Gizmodo...Sam Altman’s Brief Ouster at OpenAI Is Getting the Movie TreatmentA movie titled "Artificial" is in development, chronicling Sam Altman's brief ousting and reinstatement as OpenAI's CEO. Luca Guadagnino is in talks to direct, with Andrew Garfield rumored to play Altman, Monica Barbaro as Mira Murati, and Yura Borisov as Ilya Sutskever. Simon Rich is writing the screenplay. The article critiques Hollywood's increasing focus on Big Tech dramas, questioning the motivations of Amazon MGM Studios, given Amazon's investment in OpenAI rival Anthropic. While acknowledging the inherent drama of the OpenAI situation, the author expresses concern that these portrayals often fail to capture the full extent of the greed, corruption, and disregard for ordinary people exhibited by tech leaders, citing "The Social Network" and "Silicon Valley" as rare exceptions.Read more →
Engadget...The head of Tesla's Optimus humanoid robot program leaves the companyMilan Kovac, the head of Tesla's Optimus humanoid robotics program, is leaving the company. Kovac cited the need to spend more time with family abroad as the sole reason for his departure, emphasizing his continued support for Elon Musk and Tesla. He joined Tesla in 2016 and began leading the Optimus group in 2022. Tesla announced the Optimus project in 2021 and debuted a prototype in 2022. Musk has stated Optimus could move at 5 mph, carry 45-pound loads, and potentially sell for $20,000, contributing to a $25 trillion market cap for Tesla. Recent demonstrations have showcased Optimus performing tasks like handling eggs and folding laundry. Production has been affected by China's export restrictions of rare earth magnets. Ashok Elluswamy, who leads Tesla's Autopilot group, will assume Kovac's responsibilities.Read more →
CNET...At WWDC, How Will Apple Address Its Lackluster First Year of Apple Intelligence?The article discusses Apple's upcoming WWDC 2025 and how the company will address its perceived shortcomings in the field of AI, particularly regarding Apple Intelligence and Siri. It highlights the contrast between the initial high expectations for Apple's AI advancements, following the AI showcases by companies like Google and OpenAI, and the subsequent slow rollout and acknowledged challenges in Siri's development. The author anticipates that Apple will likely acknowledge the situation with humor and then shift focus to new developments. The article also touches on the playful rivalry between tech giants like Apple and Google, referencing a jab made by Google at Apple's AI implementation. The WWDC 2025 keynote presentation is scheduled for June 9, and the CNET team will be reporting live from the event.Read more →
The Verge...Apple is on defense at WWDCThe article discusses the author's expectations and concerns regarding Apple's upcoming WWDC. The author notes a lack of anticipation compared to previous years, citing legal challenges related to app store commissions and disappointment with the rollout of Apple Intelligence. The author expresses a desire for Apple to acknowledge its missteps and show humility, rather than presenting a facade of confidence. The article highlights the absence of Apple executives from John Gruber's Talk Show, which the author interprets as a sign that Apple will likely continue to project confidence instead of addressing its issues. The author believes that Apple's usual approach may not be sufficient to overcome the current challenges, given the company's stock performance, legal pressures, and the perceived shortcomings of Apple Intelligence.Read more →
Gizmodo...Bad News: Google’s Veo 3 Can Make VR Slop, TooGoogle's Veo 3 AI video generator can now create 360-degree videos for VR viewing. Users can generate these videos by including "make it 360 degrees" in their Veo 3 prompts. A user, Henry Daubrez, discovered this trick and shared the method on X, detailing the need to adjust the video's metadata using ExifTool for proper VR playback. While the generated content may have some imperfections, like black bars, it functions surprisingly well, especially considering it's likely an unplanned feature. The author expresses cautious optimism about this development, noting the potential for low-quality content but also the opportunity to create novel VR experiences given the current limitations of the VR content catalog.Read more →
CNET...I've Been Using the Photoshop iPhone App for Months, and Now Android Users Can TooAdobe has released a beta version of its Photoshop mobile app for Android users, available for free download on the Google Play Store. This release aims to provide professional photographers and creators with on-the-go editing capabilities, mirroring many features of the original desktop version, including layers, masking, and the AI-powered generative fill. The mobile app includes tools optimized for mobile use, such as 'Tap select,' alongside familiar tools like the magic wand and object select. Users also gain access to Adobe's stock library. This move is intended to cater to the increasing amount of visual content created and edited on phones and to maintain competitiveness with services like Canva and emerging AI image generation technologies. Adobe plans to add more features in the future, leveraging its Firefly AI models across its creative suite.Read more →
