An anonymous reader shares a report: Google has announced Pixel Screenshots, a new AI-powered app for its Pixel 9 lineup that lets you save, organize, and surface information from screenshots. Pixel Screenshot uses Google's private, on-device Gemini Nano AI model to analyze the content of an image and make it searchable.

During a demo at its Pixel launch event, Google showed how you can take a screenshot and then save it to a collection, like "gift ideas." You can also search through all your other screenshots by typing in a keyword, like "bikes" or "shoes." Pixel Screenshots will then pull up all relevant results. Additionally, Pixel Screenshots can give you information about what's inside an image. Further reading: Microsoft Postpones Windows Recall After Major Backlash.

Google unveiled its latest Pixel smartphone series on Tuesday, introducing four new models with enhanced AI capabilities and updated designs. The Pixel 9 lineup includes the standard Pixel 9, two Pro models, and a foldable device. The new Pixel phones feature flat sides and an elongated camera module on the rear, departing from the curved edges of previous generations. Screen sizes range from 6.3 inches on the standard Pixel 9 to 6.8 inches on the Pixel 9 Pro XL.

All models are powered by Google's new Tensor G4 processor and come with increased RAM, with Pro models boasting 16GB. The devices run on Android 14 and will receive seven years of OS updates and security patches. Google has significantly expanded the AI capabilities of the new Pixels. An updated on-device Gemini Nano model can now analyze images and speech in addition to text. New features include automatic screenshot cataloging and retrieval, and an AI-powered illustration generator called Pixel Studio. Camera improvements are a key focus, with all models receiving upgraded ultrawide lenses and the Pro versions featuring a new 42-megapixel selfie camera with autofocus. Google has introduced "Magic Editor," allowing users to transform parts of an image using text prompts and generative AI.

The Pixel 9 Pro Fold, Google's second-generation foldable device, is thinner than its predecessor at 5.1mm when unfolded. It features a larger 8-inch inner display with increased brightness, reaching up to 2,700 nits in peak mode. Pricing for the new Pixel lineup starts at $799 for the standard Pixel 9, representing a $100 increase from last year's model. The Pixel 9 Pro and Pro XL are priced at $999 and $1,099 respectively, while the Pixel 9 Pro Fold will retail for $1,799. The devices will be released in stages, with the Pixel 9 and 9 Pro XL available from August 22, followed by the 9 Pro in September and the Pro Fold on September 4.

Dutch-based copyright enforcement group BREIN has taken down a large language dataset that was being offered for use in training AI models, the organization said on Tuesday. From a report: The dataset included information collected without permission from tens of thousands of books, news sites, and Dutch language subtitles harvested from "countless" films and TV series, BREIN said in a statement. Director Bastiaan van Ramshorst told Reuters it was not clear whether or how widely the dataset may already have been used by AI companies. "It's very difficult to know, but we are trying to be on time" to avoid future lawsuits, he said. He said the European Union's AI Act will require AI firms to disclose what datasets they have used to train their models.

AI PCs accounted for 14% of all PC shipped in the second quarter with Apple leading the way, research firm Canalys said on Tuesday, as added AI capabilities help reinvigorate demand. From a report: PC providers and chipmakers have pinned high hopes on devices that can perform AI tasks directly on the system, bypassing the cloud, as the industry slowly emerges from its worst slump in years. These devices typically feature neural processing units dedicated to performing AI tasks.

Apple commands about 60% of the AI PC market, the research firm said in the report, pointing to its Mac portfolio incorporating M-series chips with a neural engine. Within Microsoft's Windows, AI PC shipments grew 127% sequentially in the quarter. The tech giant debuted its "Copilot+" AI PCs in May, with Qualcomm's Snapdragon PC chips based on Arm Holdings' architecture.

About half of all job seekers are using AI tools to apply for roles, inundating employers and recruiters with low-quality applications in an already squeezed labour market. From a report: Candidates are turning increasingly to generative AI -- the type used in chatbot products such as ChatGPT and Gemini to produce conversational passages of text -- to assist them in writing their CVs, cover letters and completing assessments. Estimates from employers and recruiters who spoke to the Financial Times, as well as multiple published surveys, have suggested the figure is as high as 50 per cent of applicants.

