A new study published in Nature Human Behaviour reveals that ChatGPT diminishes the diversity of ideas generated during brainstorming sessions. Researchers from the University of Pennsylvania's Wharton School found [PDF] that while generative AI tools may enhance individual creativity, they simultaneously reduce the collective diversity of novel content.

The investigation responds to previous research that examined ChatGPT's impact on creativity. Their findings align with separate research published in Science Advances suggesting AI-generated content tends toward homogeneity. This phenomenon mirrors what researchers call the "fourth grade slump in creativity," referencing earlier studies on how structured approaches can limit innovative thinking.

An anonymous reader shares a report: Microsoft's recently announced job cuts fell hardest on the people who build the company's products, showing that even software developers are at risk in the age of artificial intelligence.

In Microsoft's home state of Washington, software engineering was by far the largest single job category to receive layoff notices, making up more than 40% of the roughly 2,000 positions cut, according to state documents reviewed by Bloomberg. Microsoft on Tuesday said it would cut about 6,000 workers across the company. The Washington state data represents about a third of the total.

Nvidia-backed CoreWeave plans to spend up to $23 billion this year on AI infrastructure and data center capacity, as it aims to meet the booming demand from clients. Reuters reports: The heavy spending plan weighed on its shares, which fell 5% after surging as much as 11% on better-than-expected revenue in its first results as a public company after debuting on the Nasdaq in March. The company's projected capital expenditure of between $3 billion and $3.5 billion for the second quarter was way above its revenue expectation of $1.06 billion to $1.1 billion. "While the revenue from Microsoft is likely secure for the next three years, CoreWeave represents overflow capacity for Microsoft, which may not need that capacity in the future," D.A. Davidson analyst Gil Luria said.

The company's revenue backlog was $25.9 billion as of March 31, with its five-year deal with OpenAI adding $11.2 billion to the pile up. As part of the deal signed in March, CoreWeave will provide AI infrastructure to OpenAI, while the ChatGPT maker will get a stake. CoreWeave expects annual revenue of $4.9 billion to $5.1 billion, above analysts' expectation of $4.61 billion, according to data compiled by LSEG. It reported revenue of $981.6 million for the first quarter, beating the estimate of $852.9 million.

Fintech startup Klarna is now recruiting humans after its AI customer service agents underperformed. The buy-now-pay-later company, which eliminated its marketing contracts in 2023 and customer service team in 2024, now plans an "Uber-type setup" with remote gig workers.

This marks a stark reversal from CEO Sebastian Siemiatkowski's 2024 claim that "AI can already do all of the jobs that we, as humans, do." Siemiatkowski told Bloomberg: "From a brand perspective, I just think it's so critical that you are clear to your customer that there will be always a human if you want." He added that "cost unfortunately seems to have been a too predominant evaluation factor" leading to "lower quality."

An anonymous reader quotes a report from Ars Technica: Google's DeepMind research division claims its newest AI agent marks a significant step toward using the technology to tackle big problems in math and science. The system, known as AlphaEvolve, is based on the company's Gemini large language models (LLMs), with the addition of an "evolutionary" approach that evaluates and improves algorithms across a range of use cases. AlphaEvolve is essentially an AI coding agent, but it goes deeper than a standard Gemini chatbot. When you talk to Gemini, there is always a risk of hallucination, where the AI makes up details due to the non-deterministic nature of the underlying technology. AlphaEvolve uses an interesting approach to increase its accuracy when handling complex algorithmic problems.

According to DeepMind, this AI uses an automatic evaluation system. When a researcher interacts with AlphaEvolve, they input a problem along with possible solutions and avenues to explore. The model generates multiple possible solutions, using the efficient Gemini Flash and the more detail-oriented Gemini Pro, and then each solution is analyzed by the evaluator. An evolutionary framework allows AlphaEvolve to focus on the best solution and improve upon it. Many of the company's past AI systems, for example, the protein-folding AlphaFold, were trained extensively on a single domain of knowledge. AlphaEvolve, however, is more dynamic. DeepMind says AlphaEvolve is a general-purpose AI that can aid research in any programming or algorithmic problem. And Google has already started to deploy it across its sprawling business with positive results. DeepMind's AlphaEvolve AI has optimized Google's Borg cluster scheduler, reducing global computing resource usage by 0.7% -- a significant cost saving at Google's scale. It also outperformed specialized AI like AlphaTensor by discovering a more efficient algorithm for multiplying complex-valued matrices. Additionally, AlphaEvolve proposed hardware-level optimizations for Google's next-gen Tensor chips.