The Verge...Popular AI apps get caught in the crosshairs of Anthropic and OpenAIThe article discusses the growing tension between major AI labs like Anthropic and OpenAI, and the AI applications that rely on their models. Anthropic has reportedly cut off Windsurf, a popular vibe coding tool, potentially due to OpenAI's interest in acquiring Windsurf. OpenAI has also introduced a feature in ChatGPT that directly competes with Granola, a meeting note-taking AI app. This raises concerns about the stability of building businesses on top of AI models, as the model providers may directly compete with their API customers. The article also touches on the debate around AI and job loss, with some CEOs downplaying fears of mass engineering job displacement, emphasizing the importance of engineers adopting AI tools. Additionally, the article includes personnel updates from various tech companies and a list of links for further reading.Read more →
TechCrunch...Meet the Finalists: VivaTech’s 5 Most Visionary Startups of 2025VivaTech is debuting its Innovation of the Year award to recognize startups that are reshaping the future through creativity, technological ingenuity, and potential for industry transformation. The five finalists for the award, who will pitch their solutions live at VivaTech 2025 on June 11th, are BeyondMath (AI-powered physics platform), Chipiron (low-cost MRI), Enerdrape (geothermal panels for urban spaces), Hua Tech International (microfluidic platform for cancer diagnostics), and Lumisync (photonic oscillator for data centers). A panel of expert judges will conduct a Q&A session after the pitches. The winner will be announced on June 12th at the VivaTech Global Awards Ceremony and will receive a free Startup Corner at VivaTech 2026 and a spot in the TechCrunch Startup Battlefield 200. The ceremony will also highlight other VivaTech Awards, including the Female Founder Challenge, the Africatech Awards, The Next Startupper Challenge, and the Tech For Change Award.Read more →
TechCrunch...TechCrunch Mobility: How Jony Ive’s LoveFrom helped Rivian and what Uber’s next-generation playbook looks likeThis TechCrunch Mobility article covers a range of topics in the transportation and technology sectors. It discusses Elon Musk's strained relationship with Donald Trump and its potential impact on Tesla, particularly concerning autonomous vehicle legislation and sales. The article also reveals Jony Ive's firm LoveFrom's involvement with Rivian. Furthermore, it highlights Joby Aviation's potential deal with Abdul Latif Jameel for electric aircraft distribution in Saudi Arabia, and funding rounds for startups like Obvio, Portless, and Toma. Uber's strategic shift towards autonomy, indicated by executive changes and CEO statements, is analyzed. Other news includes Tesla's robotaxi trademark efforts, Redwood Materials declining a Department of Energy loan, Walmart and Wing expanding drone delivery, and Trevor Milton's legal battles.Read more →
CNET...Google's AI Mode Now Creates Interactive Stock Charts For YouGoogle has launched an AI Mode experiment within Google Labs that allows users to create interactive charts for stocks and mutual funds using its Gemini AI engine. Users can input queries, such as comparing stock performances or requesting top-performing mutual funds, and the AI will generate an interactive graph with explanations. While the feature aims to provide accessible financial data visualization, experts caution users to verify the AI's output due to potential accuracy issues in quantitative reasoning. The AI provides links to data sources, but users are advised to double-check the information and avoid seeking investment advice from the AI, as it is not a credentialed financial advisor and AI-powered investment tools have shown underperformance compared to benchmarks like the S&P 500. The feature, currently limited to stocks and mutual funds, is expected to expand to other topics in the future.Read more →
CNET...My WWDC 2025 Predictions: iOS 26, Gaming, Health and More Expected MondayApple's WWDC 2025, commencing June 9th, faces pressure to innovate following the launches of Vision Pro and Apple Intelligence in previous years. While past events introduced entries into AR/VR and generative AI, both have experienced slow adoption and criticism. This year's focus may shift towards operating system redesigns and incremental improvements across Apple's product lines. Key expectations include a potential cross-OS redesign, possibly named after the year (e.g., iOS 26), and a new "glass" look for the user interface, potentially mirroring VisionOS. Other anticipated updates involve AI enhancements for the Apple Watch, including message summaries and health insights, alongside potential battery life improvements via AI. Gaming may also receive attention with a possible new game hub app and VR gaming expansions for Vision Pro. Additional rumors suggest updates to the Messages app, iPadOS, and the iPhone Camera app. A new HomePod with a screen is also speculated. The event will clarify Apple's AI strategy and potentially reveal new products.Read more →
TechCrunch...Why investing in growth-stage AI startups is getting riskier and more complicatedThe article discusses the current state of investing in AI startups, highlighting both the excitement and the risks involved. Jill Chase from CapitalG notes the trend of very young companies reaching significant revenue and valuation milestones rapidly. While this fast growth is appealing, it also presents challenges for investors due to the lack of established infrastructure and the rapidly evolving AI landscape. Chase emphasizes the importance of assessing a founder's ability to adapt and anticipate future changes in the AI field. She cites AI coding startup Cursor as an example of a company that capitalized on a timely use case but will need to innovate to stay ahead as AI technology advances, particularly with the anticipated emergence of AI software engineers.Read more →