A "barrage" of AI-powered applications had led to more than double the number of candidates per job while the "barrier to entry is lower," said Khyati Sundaram, chief executive of Applied, a recruitment platform. "We're definitely seeing higher volume and lower quality, which means it is harder to sift through," she added. "A candidate can copy and paste any application question into ChatGPT, and then can copy and paste that back into that application form."

In recent months, recruiters have received more applications for each job because labour markets on both sides of the Atlantic have weakened. Employers need to fill fewer vacancies, and more people are job-hunting after being made redundant. Longer-term trends, such as the rise of online job boards that make openings visible to a broader pool of potential candidates and make applying easy, have already boosted the number of applications. About 46 per cent of job hunters are using generative AI to search and apply for posts, according to a survey of 2,500 UK workers from HR start-up Beamery. In a separate poll of 5,000 global job seekers by creative platform Canva, 45 per cent had used generative AI to build or improve their CVs.

Google is rolling out a new feature for Google Wallet that uses AI to generate a digital version of IDs, tickets, and other passes. "Replacing the old 'Photo' option, Everything else lets you 'Scan a photo of any pass like an event ticket, gym membership, insurance card, and more' to create a digital version that appears in Google Wallet," writes 9to5Google's Abner Li. "The app explains how AI is leveraged to 'determine what kind of pass you're adding and to suggest the content of the pass.'" From the report: If you're adding something sensitive with health or government ID information, it will be classified as private and not get synced to other devices, while authentication is required before opening. However, you can change the private pass classification later. After taking a picture of the pass, Google will extract the information and let you edit common fields, as well as add your own. At this stage, you can change the pass type [...]. When finalized, it will appear below your carousel of credit/debit cards. Google will let you access the original "Pass photos" when viewing the digital copy.

In an interview with The Verge's Nilay Patel, Replika founder and CEO Eugenia Kuyda discusses the role AI will play in the future of human relationships. Replika is an AI-powered chatbot that offers personalized, empathetic conversations to users, serving as a virtual companion for emotional support, mental health, and social interaction. It allows users to engage in meaningful, human-like conversations, enhancing their well-being through AI-driven companionship. Here is an excerpt from the interview: Where have you landed with Replika now? Is it still sort of romantic? Is it mostly friendly? Have you gotten the user base to stop thinking of it as dating in that way?

It's mostly friendship and a long-term one-on-one connection, and that's been the case forever for Replika. That's what our users come for. That's how they find Replika. That's what they do there. They're looking for that connection. My belief is that there will be a lot of flavors of AI. People will have assistants, they will have agents that are helping them at work, and then, at the same time, there will be agents or AIs that are there for you outside of work. People want to spend quality time together, they want to talk to someone, they want to watch TV with someone, they want to play video games with someone, they want to go for walks with someone, and that's what Replika is for.

You've said "someone" several times now. Is that how you think of a Replika AI avatar -- as a person? Is it how users think of it? Is it meant to replace a person?

It's a virtual being, and I don't think it's meant to replace a person. We're very particular about that. For us, the most important thing is that Replika becomes a complement to your social interactions, not a substitute. The best way to think about it is just like you might a pet dog. That's a separate being, a separate type of relationship, but you don't think that your dog is replacing your human friends. It's just a completely different type of being, a virtual being. Or, at the same time, you can have a therapist, and you're not thinking that a therapist is replacing your human friends. In a way, Replika is just another type of relationship. It's not just like your human friends. It's not just like your therapist. It's something in between those things.

With an AI that kind of feels like a person and is meant to complement your friends, the boundaries of that relationship are still pretty fuzzy. In the culture, I don't think we quite understand them. You've been running Replika for a while. Where do you think those boundaries are with an AI companion?

