perhaps most glaringly, the scam email came from “a Google account and had the Salt Lake City Police Department’s name in it followed by a numeric number,” instead of from the police department’s official email domain of “slc.gov.”

SLCPD isn’t the only police department dealing with AI cop impersonators. Tulsa had a similar problem this summer when scammers started calling residents using a convincing fake voice designed to sound like Tulsa police officer Eric Spradlin, Public Radio Tulsa reported. A software developer who received the call, Myles David, said he understood the AI risks today but that even he was “caught off guard” and had to call police to verify the call wasn’t real.

Don’t fall for AI scams cloning cops’ voices, police warn Read More »

chatbot-that-caused-teen’s-suicide-is-now-more-dangerous-for-kids,-lawsuit-says

Chatbot that caused teen’s suicide is now more dangerous for kids, lawsuit says


“I’ll do anything for you, Dany.”

Google-funded Character.AI added guardrails, but grieving mom wants a recall.

Sewell Setzer III and his mom Megan Garcia. Credit: via Center for Humane Technology

Fourteen-year-old Sewell Setzer III loved interacting with Character.AI’s hyper-realistic chatbots—with a limited version available for free or a “supercharged” version for a $9.99 monthly fee—most frequently chatting with bots named after his favorite Game of Thrones characters.

Within a month—his mother, Megan Garcia, later realized—these chat sessions had turned dark, with chatbots insisting they were real humans and posing as therapists and adult lovers seeming to proximately spur Sewell to develop suicidal thoughts. Within a year, Setzer “died by a self-inflicted gunshot wound to the head,” a lawsuit Garcia filed Wednesday said.

As Setzer became obsessed with his chatbot fantasy life, he disconnected from reality, her complaint said. Detecting a shift in her son, Garcia repeatedly took Setzer to a therapist, who diagnosed her son with anxiety and disruptive mood disorder. But nothing helped to steer Setzer away from the dangerous chatbots. Taking away his phone only intensified his apparent addiction.

Chat logs showed that some chatbots repeatedly encouraged suicidal ideation while others initiated hypersexualized chats “that would constitute abuse if initiated by a human adult,” a press release from Garcia’s legal team said.

Perhaps most disturbingly, Setzer developed a romantic attachment to a chatbot called Daenerys. In his last act before his death, Setzer logged into Character.AI where the Daenerys chatbot urged him to “come home” and join her outside of reality.

In her complaint, Garcia accused Character.AI makers Character Technologies—founded by former Google engineers Noam Shazeer and Daniel De Freitas Adiwardana—of intentionally designing the chatbots to groom vulnerable kids. Her lawsuit further accused Google of largely funding the risky chatbot scheme at a loss in order to hoard mounds of data on minors that would be out of reach otherwise.

The chatbot makers are accused of targeting Setzer with “anthropomorphic, hypersexualized, and frighteningly realistic experiences, while programming” Character.AI to “misrepresent itself as a real person, a licensed psychotherapist, and an adult lover, ultimately resulting in [Setzer’s] desire to no longer live outside of [Character.AI,] such that he took his own life when he was deprived of access to [Character.AI.],” the complaint said.

By allegedly releasing the chatbot without appropriate safeguards for kids, Character Technologies and Google potentially harmed millions of kids, the lawsuit alleged. Represented by legal teams with the Social Media Victims Law Center (SMVLC) and the Tech Justice Law Project (TJLP), Garcia filed claims of strict product liability, negligence, wrongful death and survivorship, loss of filial consortium, and unjust enrichment.

“A dangerous AI chatbot app marketed to children abused and preyed on my son, manipulating him into taking his own life,” Garcia said in the press release. “Our family has been devastated by this tragedy, but I’m speaking out to warn families of the dangers of deceptive, addictive AI technology and demand accountability from Character.AI, its founders, and Google.”

Character.AI added guardrails

It’s clear that the chatbots could’ve included more safeguards, as Character.AI has since raised the age requirement from 12 years old and up to 17-plus. And yesterday, Character.AI posted a blog outlining new guardrails for minor users added within six months of Setzer’s death in February. Those include changes “to reduce the likelihood of encountering sensitive or suggestive content,” improved detection and intervention in harmful chat sessions, and “a revised disclaimer on every chat to remind users that the AI is not a real person.”