TechCrunch...Anthropic appoints a national security expert to its governing trustAnthropic has appointed Richard Fontaine, a national security expert, to its long-term benefit trust, a governance mechanism designed to prioritize safety over profit and influence the company's board of directors. This move follows Anthropic's recent announcement of new AI models for U.S. national security applications and reflects the company's increasing engagement with the defense sector, including a partnership with Palantir and AWS to sell AI to defense customers. CEO Dario Amodei stated that Fontaine's expertise will strengthen the trust's ability to navigate complex decisions related to AI and security. Fontaine's background includes advising the late Sen. John McCain and leading the Center for A New American Security. The appointment is part of Anthropic's broader strategy to strengthen its executive ranks, as seen with the addition of Reed Hastings to its board in May. Other AI companies, including OpenAI, Meta, Google, and Cohere, are also pursuing opportunities in the defense sector.Read more →
TechCrunch...AI startups continue fueling San Francisco’s office recoveryEarly-stage AI startups, many backed by Y Combinator, are revitalizing San Francisco's Northern Waterfront by leasing significant office space. Five AI companies recently leased 23,900 square feet at Waterfront Plaza, contributing to a citywide trend where AI firms are expanding and occupying a total of 5 million square feet. CBRE projects that AI companies could occupy 21 million square feet within five years, potentially halving San Francisco's vacancy rate and creating numerous jobs. This expansion could significantly boost the city's downtown vibrancy. Waterfront Plaza, a 442,000-square-foot complex, previously housed WeWork.Read more →
TechCrunch...Tesla’s Optimus robot VP is reportedly leaving the companyMilan Kovac, the head of Tesla's Optimus humanoid robot program, is leaving the company. This departure occurs as Tesla CEO Elon Musk projects the deployment of thousands of Optimus robots in Tesla factories by the end of the year, with plans to rapidly scale production to millions of units annually. Kovac, who previously led engineering teams for Tesla's Autopilot software, began leading the Optimus project in 2022 and became a vice president overseeing the program in late 2024. Ashok Elluswamy, Tesla's vice president of AI software, will assume leadership of the Optimus project.Read more →
Gizmodo...Live Updates From Apple WWDC 2025 🔴Apple is set to unveil significant software updates for its entire ecosystem at WWDC 2025 on June 9th. The updates, expected to jump to version 26 across iOS, iPadOS, macOS, watchOS, visionOS, and tvOS, will feature a unified visual design inspired by the Vision Pro. The iPhone is anticipated to receive its first major interface overhaul since iOS 7 in 2013. A key focus will be Apple's approach to artificial intelligence, dubbed "Apple Intelligence," and whether the company will address delays in its next-generation Siri voice assistant. Gizmodo's tech team will provide live coverage from Apple Park.Read more →
The Verge...Google Gemini can now handle scheduled tasks like an assistantGoogle is introducing "scheduled actions" for Gemini AI Pro and AI Ultra subscribers, enabling the AI assistant to perform tasks at specific times. Users can schedule recurring tasks like daily calendar summaries or weekly blog post ideas, as well as one-off tasks such as summarizing events. The feature, accessible through the Gemini app's settings, aims to make Gemini a more proactive assistant, similar to the recurring actions feature offered by OpenAI's ChatGPT.Read more →
TechCrunch...Figure AI CEO skips live demo, sidesteps BMW deal questions on stage at tech conferenceFigure AI's CEO, Brett Adcock, addressed skepticism surrounding the company's progress and its relationship with BMW at the Bloomberg Tech conference. While not providing specific contractual details, Adcock emphasized the technical benefits of using robots in factory settings. He confirmed a contract with a second, unnamed customer, reportedly UPS, for initial deployment. Figure AI aims to manufacture and deploy approximately 100,000 robots within four years. The company is currently seeking to raise $1.5 billion at a $39.5 billion valuation, a significant increase from its previous valuation. Figure AI has also been issuing cease-and-desist letters to secondary market brokers marketing its shares.Read more →
The Verge...The Ray-Ban Meta smart glasses are on sale for their best price to dateThe article promotes Ray-Ban Meta Smart Glasses as a practical Father's Day gift. These smart glasses, available at a discounted price of $239.20 at Amazon, Best Buy, and Target, offer features beyond basic eye protection. They include a built-in 12MP camera for hands-free photos and videos, direct livestreaming capabilities, and high-quality audio for music and calls. The latest model incorporates AI features, enabling voice-activated photo and video capture, messaging, information retrieval, translations, object recognition, and QR code scanning.Read more →
TechCrunch...Building Your AI Engine: How OpenAI Works with StartupsHao Sang from OpenAI's Startups Team discussed at TechCrunch Sessions: AI how startups can leverage OpenAI's resources to gain a competitive advantage in the AI field. The discussion covered technical guidance, access to advanced models, and how startup feedback influences OpenAI's product development roadmap.Read more →