I actually think, just like a therapist has agency to fire you, the dog has agency to run away or bite or shit all over your carpet. It's not really that you're getting this subservient, subordinate thing. I think, actually, we're all used to different types of relationships, and we understand these new types of relationships pretty easily. People don't have a lot of confusion that their therapist is not their friend. I mean, some people do project and so on, but at the same time, we understand that, yes, the therapist is there, and he or she is providing this service of listening and being empathetic. That's not because they love you or want to live with you. So we actually already have very different relationships in our lives. We have empathy for hire with therapists, for instance, and we don't think that's weird. AI friends are just another type of that -- a completely different type. People understand boundaries. At the end of the day, it's a work in progress, but I think people understand quickly like, 'Okay, well, that's an AI friend, so I can text or interact with it anytime I want.' But, for example, a real friend is not available 24/7. That boundary is very different. You know these things ahead of time, and that creates a different setup and a different boundary than, say, with your real friend. In the case of a therapist, you know a therapist will not hurt you. They're not meant to hurt you. Replika probably won't disappoint you or leave you. So there's also that. We already have relationships with certain rules that are different from just human friendships. The full transcript can be read here. You can also listen to the interview on the latest episode of Decoder with Nilay Patel.

Civil servants in the UK are locked in a row with the government over plans to use AI to answer taxpayer complaints. The Telegraph: Letters and emails to the Treasury are already being read by an AI tool which summarises the contents and suggests responses for civil servants. The government is now in talks to use AI across more departments, The Telegraph understands. A government spokesman said: "We make no apology for exploring innovations which improve public services. This tool has already been in use for four months and has increased productivity by 30pc, helping us save taxpayers' money by stopping the need to use expensive contractors during busy periods.

"Staff at the Treasury will continue to write all public correspondence and make decisions on cases. This tool, which was developed by data science experts in the department, helps civil servants respond to thousands more questions quickly and on time." But the civil serrvice trade body, the Public and Commercial Services (PCS) union, warned that the AI tool could be developed "on the cheap," leading to correspondence being misread. Fran Heathcote, PCS general secretary, said that while the union has no objection in principle to AI, that training the tool "properly takes a long time and considerable resources." Ms Heathcote added: "Further, in anticipation of AI working well, staff numbers are cut so you get the worst of all worlds: a poorly functioning AI system with too few humans left behind to pick up the broken pieces."

With its hardware event on Tuesday, Alphabet's Google is trying to outshine Apple's annual iPhone launch -- and is letting longtime executive Rick Osterloh take center stage. Bloomberg: Osterloh, the former president of Motorola who joined Google in 2016, will helm the first major product launch after the company this year unified under his leadership the teams developing hardware and the Android operating system. The reorganization expanded Osterloh's influence in the company and signaled that Google intends to compete in hardware for the long term.

In a sign of a more aggressive push into consumer devices, Google moved up its annual flagship Pixel smartphone launch to August from October, preempting the next Apple. iPhone debut and seizing attention during a typically quiet period for the industry. [...] By holding its hardware showcase a month ahead of the iPhone maker's largest annual event, Google is "frontrunning Apple and also making a statement that we are likely way ahead of what Apple will show for iPhone 16 at least," said Mandeep Singh, an analyst with Bloomberg Intelligence.

Google has at least a six-month head start on Apple, which has invested less in AI over the years than some of its Big Tech peers, he added. Google's strategy -- tying together the development of hardware, software and services -- carries echoes of Apple's successful approach to designing devices. Yet, as Osterloh seeks to capitalize on the opportunity presented by AI, he faces a perennial challenge for Google: bringing the fight to Apple without threatening key relationships with hardware giants such as Xiaomi that rely on the Android operating system.

Proximity to natural gas lines could become just as desirable for datacenter operators as high-speed fiber-optic networks as they scramble to satiate AI's ever growing thirst for power. From a report: Speaking to analysts during their respective earnings call this week, executives at Energy Transfer LP and Williams Companies, both of which operate pipelines across the US, revealed they were in talks with datacenter operators to supply them with large quantities of natural gas. "We are, in four different states, in discussions with multiple datacenters of different sizes. Some of them, or many of them, want to put generation on site ... So it's an enormous opportunity for us," Mackie McCrea, co-CEO of Energy Transfer LP, told Wall Street, according to a transcript.

Energy Transfer LP's pipelines currently span 15 states in the USA, serving 185 power plants. Looking at the opportunity afforded by datacenter hookups, McCrea estimated that power demand could increase by 30 to 40 gigawatts over the next six to eight years. "We believe we are extremely well positioned to benefit from the anticipated rise in natural gas needs," Energy Transfer LP co-CEO Tom Long added. Energy Transfer LP isn't the only pipeline operator eager to take advantage of skyrocketing datacenter power demands. Speaking to analysts earlier this week, Williams Companies CEO Alan Armstrong expressed optimism about the firm's ability to capitalize on this demand.