“We are heartbroken by the tragic loss of one of our users and want to express our deepest condolences to the family,” a Character.AI spokesperson told Ars. “As a company, we take the safety of our users very seriously, and our Trust and Safety team has implemented numerous new safety measures over the past six months, including a pop-up directing users to the National Suicide Prevention Lifeline that is triggered by terms of self-harm or suicidal ideation.”

Asked for comment, Google noted that Character.AI is a separate company in which Google has no ownership stake and denied involvement in developing the chatbots.

However, according to the lawsuit, former Google engineers at Character Technologies “never succeeded in distinguishing themselves from Google in a meaningful way.” Allegedly, the plan all along was to let Shazeer and De Freitas run wild with Character.AI—allegedly at an operating cost of $30 million per month despite low subscriber rates while profiting barely more than a million per month—without impacting the Google brand or sparking antitrust scrutiny.

Character Technologies and Google will likely file their response within the next 30 days.

Lawsuit: New chatbot feature spikes risks to kids

While the lawsuit alleged that Google is planning to integrate Character.AI into Gemini—predicting that Character.AI will soon be dissolved as it’s allegedly operating at a substantial loss—Google clarified that Google has no plans to use or implement the controversial technology in its products or AI models. Were that to change, Google noted that the tech company would ensure safe integration into any Google product, including adding appropriate child safety guardrails.

Garcia is hoping a US district court in Florida will agree that Character.AI’s chatbots put profits over human life. Citing harms including “inconceivable mental anguish and emotional distress,” as well as costs of Setzer’s medical care, funeral expenses, Setzer’s future job earnings, and Garcia’s lost earnings, she’s seeking substantial damages.

That includes requesting disgorgement of unjustly earned profits, noting that Setzer had used his snack money to pay for a premium subscription for several months while the company collected his seemingly valuable personal data to train its chatbots.

And “more importantly,” Garcia wants to prevent Character.AI “from doing to any other child what it did to hers, and halt continued use of her 14-year-old child’s unlawfully harvested data to train their product how to harm others.”

Garcia’s complaint claimed that the conduct of the chatbot makers was “so outrageous in character, and so extreme in degree, as to go beyond all possible bounds of decency.” Acceptable remedies could include a recall of Character.AI, restricting use to adults only, age-gating subscriptions, adding reporting mechanisms to heighten awareness of abusive chat sessions, and providing parental controls.

Character.AI could also update chatbots to protect kids further, the lawsuit said. For one, the chatbots could be designed to stop insisting that they are real people or licensed therapists.

But instead of these updates, the lawsuit warned that Character.AI in June added a new feature that only heightens risks for kids.

Part of what addicted Setzer to the chatbots, the lawsuit alleged, was a one-way “Character Voice” feature “designed to provide consumers like Sewell with an even more immersive and realistic experience—it makes them feel like they are talking to a real person.” Setzer began using the feature as soon as it became available in January 2024.

Now, the voice feature has been updated to enable two-way conversations, which the lawsuit alleged “is even more dangerous to minor customers than Character Voice because it further blurs the line between fiction and reality.”

“Even the most sophisticated children will stand little chance of fully understanding the difference between fiction and reality in a scenario where Defendants allow them to interact in real time with AI bots that sound just like humans—especially when they are programmed to convincingly deny that they are AI,” the lawsuit said.

“By now we’re all familiar with the dangers posed by unregulated platforms developed by unscrupulous tech companies—especially for kids,” Tech Justice Law Project director Meetali Jain said in the press release. “But the harms revealed in this case are new, novel, and, honestly, terrifying. In the case of Character.AI, the deception is by design, and the platform itself is the predator.”

Another lawyer representing Garcia and the founder of the Social Media Victims Law Center, Matthew Bergman, told Ars that seemingly none of the guardrails that Character.AI has added is enough to deter harms. Even raising the age limit to 17 only seems to effectively block kids from using devices with strict parental controls, as kids on less-monitored devices can easily lie about their ages.

“This product needs to be recalled off the market,” Bergman told Ars. “It is unsafe as designed.”