TechCrunch...AI startups take to the stage to pitch their companies to a panel of judgesAt TechCrunch Sessions: AI, Fluix AI, Clicka, and Narada AI participated in the "So You Think You Can Pitch?" event. Each company was given four minutes to present their products, platforms, solutions, and leadership to judges from Initiate Ventures, Felicis, and Recursive Ventures.Read more →
TechCrunch...Building More Scalable GenAI Applications for Startups and DevelopersOracle presented at TechCrunch Sessions: AI, detailing how MySQL HeatWave facilitates the development of AI-based solutions. HeatWave's features, including its built-in vector store, in-database LLMs, machine learning capabilities, and parallel processing architecture, enable the creation of GenAI applications with real-time data integration, personalization, and retrieval-augmented generation. This simplifies the development process, accelerates time-to-market, and allows developers to concentrate on innovation rather than infrastructure management, resulting in more powerful and scalable GenAI solutions.Read more →
Gizmodo...Apple Has a Huge Siri Problem That WWDC 2025 Probably Won’t FixThe upcoming Apple Worldwide Developers Conference (WWDC) 2025 is anticipated to be pivotal, primarily due to concerns surrounding Apple's progress in artificial intelligence, particularly with its Siri voice assistant. The article highlights delays and reported struggles in developing Apple Intelligence, with its LLM-powered Siri facing significant setbacks. Despite these challenges, the author suggests that the pressure on Apple is currently more about perception than practical impact, as consumer interest in AI features on phones remains lukewarm. The conference may serve as an opportunity for Apple to address these AI shortcomings, potentially by shifting focus to other expected updates, such as a redesigned iOS inspired by visionOS, a gaming hub, and improved battery management. While not a make-or-break moment, WWDC 2025 presents a chance for Apple to demonstrate its vision for AI and avoid falling further behind competitors in the rapidly evolving AI landscape.Read more →
TechCrunch...EleutherAI releases massive AI training dataset of licensed and open domain textEleutherAI has released The Common Pile v0.1, an 8-terabyte dataset of licensed and open-domain text intended for training AI models. This project, a collaboration with AI startups and academic institutions, aims to provide a transparent alternative to datasets that rely on scraped, potentially copyrighted material. EleutherAI argues that lawsuits surrounding AI training data have reduced transparency in the field, hindering research. The organization claims that models trained on The Common Pile v0.1, named Comma v0.1-1T and Comma v0.1-2T, perform comparably to those trained on unlicensed data, rivaling models like Meta’s Llama AI on benchmarks for coding, image understanding, and math. The dataset, available on Hugging Face and GitHub, includes public domain books and transcribed audio content. EleutherAI's executive director, Stella Biderman, asserts that unlicensed text is not essential for achieving high model performance and that the quality of models trained on openly licensed content will improve as the amount of accessible data grows. This release also addresses EleutherAI's past use of copyrighted material in its earlier dataset, The Pile, and signals a commitment to more frequent releases of open datasets.Read more →
TechCrunch...Disruption playbook: How to beat AI incumbents at their own gameA panel discussion at TechCrunch Sessions: AI, featuring Odyssey co-founder Oliver Cameron, Linear COO Cristina Cordova, and NEA partner Ann Bordetsky, explored strategies for new AI companies to succeed against well-funded, established competitors. The conversation, moderated by TechCrunch's AI editor Kyle Wiggers, focused on how these companies can break through in a sector dominated by incumbents.Read more →
CNET...WatchOS 12 Is Coming! Here's What We're Manifesting for Apple WatchThe article previews potential updates to WatchOS 12, expected to be unveiled at Apple's Worldwide Developers Conference (WWDC) on June 9. It highlights desired features from CNET editors and Apple Watch users, focusing on improved battery life, UI customization (including a watch face store and redesigned app list), enhanced gesture controls, and the integration of Apple Intelligence features like Genmoji and smarter notifications. The article also discusses the need for smarter Siri, customizable workout options, and more proactive fitness and sleep coaching, drawing comparisons to competitors like Google and Oura. It mentions rumors of a UI redesign inspired by visionOS and a secret project ("Project Mulberry") aimed at revamping the Health app with smart coaching. The piece emphasizes the desire for the Apple Watch to provide more actionable insights from health data and personalized recommendations, while also addressing concerns about subscription costs.Read more →