"Algorithmic price-fixing appears to be spreading to more and more industries," warns the Atlantic. "And existing laws may not be equipped to stop it."

They start with RealPage's rental-property software (pointing out that "a series of lawsuits says it's something else: an AI-enabled price-fixing conspiracy" and "The lawsuits also argue that RealPage pressures landlords to comply with its pricing suggestions.") But the most important point is that RealPage isn't the only company doing this: Its main competitor, Yardi, is involved in a similar lawsuit. One of RealPage's subsidiaries, a service called Rainmaker, faces multiple legal challenges for allegedly facilitating price-fixing in the hotel industry. (Yardi and Rainmaker deny wrongdoing.) Similar complaints have been brought against companies in industries as varied as health insurance, tire manufacturing, and meat processing. But winning these cases is proving difficult.
The article notes that "Agreeing to fix prices is punishable with up to 10 years in prison and a $100 million fine." But it also notes concerns that algorithms could produce price-fixing-like behavior that's "almost impossible to prosecute under existing antitrust laws. Price-fixing, in other words, has entered the algorithmic age, but the laws designed to prevent it have not kept up." Last week, San Francisco passed a first-of-its-kind ordinance banning "both the sale and use of software which combines non-public competitor data to set, recommend or advise on rents and occupancy levels."

Whether other jurisdictions follow suit remains to be seen.

In the meantime, more and more companies are figuring out ways to use algorithms to set prices. If these really do enable de facto price-fixing, and manage to escape legal scrutiny, the result could be a kind of pricing dystopia in which competition to create better products and lower prices would be replaced by coordination to keep prices high and profits flowing. That would mean permanently higher costs for consumers — like an inflation nightmare that never ends.

Mozilla's interim CEO Laura Chambers "says the company is reinvesting in Firefox after letting it languish in recent years," reports Fast Company, "hoping to reestablish the browser as independent alternative to the likes of Google's Chrome and Apple's Safari.

"But some of those investments, which also include forays into generative AI, may further upset the community that's been sticking with Firefox all these years..." Chambers acknowledges that Mozilla lost sight of Firefox in recent years as it chased opportunities outside the browser, such as VPN service and email masking. When she replaced Mitchell Baker as CEO in February, the company scaled back those other efforts and made Firefox a priority again. "Yes, Mozilla is refocusing on Firefox," she says. "Obviously, it's our core product, so it's an important piece of the business for us, but we think it's also really an important part of the internet."

Some of that focus involves adding features that have become table-stakes in other browsers. In June, Mozilla added vertical tab support in Firefox's experimental branch, echoing a feature that Microsoft's Edge browser helped popularize three years ago. It's also working on tab grouping features and an easier way to switch between user profiles. Mozilla is even revisiting the concept of web apps, in which users can install websites as freestanding desktop applications. Mozilla abandoned work on Progressive Web Apps in Firefox a few years ago to the dismay of many power users, but now it's talking with community members about a potential path forward.

"We haven't always prioritized those features as highly as we should have," Chambers says. "That's been a real shift that's been very felt in the community, that the things they're asking for . . . are really being prioritized and brought to life."
Firefox was criticized for testing a more private alternative to tracking cookies which could make summaries of aggregated data available to advertisers. (Though it was only tested on a few sites, "Privacy-Preserving Attribution" was enabled by default.) But EFF staff technologist Lena Cohen tells Fast Company that approach was "much more privacy-preserving" than Google's proposal for a "Privacy Sandbox." And according to the article, "Mozilla's system only measures the success rate of ads — it doesn't help companies target those ads in the first place — and it's less susceptible to abuse due to limits on how much data is stored and which parties are allowed to access it." In June, Mozilla also announced its acquisition of Anonym, a startup led by former Meta executives that has its own privacy-focused ad measurement system. While Mozilla has no plans to integrate Anonym's tech in Firefox, the move led to even more anxiety about the kind of company Mozilla was becoming. The tension around Firefox stems in part from Mozilla's precarious financial position, which is heavily dependent on royalty payments from Google. In 2022, nearly 86% of Mozilla's revenue came from Google, which paid $510 million to be Firefox's default search engine. Its attempts to diversify, through VPN service and other subscriptions, haven't gained much traction.