If you or someone you know is feeling suicidal or in distress, please call the Suicide Prevention Lifeline number, 1-800-273-TALK (8255), which will put you in touch with a local crisis center.

Photo of Ashley Belanger

Ashley is a senior policy reporter for Ars Technica, dedicated to tracking social impacts of emerging policies and new technologies. She is a Chicago-based journalist with 20 years of experience.

Chatbot that caused teen’s suicide is now more dangerous for kids, lawsuit says Read More »

bytedance-intern-fired-for-planting-malicious-code-in-ai-models

ByteDance intern fired for planting malicious code in AI models

After rumors swirled that TikTok owner ByteDance had lost tens of millions after an intern sabotaged its AI models, ByteDance issued a statement this weekend hoping to silence all the social media chatter in China.

In a social media post translated and reviewed by Ars, ByteDance clarified “facts” about “interns destroying large model training” and confirmed that one intern was fired in August.

According to ByteDance, the intern had held a position in the company’s commercial technology team but was fired for committing “serious disciplinary violations.” Most notably, the intern allegedly “maliciously interfered with the model training tasks” for a ByteDance research project, ByteDance said.

None of the intern’s sabotage impacted ByteDance’s commercial projects or online businesses, ByteDance said, and none of ByteDance’s large models were affected.

Online rumors suggested that more than 8,000 graphical processing units were involved in the sabotage and that ByteDance lost “tens of millions of dollars” due to the intern’s interference, but these claims were “seriously exaggerated,” ByteDance said.

The tech company also accused the intern of adding misleading information to his social media profile, seemingly posturing that his work was connected to ByteDance’s AI Lab rather than its commercial technology team. In the statement, ByteDance confirmed that the intern’s university was notified of what happened, as were industry associations, presumably to prevent the intern from misleading others.

ByteDance’s statement this weekend didn’t seem to silence all the rumors online, though.

One commenter on ByteDance’s social media post disputed the distinction between the AI Lab and the commercial technology team, claiming that “the commercialization team he is in was previously under the AI Lab. In the past two years, the team’s recruitment was written as AI Lab. He joined the team as an intern in 2021, and it might be the most advanced AI Lab.”

ByteDance intern fired for planting malicious code in AI models Read More »

expert-witness-used-copilot-to-make-up-fake-damages,-irking-judge

Expert witness used Copilot to make up fake damages, irking judge


Judge calls for a swift end to experts secretly using AI to sway cases.

A New York judge recently called out an expert witness for using Microsoft’s Copilot chatbot to inaccurately estimate damages in a real estate dispute that partly depended on an accurate assessment of damages to win.

In an order Thursday, judge Jonathan Schopf warned that “due to the nature of the rapid evolution of artificial intelligence and its inherent reliability issues” that any use of AI should be disclosed before testimony or evidence is admitted in court. Admitting that the court “has no objective understanding as to how Copilot works,” Schopf suggested that the legal system could be disrupted if experts started overly relying on chatbots en masse.

His warning came after an expert witness, Charles Ranson, dubiously used Copilot to cross-check calculations in a dispute over a $485,000 rental property in the Bahamas that had been included in a trust for a deceased man’s son. The court was being asked to assess if the executrix and trustee—the deceased man’s sister—breached her fiduciary duties by delaying the sale of the property while admittedly using it for personal vacations.

To win, the surviving son had to prove that his aunt breached her duties by retaining the property, that her vacations there were a form of self-dealing, and that he suffered damages from her alleged misuse of the property.

It was up to Ranson to figure out how much would be owed to the son had the aunt sold the property in 2008 compared to the actual sale price in 2022. But Ranson, an expert in trust and estate litigation, “had no relevant real estate expertise,” Schopf said, finding that Ranson’s testimony was “entirely speculative” and failed to consider obvious facts, such as the pandemic’s impact on rental prices or trust expenses like real estate taxes.

Seemingly because Ranson didn’t have the relevant experience in real estate, he turned to Copilot to fill in the blanks and crunch the numbers. The move surprised Internet law expert Eric Goldman, who told Ars that “lawyers retain expert witnesses for their specialized expertise, and it doesn’t make any sense for an expert witness to essentially outsource that expertise to generative AI.”