Gizmodo...OpenAI Appeals ‘Sweeping, Unprecedented Order’ Requiring It Maintain All ChatGPT LogsA federal judge ordered OpenAI to preserve all ChatGPT data due to an ongoing copyright lawsuit filed by The New York Times, who allege their articles were used without permission to train OpenAI's language models. OpenAI is appealing the decision, arguing that the order violates user privacy and that the Times' claims are without merit, citing "fair use". OpenAI COO Brad Lightcap stated the data retention demand abandons privacy norms, while CEO Sam Altman called for "AI privilege" akin to attorney-client privilege. The court order has sparked user concerns about privacy, with some expressing worry over the potential exposure of sensitive information shared with ChatGPT. The article also raises broader questions about the ethics of AI training data and the need for consent when using copyrighted material.Read more →
TechCrunch...X names Polymarket as its official prediction market partnerX, formerly Twitter, has partnered with Polymarket, a cryptocurrency-based prediction market platform, to integrate prediction data with X's real-time insights. The partnership aims to provide users with data-driven insights and recommendations by combining Polymarket's prediction probabilities with Grok's analysis and X's real-time data. The initial product will integrate Polymarket predictions with X data, offering live insights and real-time annotations of market moves. Polymarket's CEO, Shayne Coplan, emphasized the partnership's potential to enhance users' ability to understand breaking news and make informed decisions. In 2024, Polymarket saw over $8 billion in predictions across various categories. This announcement follows Polymarket's previous partnerships with Perplexity and Substack.Read more →
TechCrunch...Startups Weekly: It’s buying seasonThis week's startup news highlights a flurry of acquisitions and significant funding rounds. Cellebrite acquired Corellium, IBM acquired Seek AI, Collibra acquired Raito, AMD acquired Brium, and Snowflake plans to buy Crunchy Data. Scale AI hired the team behind Pesto AI, which is shutting down. Airtime laid off employees, and KiranaPro confirmed a data breach. Deel and Rippling continue their legal battle. In venture capital, Anduril raised $2.5 billion, Anysphere raised $900 million, Neuralink secured $650 million, Bolttech completed a $147 million Series C, Speedata raised $44 million, Nomupay locked in a $40 million Series C, and Obvio raised a $22 million Series A. North American AI startups attracted significantly more venture capital than Europe and Asia combined. Elad Gil is now investing in traditional businesses that AI can help reinvent.Read more →
CNET...ChatGPT Built a Budget for Me ASAP, but It Has Several LimitationsThe article discusses using ChatGPT for budget creation, highlighting its potential benefits and drawbacks. It details a hands-on experiment where the author used ChatGPT to build a budget for a hypothetical person, outlining the steps involved, including providing income, expenses, and savings goals. The article points out that while ChatGPT can be helpful for suggesting budget categories, setting spending limits, and providing savings goals, it requires precise prompts and manual data entry. It lacks the automated transaction tracking and categorization features found in dedicated budgeting apps. The author concludes that ChatGPT can be a free and basic tool for initial budget creation and customization, but it's not suitable for ongoing budget maintenance or complex financial advice. The article recommends double-checking ChatGPT's suggestions and being cautious about sharing sensitive financial information due to potential privacy and security risks. Budgeting apps are presented as a more effective alternative for comprehensive budget management.Read more →
TechCrunch...Humans provide neccessary ‘checks and balance’ for AI, says Lattice CEOSarah Franklin, CEO of Lattice, a $3 billion employee performance software company, emphasizes the importance of balance in AI adoption, advocating for prioritizing people and customers over solely pursuing cost savings through workforce replacement. Speaking at SXSW London and in an interview with TechCrunch, Franklin highlighted the need for transparency, accountability, and responsibility in AI implementation, asserting that AI should serve humans, not the other way around. Lattice has developed an AI HR agent to assist employees, but Franklin insists on human oversight to maintain checks and balances. She believes companies that prioritize human connection will be the most successful in the AI era.Read more →
TechCrunch...The best ways to build on top of foundation models, with DeepMind, Twelve Labs, and AmazonThe news article discusses the challenge and opportunity for startups to build sustainable businesses amidst the rapidly evolving landscape of AI models. Experts from DeepMind, Twelve Labs, and Amazon shared their strategies at the TechCrunch Sessions: AI event, emphasizing that the evolving AI infrastructure provides founders with more tools and resources than ever before to build what they want.Read more →
Engadget...Tesla is reportedly blocking the city of Austin from releasing Robotaxi recordsTesla is attempting to prevent the city of Austin from releasing public records related to its robotaxi operations, which are planned to expand this month. This follows a similar request to prevent the National Highway Traffic Safety Administration from releasing crash data involving Tesla's self-driving features. Reuters requested two years of communications between Tesla and Austin officials after Elon Musk announced Austin would be a robotaxi testing ground. Tesla's attorney argued the information is confidential and proprietary. An Austin official stated the city doesn't take a position on the confidentiality of the materials. Tesla recently announced it has been testing driverless Model Ys in Austin ahead of schedule. Waymo has been operating autonomous taxis in Austin since 2023 and has partnered with Uber since March. The Texas Attorney General's office will decide next week whether Austin must release the records.Read more →