Chambers says that becoming less dependent on Google is "absolutely a priority," and acknowledges that building an ad-tech business is one way of doing that. Mozilla is hoping that emerging privacy regulations and wider adoption of anti-tracking tools in web browsers will increase demand for services like Anonym and for systems like Firefox's privacy-preserving ad measurements. Other revenue-generating ideas are forthcoming. Chambers says Mozilla plans to launch new products outside of Firefox under a "design sprint" model, aimed at quickly figuring out what works and what doesn't. It's also making forays into generative AI in Firefox, starting with a chatbot sidebar in the browser's experimental branch.
Chambers "says to expect a bigger marketing push for Firefox in the United States soon, echoing a 'Challenge the default' ad campaign that was successful in Germany last summer. Mozilla's nonprofit ownership structure, and the idea that it's not beholden to corporate interests, figures heavily into those plans."

A new article in ScienceAlert describes new research into the dangers of "heavily processed sources of digital nourishment" for generative AI: A new study by researchers from Rice University and Stanford University in the US offers evidence that when AI engines are trained on synthetic, machine-made input rather than text and images made by actual people, the quality of their output starts to suffer.

The researchers are calling this effect Model Autophagy Disorder (MAD). The AI effectively consumes itself, which means there are parallels for mad cow disease — a neurological disorder in cows that are fed the infected remains of other cattle. Without fresh, real-world data, content produced by AI declines in its level of quality, in its level of diversity, or both, the study shows. It's a warning about a future of AI slop from these models.

"Our theoretical and empirical analyses have enabled us to extrapolate what might happen as generative models become ubiquitous and train future models in self-consuming loops," says computer engineer Richard Baraniuk, from Rice University. "Some ramifications are clear: without enough fresh real data, future generative models are doomed to MADness."
The article notes that "faces began to look more and more like each other when fresh, human-generated training data wasn't involved. In tests using handwritten numbers, the numbers gradually became indecipherable.

"Where real data was used but in a fixed way without new data being added, the quality of the output was still degraded, merely taking a little longer to break down. It appears that freshness is crucial."

Thanks to long-time Slashdot reader schwit1 for sharing the news.

Cisco will cut thousands of jobs in a second round of layoffs this year as the U.S. networking equipment maker shifts focus to higher-growth areas, including cybersecurity and AI, Reuters reported Friday, citing sources. From the report: The number of people affected could be similar to or slightly higher than the 4,000 employees Cisco laid off in February, and will likely be announced as early as Wednesday with the company's fourth-quarter results, said the sources, who were not authorized to speak publicly.

The FCC has proposed new rules governing the use of AI-generated phone calls and texts. Part of the proposal centers on create a clear definition for AI-generated calls, with the rest focuses on consumer protection by making companies disclose when AI is being used in calls or texts. A report adds: "This provides consumers with an opportunity to identify and avoid those calls or texts that contain an enhanced risk of fraud and other scams," the FCC said. The agency is also looking ensure that legitimate uses of AI to assist people with disabilities to communicate remains protected.

OpenAI's latest AI model, GPT-4o, exhibits unusual behaviors, including voice cloning and random shouting, according to a new "red teaming" report. The model, which powers ChatGPT's Advanced Voice Mode alpha, is OpenAI's first trained on voice, text, and image data. In high-noise environments, GPT-4o occasionally mimics users' voices, a quirk OpenAI attributes to difficulties processing distorted speech. The company said it has implemented a "system-level mitigation" to address this issue. The report also reveals GPT-4o's tendency to generate inappropriate vocalizations and sound effects when prompted.

An anonymous reader quotes a report from The Guardian: Amazon's $4 billion investment into US artificial intelligence startup Anthropic is to be examined in the latest investigation into technology tie-ups by the UK's competition watchdog. The Competition and Markets Authority (CMA) said on Thursday that it was launching a preliminary investigation into the deal, before deciding whether to refer it for an in-depth review. The deal, announced in March, included a $4 billion investment in Anthropic from Amazon, and a commitment from Anthropic to use Amazon Web Services "as its primary cloud provider for mission critical workloads, including safety research and future foundation model development." The regulator said it was "considering whether it is or may be the case that Amazon's partnership with Anthropic has resulted in the creation of a relevant merger situation." "We are an independent company. Our strategic partnerships and investor relationships do not diminish our corporate governance independence or our freedom to partner with others," said an Anthropic spokesperson said in a statement. "Amazon does not have a seat on Anthropic's board, nor does it have any board observer rights. We intend to cooperate with the CMA and provide them with a comprehensive understanding of Amazon's investment and our commercial collaboration."