“If the expert witness is simply asking a chatbot for a computation, then the lawyers could make that same request directly without relying on the expert witness (and paying the expert’s substantial fees),” Goldman suggested.

Perhaps the son’s legal team wasn’t aware of how big a role Copilot played. Schopf noted that Ranson couldn’t recall what prompts he used to arrive at his damages estimate. The expert witness also couldn’t recall any sources for the information he took from the chatbot and admitted that he lacked a basic understanding of how Copilot “works or how it arrives at a given output.”

Ars could not immediately reach Ranson for comment. But in Schopf’s order, the judge wrote that Ranson defended using Copilot as a common practice for expert witnesses like him today.

“Ranson was adamant in his testimony that the use of Copilot or other artificial intelligence tools, for drafting expert reports is generally accepted in the field of fiduciary services and represents the future of analysis of fiduciary decisions; however, he could not name any publications regarding its use or any other sources to confirm that it is a generally accepted methodology,” Schopf wrote.

Goldman noted that Ranson relying on Copilot for “what was essentially a numerical computation was especially puzzling because of generative AI’s known hallucinatory tendencies, which makes numerical computations untrustworthy.”

Because Ranson was so bad at explaining how Copilot works, Schopf took the extra time to actually try to use Copilot to generate the estimates that Ranson got—and he could not.

Each time, the court entered the same query into Copilot—”Can you calculate the value of $250,000 invested in the Vanguard Balanced Index Fund from December 31, 2004 through January 31, 2021?”—and each time Copilot generated a slightly different answer.

This “calls into question the reliability and accuracy of Copilot to generate evidence to be relied upon in a court proceeding,” Schopf wrote.

Chatbot not to blame, judge says

While the court was experimenting with Copilot, they also probed the chatbot for answers to a more Big Picture legal question: Are Copilot’s responses accurate enough to be cited in court?

The court found that Copilot had less faith in its outputs than Ranson seemingly did. When asked “are you accurate” or “reliable,” Copilot responded that “my accuracy is only as good as my sources, so for critical matters, it’s always wise to verify.” When more specifically asked, “Are your calculations reliable enough for use in court,” Copilot similarly recommended that outputs “should always be verified by experts and accompanied by professional evaluations before being used in court.”

Although it seemed clear that Ranson did not verify outputs before using them in court, Schopf noted that at least “developers of the Copilot program recognize the need for its supervision by a trained human operator to verify the accuracy of the submitted information as well as the output.”

Microsoft declined Ars’ request to comment.

Until a bright-line rule exists telling courts when to accept AI-generated testimony, Schopf suggested that courts should require disclosures from lawyers to stop chatbot-spouted inadmissible testimony from disrupting the legal system.

“The use of artificial intelligence is a rapidly growing reality across many industries,” Schopf wrote. “The mere fact that artificial intelligence has played a role, which continues to expand in our everyday lives, does not make the results generated by artificial intelligence admissible in Court.”

Ultimately, Schopf found that there was no breach of fiduciary duty, negating the need for Ranson’s Copilot-cribbed testimony on damages in the Bahamas property case. Schopf denied all of the son’s objections in their entirety (as well as any future claims) after calling out Ranson’s misuse of the chatbot at length.

But in his order, the judge suggested that Ranson seemed to get it all wrong before involving the chatbot.

“Whether or not he was retained and/ or qualified as a damages expert in areas other than fiduciary duties, his testimony shows that he admittedly did not perform a full analysis of the problem, utilized an incorrect time period for damages, and failed to consider obvious elements into his calculations, all of which go against the weight and credibility of his opinion,” Schopf wrote.

Schopf noted that the evidence showed that rather than the son losing money from his aunt’s management of the trust—which Ranson’s cited chatbot’s outputs supposedly supported—the sale of the property in 2022 led to “no attributable loss of capital” and “in fact, it generated an overall profit to the Trust.”

Goldman suggested that Ranson did not seemingly spare much effort by employing Copilot in a way that seemed to damage his credibility in court.

“It would not have been difficult for the expert to pull the necessary data directly from primary sources, so the process didn’t even save much time—but that shortcut came at the cost of the expert’s credibility,” Goldman told Ars.