TechCrunch...OpenAI’s marketing head takes leave to undergo breast cancer treatmentKate Rouch, OpenAI's head of marketing, is taking a three-month leave of absence to undergo treatment for invasive breast cancer. Gary Briggs, formerly Meta's CMO, will serve as interim head of marketing. Rouch announced her diagnosis via LinkedIn, stating she was diagnosed shortly after starting her role at OpenAI and has been undergoing chemotherapy while leading the marketing team. She emphasizes her positive prognosis and encourages women to prioritize their health and undergo routine exams, citing statistics about the prevalence of breast cancer. Rouch joined OpenAI in December and previously held marketing leadership positions at Coinbase and Meta.Read more →
TechCrunch...A deep dive on AI safety and ethics with Databricks and ElevenLabsElevenLabs' Head of AI Safety, Artemis Seaford, and Databricks co-founder, Ion Stoica, discussed the ethical challenges of AI with TechCrunch AI editor Kyle Wiggers. The discussion focused on the increasing importance of AI safety and ethics due to the growing accessibility and affordability of AI tools. They explored methods to prevent misuse, such as deepfakes, and addressed the complexities of defining ethical boundaries in AI development and deployment.Read more →
TechCrunch...iOS 19: All the rumored changes Apple could be bringing to its new operating systemAhead of Apple's WWDC 2025, rumors are circulating about the upcoming iOS 19, potentially to be renamed iOS 26. Key anticipated changes include a major design overhaul inspired by the Vision Pro, featuring translucent panels and circular app icons across Apple's ecosystem. Specific apps like Phone, Camera, and Safari are expected to receive significant updates. A dedicated gaming app integrating Apple Arcade and App Store games is rumored, along with a virtual AI health coach offering personalized advice. Battery management may be improved with AI, and reverse wireless charging is being tested. The Messages app could gain AI-powered translation and polling features, plus background image customization. The Preview app from macOS might come to iOS, and Genmoji could allow emoji combinations. Notably absent from the rumors are the previously highlighted Siri capabilities, which have been delayed. The focus is expected to be on design changes and improvements to existing Apple Intelligence features.Read more →
TechCrunch...WWDC 2025: What to expect from this year’s conferenceApple's WWDC 2025 is anticipated to showcase significant software updates and new technologies, with a strong emphasis on building upon its AI capabilities. The event is expected to feature a major redesign of iOS, potentially named iOS 26, with visual elements inspired by visionOS, including a transparent interface and circular app icons. This redesign aims for a more seamless experience across Apple's ecosystem, including CarPlay. Other key announcements may include a new naming system for Apple's operating systems, aligning them with the year of release. While Apple may not unveil groundbreaking AI advancements, updates to existing AI tools and potential integration of third-party AI chatbots are expected. The Health app may incorporate AI for personalized health suggestions, and Messages could gain AI-powered translation and poll features. A dedicated gaming app is also rumored to replace Game Center, offering access to Apple Arcade and social gaming features. AirPods are expected to receive new features like live language translation, head gesture controls, and improved audio management. The Apple Pencil may gain a virtual calligraphy feature for Arabic script. While new hardware announcements are unlikely, support for spatial controllers for the Vision Pro and potential unveiling of "HomeOS" for a smart home tablet are possibilities.Read more →
TechCrunch...AMD acqui-hires the employees behind Untether AIAMD has acquired the team behind Untether AI, a Toronto-based startup specializing in AI inference chips. Untether AI, founded in 2018, claimed its chips were faster and more energy-efficient than competitors and had raised over $150 million in venture capital. The acquisition follows AMD's recent purchase of AI software optimization platform Brium. The terms of the Untether AI deal were not disclosed. TechCrunch has contacted AMD for further details.Read more →
The Verge...OpenAI is storing deleted ChatGPT conversations as part of its NYT lawsuitOpenAI is appealing a court order related to The New York Times' copyright lawsuit that compels the company to indefinitely store deleted ChatGPT conversations. OpenAI COO Brad Lightcap stated the company believes the order is an overreach that compromises user privacy. The order requires OpenAI to preserve all output log data, even if users request deletion or privacy laws mandate it. This impacts free, Pro, Plus, and Team ChatGPT users, but not Enterprise or Edu customers, or those with zero data retention agreements. OpenAI assures that the data will not be public and access will be limited to a small legal and security team. The New York Times sued OpenAI and Microsoft in 2023, alleging copyright infringement for using millions of articles to train AI models. OpenAI CEO Sam Altman has publicly stated that the company will fight any demand that compromises user privacy.Read more →