The founder of Cambridge-based Riverlane, Steve Brierley, predicts quantum computing will have its "Sputnik" breakthrough within years. "Quantum computing is not going to be just slightly better than the previous computer, it's going to be a huge step forward," he said. Phys.org reports: His company produces the world's first dedicated quantum decoder chip, which detects and corrects the errors currently holding the technology back. In a sign of confidence in Riverlane's work and the sector in general, the company announced on Tuesday that it had raised $75 million in Series C funding, typically the last round of venture capital financing prior to an initial public offering. "Over the next two to three years, we'll be able to get to systems that can support a million error-free operations," said Earl Campbell, vice president of quantum science at Riverlane. This is the threshold where a quantum computer should be able to perform certain tasks better than conventional computers, he added.

Quantum computers are "really good at simulating other quantum systems", explained Brierley, meaning they can simulate interactions between particles, atoms and molecules. This could open the door to revolutionary medicines and also promises huge efficiency improvements in how fertilizers are made, transforming an industry that today produces around two percent of global CO2 emissions. It also paves the way for much more efficient batteries, another crucial weapon in the fight against climate change. "I think most people are more familiar with exponential after COVID, so we know how quickly something that's exponential can spread," said Campbell, inside Riverlane's testing lab, a den of oscilloscopes and chipboards. [...]

While today's quantum computers can only perform around 1,000 operations before being overwhelmed by errors, the quality of the actual components has "got to the point where the physical qubits are good enough," said Brierley. "So this is a super exciting time. The challenge now is to scale up... and to add error correction into the systems," he added. Such progress, along with quantum computing's potential to crack all existing cryptography and create potent new materials, is spurring regulators into action. "There's definitely a scrambling to understand what's coming next in technology. It's really important that we learn the lessons from AI, to not be surprised by the technology and think early about what those implications are going to be," said Brierley. "I think there will ultimately be regulation around quantum computing, because it's such an important technology. And I think this is a technology where no government wants to come second."

Things aren't working out well for Humane, a heavily-funded startup that launched an eponymous AI device earlier this year. Despite significant funding from prominent Silicon Valley figures, the product has been grappling with negative reviews -- and now more pressing issues are emerging. An anonymous reader shares a report: Shortly after Humane released its $699 AI Pin in April, the returns started flowing in. Between May and August, more AI Pins were returned than purchased, according to internal sales data obtained by The Verge. By June, only around 8,000 units hadn't been returned, a source with direct knowledge of sales and return data told me. As of today, the number of units still in customer hands had fallen closer to 7,000, a source with direct knowledge said.

At launch, the AI Pin was met with overwhelmingly negative reviews. Our own David Pierce said it "just doesn't work," and Marques Brownlee called it "the worst product" he's ever reviewed. Now, Humane is attempting to stabilize its operations and maintain confidence among staff and potential acquirers. The New York Times reported in June that HP is considering purchasing the company, and The Information reported last week that Humane is negotiating with its current investors to raise debt, which could later be converted into equity.

AI is upending India's technology outsourcing business. The industry is pivoting to adapt, but the changes could cost a large number of coveted jobs. From a report: The country's big outsourcing companies are already using AI and have plans to integrate it throughout their businesses. That might not save the low-end operations that run call centers or do other basic tasks within the so-called business process outsourcing sector.Â

AI is threatening to disrupt most businesses around the world, not just India's $250 billion outsourcing industry. The outsourcing boom in India over the past few decades created the "getting Bangalore-d" phenomenon in the U.S., often used for Americans who lost their jobs to more affordable Indian talent. AI's impact could have big repercussions as the industry employs 5.4 million people, according to tech-industry body Nasscom, and contributes about 8% of the country's economy. More than 80% of companies in the S&P 500 outsource some operations to India, according to HSBC.