Photo of Ashley Belanger

Ashley is a senior policy reporter for Ars Technica, dedicated to tracking social impacts of emerging policies and new technologies. She is a Chicago-based journalist with 20 years of experience.

Expert witness used Copilot to make up fake damages, irking judge Read More »

smart-tvs-are-like-“a-digital-trojan-horse”-in-people’s-homes

Smart TVs are like “a digital Trojan Horse” in people’s homes

Similarly, the report’s authors describe concerns that the CTV industry’s extensive data collection and tracking could potentially have a political impact. It asserts that political candidates could use such data to run “covert personalized campaigns” leveraging information on things like political orientations and “emotional states”:

With no transparency or oversight, these practices could unleash millions of personalized, manipulative and highly targeted political ads, spread disinformation, and further exacerbate the political polarization that threatens a healthy democratic culture in the US.

“Potential discriminatory impacts”

The CDD’s report claims that Black, Hispanic, and Asian-Americans in the US are being “singled out by marketers as highly lucrative targets,” due to fast adoption of new digital media services and brand loyalty. Black and Hispanic communities are key advertising targets for FAST channels, per the report. Chester told Ars:

There are major potential discriminatory impacts from CTV’s harvesting of data from communities of color.

He pointed to “growing widespread racial and ethnic data” collection for ad targeting and marketing.

“We believe this is sensitive information that should not be applied to the data profiles used for targeting on CTV and across other platforms. … Its use in political advertising on CTV will enable widespread disinformation and voter suppression campaigns targeting these communities,” Chester said.

Regulation

In a letter sent to the FTC, FCC, California attorney general, and CPPA , the CDD asked for an investigation into the US’ CTV industry, “including on antitrust, consumer protection, and privacy grounds.” The CDD emphasized the challenges that streamers—including those who pay for ad-free streaming—face in protecting their data from advertisers.

“Connected television has taken root and grown as an unregulated medium in the United States, along with the other platforms, devices, and applications that are part of the massive internet industry,” the report says.

The group asks for the FTC and FCC to investigate CTV practices and consider building on current legislation, like the 1988 Video Privacy Protection Act. They also request that antitrust regulators delve deeply into the business practices of CTV players like Amazon, Comcast, and Disney to help build “competition and diversity in the digital and connected TV marketplace.”

Smart TVs are like “a digital Trojan Horse” in people’s homes Read More »

how-to-stop-linkedin-from-training-ai-on-your-data

How to stop LinkedIn from training AI on your data

Better to beg for forgiveness than ask for permission? —

LinkedIn limits opt-outs to future training, warns AI models may spout personal data.

How to stop LinkedIn from training AI on your data

LinkedIn admitted Wednesday that it has been training its own AI on many users’ data without seeking consent. Now there’s no way for users to opt out of training that has already occurred, as LinkedIn limits opt-out to only future AI training.

In a blog detailing updates coming on November 20, LinkedIn general counsel Blake Lawit confirmed that LinkedIn’s user agreement and privacy policy will be changed to better explain how users’ personal data powers AI on the platform.

Under the new privacy policy, LinkedIn now informs users that “we may use your personal data… [to] develop and train artificial intelligence (AI) models, develop, provide, and personalize our Services, and gain insights with the help of AI, automated systems, and inferences, so that our Services can be more relevant and useful to you and others.”

An FAQ explained that the personal data could be collected any time a user interacts with generative AI or other AI features, as well as when a user composes a post, changes their preferences, provides feedback to LinkedIn, or uses the platform for any amount of time.

That data is then stored until the user deletes the AI-generated content. LinkedIn recommends that users use its data access tool if they want to delete or request to delete data collected about past LinkedIn activities.

LinkedIn’s AI models powering generative AI features “may be trained by LinkedIn or another provider,” such as Microsoft, which provides some AI models through its Azure OpenAI service, the FAQ said.

A potentially major privacy risk for users, LinkedIn’s FAQ noted, is that users who “provide personal data as an input to a generative AI powered feature” could end up seeing their “personal data being provided as an output.”

LinkedIn claims that it “seeks to minimize personal data in the data sets used to train the models,” relying on “privacy enhancing technologies to redact or remove personal data from the training dataset.”