The AI remains too complex for public release but that may change in the future as it gets integrated into smaller research tools.

Microsoft quietly announced earlier this week that it plans to shut down a long-standing tool supplying search engine startups and other software developers with a raw feed of Bing search results. From a report: The Bing Search APIs, or application programming interfaces, were once vital to many niche Google alternatives, but fell out of favor more recently as Microsoft hiked fees for the service and restricted its use.

The shutoff, which is scheduled to begin on August 11, still came as a surprise to several developers who spoke with WIRED. Customers learned of it on Monday via an email from Microsoft and a post on its website. They were directed to consider using "Grounding with Bing Search as part of Azure AI Agents," a Microsoft service that allows chatbots like ChatGPT to augment AI-generated responses with "real-time public web data." Some developers view the AI-centric alternative as an insufficient replacement. Larger customers like DuckDuckGo told Wired they won't be affected.

Palantir CEO Alex Karp criticized Europe's AI adoption while praising Saudi Arabia's engineering talent at Tuesday's Saudi-US Investment Forum in Riyadh. "It's like people have given up," Karp said of Europe, while commending Saudi engineers for their "meritocracy and patriotism" and "deep tradition in engineering excellence."

A California judge slammed a pair of law firms for the undisclosed use of AI after he received a supplemental brief with "numerous false, inaccurate, and misleading legal citations and quotations." From a report: In a ruling submitted last week, Judge Michael Wilner imposed $31,000 in sanctions against the law firms involved, saying "no reasonably competent attorney should out-source research and writing" to AI, as pointed out by law professors Eric Goldman and Blake Reid on Bluesky.

"I read their brief, was persuaded (or at least intrigued) by the authorities that they cited, and looked up the decisions to learn more about them -- only to find that they didn't exist," Judge Milner writes. "That's scary. It almost led to the scarier outcome (from my perspective) of including those bogus materials in a judicial order."

President Donald Trump's administration has taken a tougher stance on Chinese technology advances, warning companies around the world that using AI chips made by Huawei could trigger criminal penalties for violating US export controls. From a report: The commerce department issued guidance to clarify that Huawei's Ascend processors were subject to export controls because they almost certainly contained, or were made with, US technology.

Its Bureau of Industry and Security, which oversees export controls, said on Tuesday it was taking a more stringent approach to foreign AI chips, including "issuing guidance that using Huawei Ascend chips anywhere in the world violates US export controls." But people familiar with the matter stressed that the bureau had not issued a new rule, but was making it clear to companies that Huawei chips are likely to have violated a measure that requires hard-to-get licences to export US technology to the Chinese company.

Google's stalwart search button has a new neighbor: AI Mode. From a report: The artificial intelligence feature is being tested directly beneath the Google search bar beside a "Google Search" button, replacing the "I'm Feeling Lucky" widget. The new feature, though not widely available yet, is being tested in a location where Google rarely makes changes.

Amazon's Audible is expanding its audiobook offerings by giving select publishers access to its AI-powered production tools that will let them more easily convert books into audiobooks with over 100 customizable AI voices to choose from. It will also launch an AI translation tool in beta later this year to help publishers translate and localize audiobooks into multiple languages, with both automated and human-assisted options. The Verge reports: Audible says its new AI narration technology leverages Amazon's advanced AI capabilities and will be made available to interested publishing partners in the coming months in one of two ways. For publishers wanting to be hands-off, an end-to-end service managed by Audible handles the "entire audiobook production process" right up to publication, while a self-service option will give publishers access to the same tools so they can independently direct the entire production process.

With both options, publishers are able to "choose from a quickly growing and improving selection of more than 100 AI-generated voices across English, Spanish, French, and Italian with multiple accent and dialect options, and will be able to access voice upgrades for their titles as our technology evolves," according to Amazon. [...] Publishers will also be able to review translations themselves or opt for a human review through Audible with a professional linguist.

An anonymous reader quotes a report from 404 Media: Late last night, House Republicans introduced new language to the Budget Reconciliation bill that will immiserate the lives of millions of Americans by cutting their access to Medicaid, and making life much more difficult for millions more by making them pay higher fees when they seek medical care. While a lot of attention will be justifiably given to these cuts, the bill has also crammed in new language that attempts to entirely stop states from enacting any regulation against artificial intelligence.

"...no State or political subdivision thereof may enforce any law or regulation regulating artificial intelligence models, artificial intelligence systems, or automated decision systems during the 10 year period beginning on the date of the enactment of this Act," says the text of the bill introduced Sunday night by Congressman Brett Guthrie of Kentucky, Chairman of the House Committee on Energy and Commerce. The text of the bill will be considered by the House at the budget reconciliation markup on May 13.