Engadget...Is it wrong the iPhone's AI battery management is the only WWDC rumor I'm excited about?The article expresses the author's lack of excitement for Apple's upcoming Worldwide Developer Conference (WWDC). The author anticipates announcements of features that are either already available or not personally useful. Specifically, the author is skeptical of a Vision Pro-inspired UI overhaul for iPhone, iPad, and Mac, questioning the need for consistency across platforms and worrying about the potential loss of device-specific functionality. The author also expresses indifference towards rumored improvements to Apple Intelligence and Siri, as well as generative AI features. The only anticipated feature of interest is an "AI-powered" battery management mode for iOS, which could address the iPhone's historically poor battery life. The author suggests Apple should focus on improving existing code and performance, similar to past "Snow Leopard" and "High Sierra" updates, rather than prioritizing new, potentially unnecessary features.Read more →
CNET...Apple's Siri Could Be More Like ChatGPT. But Is That What You Want?The article discusses the growing public apprehension towards AI, particularly concerning realistic AI-generated content like Google's Veo 3, which blurs the line between reality and fiction. It also highlights concerns raised by Anthropic's CEO about AI potentially eliminating numerous entry-level white-collar jobs, prompting calls for government intervention. The author questions the current consumer demand for AI and Apple's approach to AI development, specifically regarding Siri. The article references reports of leadership changes within the Siri team and differing opinions on whether Siri should evolve into a chatbot or focus on improving device control. The author anticipates criticism of Siri and Apple's AI after the upcoming WWDC presentation, but hopes this feedback will guide the future development of the assistant. The author encourages viewers to subscribe to the YouTube channel for more Apple news and analysis.Read more →
TechCrunch...Peloton explores placing its equipment in gyms, launches marketplace for used gearPeloton CEO Peter Stern announced the company is pursuing strategies to broaden its customer reach. These strategies include distributing Peloton equipment to commercial gyms through its Precor subsidiary, integrating Peloton workouts into Precor equipment, and launching a peer-to-peer marketplace called Repowered for used Peloton equipment to streamline second-hand sales. Additionally, Peloton aims to leverage AI to enhance personalization by creating individualized workout programs for its members.Read more →
TechCrunch...Peloton explores placing its equipment in gyms, launching marketplace for used gearPeloton CEO Peter Stern announced the company is pursuing strategies to broaden its customer reach. These include distributing Peloton equipment to commercial gyms through its Precor subsidiary, integrating Peloton workouts into Precor equipment, and launching "Repowered," a peer-to-peer marketplace for used Peloton equipment to provide a more streamlined and premium experience compared to existing secondhand sales platforms. Stern also highlighted the company's intention to leverage AI to enhance personalization by creating individualized workout programs for its members.Read more →
The Verge...The Vergecast’s WWDC 2025 previewThis Vergecast episode previews Apple's Worldwide Developer Conference (WWDC), discussing potential operating system redesigns, AI integrations, and developer frustrations. The hosts, Nilay Patel, David Pierce, and Joanna Stern, explore rumors surrounding multitasking, AirPods functionality, and AI's impact on battery life. The episode also covers the Nintendo Switch 2, Meta's smart glasses, and a theory about Jony Ive and OpenAI. A lightning round includes fediverse news.Read more →
CNET...He Got Us Talking to Alexa. Now He Wants to Kill Off AI HallucinationsThe article discusses William Tunstall-Pedoe's journey in AI, from creating the voice assistant Evi, which was later acquired by Amazon and evolved into Alexa, to his current focus on combating AI hallucinations. Tunstall-Pedoe emphasizes the risks associated with AI-generated inaccuracies, particularly in critical sectors like medicine and finance. He founded Unlikely AI to address this issue by grounding AI outputs in logic and logging the decision-making processes of algorithms. While acknowledging the inherent limitations of current AI technology, Tunstall-Pedoe remains optimistic about AI's potential to positively impact society, especially as companies like his work towards achieving greater accuracy.Read more →
Engadget...OpenAI appeals court order forcing it to preserve all ChatGPT dataThe New York Times and OpenAI are in conflict over data preservation related to a copyright infringement case. The New York Times seeks access to ChatGPT data logs to track potential copyright violations, arguing that OpenAI's technology has induced users to plagiarize its content. OpenAI is appealing a court order to retain and segregate ChatGPT output data, citing user privacy concerns. CEO Sam Altman stated the order compromises user privacy and sets a bad precedent. OpenAI argues that preserving data, even from users who requested deletion, conflicts with its privacy commitments. A federal judge initially sided with The New York Times, allowing the copyright case to proceed. The broader issue of AI companies using copyrighted material for training purposes remains unresolved, with tech companies claiming 'fair use' and content creators arguing that AI infringes on their copyrights and livelihoods. The court order does not affect ChatGPT Enterprise or ChatGPT Edu customers.Read more →