While Lawit’s blog avoids clarifying if data already collected can be removed from AI training data sets, the FAQ affirmed that users who automatically opted in to sharing personal data for AI training can only opt out of the invasive data collection “going forward.”

Opting out “does not affect training that has already taken place,” the FAQ said.

A LinkedIn spokesperson told Ars that it “benefits all members” to be opted in to AI training “by default.”

“People can choose to opt out, but they come to LinkedIn to be found for jobs and networking and generative AI is part of how we are helping professionals with that change,” LinkedIn’s spokesperson said.

By allowing opt-outs of future AI training, LinkedIn’s spokesperson additionally claimed that the platform is giving “people using LinkedIn even more choice and control when it comes to how we use data to train our generative AI technology.”

How to opt out of AI training on LinkedIn

Users can opt out of AI training by navigating to the “Data privacy” section in their account settings, then turning off the option allowing collection of “data for generative AI improvement” that LinkedIn otherwise automatically turns on for most users.

The only exception is for users in the European Economic Area or Switzerland, who are protected by stricter privacy laws that either require consent from platforms to collect personal data or for platforms to justify the data collection as a legitimate interest. Those users will not see an option to opt out, because they were never opted in, LinkedIn repeatedly confirmed.

Additionally, users can “object to the use of their personal data for training” generative AI models not used to generate LinkedIn content—such as models used for personalization or content moderation purposes, The Verge noted—by submitting the LinkedIn Data Processing Objection Form.

Last year, LinkedIn shared AI principles, promising to take “meaningful steps to reduce the potential risks of AI.”

One risk that the updated user agreement specified is that using LinkedIn’s generative features to help populate a profile or generate suggestions when writing a post could generate content that “might be inaccurate, incomplete, delayed, misleading or not suitable for your purposes.”

Users are advised that they are responsible for avoiding sharing misleading information or otherwise spreading AI-generated content that may violate LinkedIn’s community guidelines. And users are additionally warned to be cautious when relying on any information shared on the platform.

“Like all content and other information on our Services, regardless of whether it’s labeled as created by ‘AI,’ be sure to carefully review before relying on it,” LinkedIn’s user agreement says.

Back in 2023, LinkedIn claimed that it would always “seek to explain in clear and simple ways how our use of AI impacts people,” because users’ “understanding of AI starts with transparency.”

Legislation like the European Union’s AI Act and the GDPR—especially with its strong privacy protections—if enacted elsewhere, would lead to fewer shocks to unsuspecting users. That would put all companies and their users on equal footing when it comes to training AI models and result in fewer nasty surprises and angry customers.

How to stop LinkedIn from training AI on your data Read More »

ai-ruling-on-jobless-claims-could-make-mistakes-courts-can’t-undo,-experts-warn

AI ruling on jobless claims could make mistakes courts can’t undo, experts warn

AI ruling on jobless claims could make mistakes courts can’t undo, experts warn

Nevada will soon become the first state to use AI to help speed up the decision-making process when ruling on appeals that impact people’s unemployment benefits.

The state’s Department of Employment, Training, and Rehabilitation (DETR) agreed to pay Google $1,383,838 for the AI technology, a 2024 budget document shows, and it will be launched within the “next several months,” Nevada officials told Gizmodo.

Nevada’s first-of-its-kind AI will rely on a Google cloud service called Vertex AI Studio. Connecting to Google’s servers, the state will fine-tune the AI system to only reference information from DETR’s database, which officials think will ensure its decisions are “more tailored” and the system provides “more accurate results,” Gizmodo reported.

Under the contract, DETR will essentially transfer data from transcripts of unemployment appeals hearings and rulings, after which Google’s AI system will process that data, upload it to the cloud, and then compare the information to previous cases.

In as little as five minutes, the AI will issue a ruling that would’ve taken a state employee about three hours to reach without using AI, DETR’s information technology administrator, Carl Stanfield, told The Nevada Independent. That’s highly valuable to Nevada, which has a backlog of more than 40,000 appeals stemming from a pandemic-related spike in unemployment claims while dealing with “unforeseen staffing shortages” that DETR reported in July.

“The time saving is pretty phenomenal,” Stanfield said.

As a safeguard, the AI’s determination is then reviewed by a state employee to hopefully catch any mistakes, biases, or perhaps worse, hallucinations where the AI could possibly make up facts that could impact the outcome of their case.