That language of the bill, how it goes on to define AI and other "automated systems," and what it considers "regulation," is broad enough to cover relatively new generative AI tools and technology that has existed for much longer. In theory, that language will make it impossible to enforce many existing and proposed state laws that aim to protect people from and inform them about AI systems. [...] In theory none of these states will be able to enforce these laws if Republicans manage to pass the Budget Reconciliation bill with this current language.

The Department of Commerce officially rescinded the Biden administration's Artificial Intelligence Diffusion Rule on Tuesday, just days before its May 15 implementation date. The rule would have imposed first-ever export restrictions on U.S.-made AI chips to dozens of countries while tightening existing controls on China and Russia.

Instead of implementing blanket restrictions, the DOC signaled a shift toward direct country-by-country negotiations. The department released interim guidance reminding companies that using Huawei's Ascend AI chips anywhere violates U.S. export rules and warned about consequences of allowing U.S. chips to train AI models in China. Commerce Secretary for Industry and Security Jeffery Kessler criticized the previous administration's approach, calling it "ill-conceived and counterproductive."

Intel has certified Shell's lubricant-based method for cooling servers more efficiently within data centers used for AI. From a report: The announcement on Tuesday, which follows the chipmaker's two-year trial of the technology, offers a way to use less energy at AI facilities, which are booming and are expected to double their electricity demand globally by 2030, consuming as much power then as all of Japan today, according to the International Energy Agency.

So far, companies have largely used giant fans to reduce temperatures inside AI data centers, which generate more heat in order to run at a higher power. Increasingly, these fans consume electricity at a rate that rivals the computers themselves, something the facilities' operators would prefer to avoid, Intel Principal Engineer Samantha Yates said in an interview.

Amazon's decade-old acquisition of Annapurna Labs has emerged as a pivotal element in its AI strategy, with the once-secretive Israeli chip design startup now powering AWS infrastructure. The $350 million deal, struck in 2015 after initial talks between Annapurna co-founder Nafea Bshara and Amazon executive James Hamilton, has equipped the tech giant with custom silicon capabilities critical to its cloud computing dominance.

Annapurna's chips, particularly the Trainium processor for AI model training and Graviton for general-purpose computing, now form the foundation of Amazon's AI infrastructure. The company is deploying hundreds of thousands of Trainium chips in its Project Rainier supercomputer being delivered to AI startup Anthropic this year. Amazon CEO Andy Jassy, who led AWS when the acquisition occurred, described it as "one of the most important moments" in AWS history.

Over 400 prominent UK media and arts figures -- including Paul McCartney, Elton John, and Ian McKellen -- have urged the prime minister to support an amendment to the Data Bill that would require AI companies to disclose which copyrighted works they use for training. The Register reports: The UK government proposes to allow exceptions to copyright rules in the case of text and data mining needed for AI training, with an opt-out option for content producers. "Government amendments requiring an economic impact assessment and reports on the feasibility of an 'opt-out' copyright regime and transparency requirements do not meet the moment, but simply leave creators open to years of copyright theft," the letter says.

The group -- which also includes Kate Bush, Robbie Williams, Tom Stoppard, and Russell T Davies -- said the amendments tabled for the Lords debate would create a requirement for AI firms to tell copyright owners which individual works they have ingested. "Copyright law is not broken, but you can't enforce the law if you can't see the crime taking place. Transparency requirements would make the risk of infringement too great for AI firms to continue to break the law," the letter states. Baroness Kidron, who proposed the amendment, said: "How AI is developed and who it benefits are two of the most important questions of our time. The UK creative industries reflect our national stories, drive tourism, create wealth for the nation, and provide 2.4 million jobs across our four nations. They must not be sacrificed to the interests of a handful of US tech companies." Baroness Kidron added: "The UK is in a unique position to take its place as a global player in the international AI supply chain, but to grasp that opportunity requires the transparency provided for in my amendments, which are essential to create a vibrant licensing market."

The letter was also signed by a number of media organizations, including the Financial Times, the Daily Mail, and the National Union of Journalists.