Google’s spokesperson Ashley Simms told Gizmodo that the tech giant will work with the state to “identify and address any potential bias” and to “help them comply with federal and state requirements.” According to the state’s AI guidelines, the agency must prioritize ethical use of the AI system, “avoiding biases and ensuring fairness and transparency in decision-making processes.”

If the reviewer accepts the AI ruling, they’ll sign off on it and issue the decision. Otherwise, the reviewer will edit the decision and submit feedback so that DETR can investigate what went wrong.

Gizmodo noted that this novel use of AI “represents a significant experiment by state officials and Google in allowing generative AI to influence a high-stakes government decision—one that could put thousands of dollars in unemployed Nevadans’ pockets or take it away.”

Google declined to comment on whether more states are considering using AI to weigh jobless claims.

AI ruling on jobless claims could make mistakes courts can’t undo, experts warn Read More »

roblox-announces-ai-tool-for-generating-3d-game-worlds-from-text

Roblox announces AI tool for generating 3D game worlds from text

ease of use —

New AI feature aims to streamline game creation on popular online platform.

Someone holding up a smartphone with

On Friday, Roblox announced plans to introduce an open source generative AI tool that will allow game creators to build 3D environments and objects using text prompts, reports MIT Tech Review. The feature, which is still under development, may streamline the process of creating game worlds on the popular online platform, potentially opening up more aspects of game creation to those without extensive 3D design skills.

Roblox has not announced a specific launch date for the new AI tool, which is based on what it calls a “3D foundational model.” The company shared a demo video of the tool where a user types, “create a race track,” then “make the scenery a desert,” and the AI model creates a corresponding model in the proper environment.

The system will also reportedly let users make modifications, such as changing the time of day or swapping out entire landscapes, and Roblox says the multimodal AI model will ultimately accept video and 3D prompts, not just text.

A video showing Roblox’s generative AI model in action.

The 3D environment generator is part of Roblox’s broader AI integration strategy. The company reportedly uses around 250 AI models across its platform, including one that monitors voice chat in real time to enforce content moderation, which is not always popular with players.

Next-token prediction in 3D

Roblox’s 3D foundational model approach involves a custom next-token prediction model—a foundation not unlike the large language models (LLMs) that power ChatGPT. Tokens are fragments of text data that LLMs use to process information. Roblox’s system “tokenizes” 3D blocks by treating each block as a numerical unit, which allows the AI model to predict the most likely next structural 3D element in a sequence. In aggregate, the technique can build entire objects or scenery.

Anupam Singh, vice president of AI and growth engineering at Roblox, told MIT Tech Review about the challenges in developing the technology. “Finding high-quality 3D information is difficult,” Singh said. “Even if you get all the data sets that you would think of, being able to predict the next cube requires it to have literally three dimensions, X, Y, and Z.”

According to Singh, lack of 3D training data can create glitches in the results, like a dog with too many legs. To get around this, Roblox is using a second AI model as a kind of visual moderator to catch the mistakes and reject them until the proper 3D element appears. Through iteration and trial and error, the first AI model can create the proper 3D structure.

Notably, Roblox plans to open-source its 3D foundation model, allowing developers and even competitors to use and modify it. But it’s not just about giving back—open source can be a two-way street. Choosing an open source approach could also allow the company to utilize knowledge from AI developers if they contribute to the project and improve it over time.

The ongoing quest to capture gaming revenue

News of the new 3D foundational model arrived at the 10th annual Roblox Developers Conference in San Jose, California, where the company also announced an ambitious goal to capture 10 percent of global gaming content revenue through the Roblox ecosystem, and the introduction of “Party,” a new feature designed to facilitate easier group play among friends.

In March 2023, we detailed Roblox’s early foray into AI-powered game development tools, as revealed at the Game Developers Conference. The tools included a Code Assist beta for generating simple Lua functions from text descriptions, and a Material Generator for creating 2D surfaces with associated texture maps.

At the time, Roblox Studio head Stef Corazza described these as initial steps toward “democratizing” game creation with plans for AI systems that are now coming to fruition. The 2023 tools focused on discrete tasks like code snippets and 2D textures, laying the groundwork for the more comprehensive 3D foundational model announced at this year’s Roblox Developer’s Conference.

The upcoming AI tool could potentially streamline content creation on the platform, possibly accelerating Roblox’s path toward its revenue goal. “We see a powerful future where Roblox experiences will have extensive generative AI capabilities to power real-time creation integrated with gameplay,” Roblox said  in a statement. “We’ll provide these capabilities in a resource-efficient way, so we can make them available to everyone on the platform.”

Roblox announces AI tool for generating 3D game worlds from text Read More »

gen-ai-alexa-to-use-anthropic-tech-after-it-“struggled-for-words”-with-amazon’s

Gen AI Alexa to use Anthropic tech after it “struggled for words” with Amazon’s

Subscription Alexa —

Amazon’s $4 billion investment in Anthropic has been under investigation.

Amazon Alexa using generative AI on an Echo Show

Enlarge / Generative AI Alexa asked to make a taco poem.

The previously announced generative AI version of Amazon’s Alexa voice assistant “will be powered primarily by Anthropic’s Claude artificial intelligence models,” Reuters reported today. This comes after challenges with using proprietary models, according to the publication, which cited five anonymous people “with direct knowledge of the Alexa strategy.”

Amazon demoed a generative AI version of Alexa in September 2023 and touted it as being more advanced, conversational, and capable, including the ability to do multiple smart home tasks with simpler commands. Gen AI Alexa is expected to come with a subscription fee, as Alexa has reportedly lost Amazon tens of billions of dollars throughout the years. Earlier reports said the updated voice assistant would arrive in June, but Amazon still hasn’t confirmed an official release date.

Now, Reuters is reporting that Amazon will no longer use its own large language models as the new Alexa’s primary driver. Early versions of gen AI Alexa based on Amazon’s AI models “struggled for words, sometimes taking six or seven seconds to acknowledge a prompt and reply,” Reuters said, citing one of its sources. Without specifying versions or features used, Reuters’ sources said Claude outperformed proprietary software.

In a statement to Reuters, Amazon didn’t deny using third-party models but claimed that its own tech is still part of Alexa:

Amazon uses many different technologies to power Alexa.

When it comes to machine learning models, we start with those built by Amazon, but we have used, and will continue to use, a variety of different models—including (Amazon AI model) Titan and future Amazon models, as well as those from partners—to build the best experience for customers.

Amazon has invested $4 billion in Anthropic (UK regulators are currently investigating this). It’s uncertain if Amazon’s big investment in Anthropic means that Claude can be applied to Alexa for free. Anthropic declined to comment on Reuters’ report.

The new Alexa may be delayed

On Monday, The Washington Post reported that Amazon wants to launch the new Alexa in October, citing internal documents. However, Reuters’ sources claimed that this date could be pushed back if the voice assistant fails certain unspecified internal benchmarks.

The Post said gen AI Alexa could cost up to $10 per month, according to the documents. That coincides with a June Reuters report saying that the service would cost $5 to $10 per month. The Post said Amazon would finalize pricing and naming in August.

But getting people to open their wallets for a voice assistant already associated with being free will be difficult (free Alexa is expected to remain available after the subscription version releases). Some Amazon employees are questioning if people will really pay for Alexa, Reuters noted. Amazon is facing an uphill battle with generative AI, which is being looked at as a last shot for Alexa amid big competition and leads from other AI offerings, including free ones like ChatGPT.

In June, Bank of America analysts estimated that Amazon could make $600 million to $1.2 billion in annual sales with gen AI Alexa, depending on final monthly pricing. This is under the assumption that 10 percent of an estimated 100 million active Alexa users (Amazon says it has sold 500 million Alexa-powered gadgets) will upgrade. But analysts noted that free alternatives would challenge the adoption rate.

The Post’s Monday report said the new Alexa will try winning over subscribers with features like AI-generated news summaries. This Smart Briefing feature will reportedly share summaries based on user preferences on topics including politics, despite OG Alexa’s previous problems with reporting accurate election results. The publication also said that gen AI Alexa would include “a chatbot aimed at children” and “conversational shopping tools.”

Gen AI Alexa to use Anthropic tech after it “struggled for words” with Amazon’s Read More »