ZDNet's Steven Vaughan-Nichols marks Reddit's 20 years of being "the front page of the internet," recalling its evolution from a scrappy startup into a cultural powerhouse that shaped online discourse, meme culture, and the way millions consume news and entertainment. Slashdot is also given a subtle nod in the opening line of the article. An anonymous reader shares an excerpt: In 2005, if you were into social networks focused on links, you probably used Digg or Slashdot. However, two guys, Steve Huffman and Alexis Ohanian, recent graduates from the University of Virginia, wanted to create a hub where users could find, share, and discuss the internet's most interesting content. Little did they know where this idea would take them. After all, their concept was nothing new. Still, after Paul Graham, co-founder of Y Combinator, the startup accelerator and seed capital firm, had shot down their first idea -- a mobile food-ordering app -- they pitched what would become Reddit to Graham, and he gave it his blessing. Drawing inspiration from sites like Delicious, a now-defunct social bookmarking service, and Slashdot, Huffman and Ohanian envisioned Reddit as a platform that would combine the best aspects of both: a place for sharing timely, ephemeral news and fostering vibrant community discussions of not just technology, but any topic users cared about. Their guiding mission was to build "the front page of the internet," a simple, user-driven site where anyone could submit content, and the community, not algorithms or editors, would decide what was most important through voting and discussion. They deliberately prioritized user participation and conversation over flashy features or heavy editorial control.

What set Reddit apart from its early rivals was its framework. Instead of one large all-in-one interface, the site borrowed the idea from pre-internet online networks, such as CompuServe, of smaller sub-networks devoted to a particular topic. These user-created communities, "subreddits," quickly set it apart from other social platforms. As Laurence Sangarde-Brown, co-founder of TechTree, wrote: "This design allows users to delve into focused discussions, ask questions, and exchange ideas on a scale unmatched by other platforms." That approach was not enough, though, to kick-start Reddit. The founders had to "fake it until they made it." They seeded the site with fake accounts to make it appear more active. Their efforts paid off, as real users soon flocked to the platform. Another crucial early change was when Reddit merged with Aaron Swartz's Infogami and introduced commenting. This move was vital for laying the groundwork for the site's interactive, community-driven experience. [...]

So, where does Reddit go from here? We'll see. Reddit's legacy is one of transformation: from a scrappy startup to a global hub for conversation, collaboration, and sometimes controversy. As it celebrates 20 years, Reddit remains a testament to how important online communities can be in a world increasingly filled with AI slop. Still, Huffman believes Reddit's true value is coming. In a recent Reddit post, he wrote: "Reddit works because it's human. It's one of the few places online where real people share real opinions. That authenticity is what gives Reddit its value. If we lose trust in that, we lose what makes RedditReddit. Our focus is, and always will be, on keeping Reddit a trusted place for human conversation." Huffman concluded: "The last 20 years have proven how powerful online communities can be — and as we look ahead, I'm even more excited for what the next 20 will bring."

An anonymous reader shares a report: After weeks of news about Google's antitrust travails, the tech giant will try to reset the narrative next week by highlighting advances it is making in artificial intelligence, cloud and Android technology at its annual I/O developer conference.

Ahead of I/O, Google has been demonstrating to employees and outside developers an array of different products, including an AI agent for software development. Known internally as a "software development lifecycle agent," it is intended to help software engineers navigate every stage of the software process, from responding to tasks to documenting code, according to three people who have seen demonstrations of the product or been told about it by Google employees. Google employees have described it as an always-on coworker that can help identify bugs to fix or flag security vulnerabilities, one of the people said, although it's not clear how close it is to being released.

Requesting concise answers from AI chatbots significantly increases their tendency to hallucinate, according to new research from Paris-based AI testing company Giskard. The study found that leading models -- including OpenAI's GPT-4o, Mistral Large, and Anthropic's Claude 3.7 Sonnet -- sacrifice factual accuracy when instructed to keep responses short.

"When forced to keep it short, models consistently choose brevity over accuracy," Giskard researchers noted, explaining that models lack sufficient "space" to acknowledge false premises and offer proper rebuttals. Even seemingly innocuous prompts like "be concise" can undermine a model's ability to debunk misinformation.

Google has launched the AI Futures Fund, a new initiative to invest in AI startups that are building with the latest tools from Google DeepMind. TechCrunch reports: The fund will back startups from seed to late stage and will offer varying degrees of support, including allowing founders to have early access to Google AI models from DeepMind, the ability to work with Google experts from DeepMind and Google Labs, and Google Cloud credits. Some startups will also have the opportunity to receive direct investment from Google.

"The AI Futures Fund doesn't follow a batch or cohort model," a Google spokesperson told TechCrunch. "Instead, we consider opportunities on a rolling basis -- there's no fixed application window or deadline. When we come across companies that align with the fund's thesis, we may choose to invest. We're not announcing a specific fund size at this time, and check sizes vary based on the company's stage and needs -- typically early to mid-stage, with flexibility for later-stage opportunities as well." Startups can apply here.