Engadget...Engadget Podcast: Answering your Nintendo Switch 2 questionsThis article summarizes a discussion about the launch of the Switch 2, including initial impressions of the console and the game Mario Kart World. The discussion also covers the launch experience. Additionally, the article previews Apple's WWDC 2025, anticipating a new naming scheme and design updates for Apple operating systems, along with further details on Apple's AI initiatives under the name Apple Intelligence. The hosts and guests share their first thoughts on the Switch 2's display and games, and discuss expectations for WWDC 2025 following the previous year's AI Siri announcement. The article also touches on electric vehicles.Read more →
TechCrunch...Cursor’s Anysphere nabs $9.9B valuation, soars past $500M ARRAnysphere, the company behind the AI coding assistant Cursor, has raised $900 million in a funding round led by Thrive Capital, valuing the company at $9.9 billion. Other participating investors include Andreessen Horowitz, Accel, and DST Global. This is Anysphere's third fundraise in under a year, following a $100 million round at a $2.5 billion pre-money valuation late last year. Anysphere's annualized revenue is reportedly doubling every two months and has surpassed $500 million, a 60% increase from mid-April. Cursor offers tiered pricing for individual developers and is now offering enterprise licenses. Anysphere previously declined acquisition offers from OpenAI and other potential buyers.Read more →
Gizmodo...A Cosmic Tech Mystery Results in Literal Lost Souls in This Sci-Fi Short StoryIn a near-future world heavily reliant on metaverse technology, a global event occurs where fifteen million people briefly disappear, experiencing a shared consciousness event lasting twenty-one seconds. This event is dubbed the "Twenty-One Second God." The story follows Corwin Sukarto, one of those affected, as he is investigated by Metaverse and WestHem Alliance. They discover that Metaverse had secretly increased bandwidth in their Hogan bridges, allowing for the unexpected hive mind to form. Sukarto is studied to understand the event's implications, but becomes disillusioned. As new, intentionally created hive minds emerge, Sukarto, feeling isolated and manipulated, decides to join one, the Moksha Mind, despite warnings from Colonel Moore, who suspects Sukarto may be a sleeper agent. The story explores themes of consciousness, grief, technological dependence, and the potential dangers of interconnected minds.Read more →
CNET...Galaxy Lockscreens Can Use AI to Show You in Outfits You Might Want to BuySamsung and Glance AI have partnered to introduce an AI-powered shopping experience for Galaxy users. This feature, rolling out this week to select devices and expected to be available to all users within 30 days, allows users to see AI-generated images of themselves wearing different outfits directly on their lock screens. The optional app, downloadable from the Samsung Galaxy Store, enables users to upload or take a selfie, then uses AI to display how they would look in various styles, with direct links to purchase the clothing. Samsung will promote the new AI technology across its devices, educating users on how to install and activate it. Glance AI leverages partnerships with over 400 clothing brands to provide recommendations and fulfill orders, aiming to predict and suggest outfits that customers will find appealing and purchase. The technology builds upon existing 'try-on' AI, like Google's, but goes further by proactively suggesting outfits based on inferred preferences.Read more →
CNET...I Switched to ChatGPT's Voice Mode. Here Are 7 Reasons Why It's Better Than TypingThe article discusses ChatGPT's Voice Mode, a hands-free feature that allows users to interact with the AI model through voice. The author expresses initial skepticism but finds the feature surprisingly effective and natural, highlighting its ability to understand conversational nuances. Voice Mode is compared to similar offerings from Google's Gemini Live, Anthropic's Claude, and Perplexity. The article details how Voice Mode works, differentiating between the free Standard Voice (using GPT-4o) and the paid Advanced Voice (using natively multimodal models for real-time conversation). The author lists seven reasons to use Voice Mode, including its conversational nature, hands-free convenience, language learning capabilities, ability to analyze real-world visuals (Advanced Voice only), accessibility for people with disabilities, faster brainstorming, and instant audio summaries of documents. The author concludes that Voice Mode offers a quicker, more natural way to use ChatGPT, making it feel more like a conversation with an expert.Read more →
TechCrunch...Perplexity received 780 million queries last month, CEO saysPerplexity CEO Aravind Srinivas announced at Bloomberg's Tech Summit that the AI search engine processed 780 million queries in May, demonstrating a month-over-month growth rate exceeding 20%. Srinivas projects that Perplexity could reach a billion queries per week if this growth continues. He highlighted the company's growth from 3,000 queries on its first day in 2022 to 30 million queries per day currently. Srinivas attributed potential future growth to Perplexity's development of a new browser called Comet. He envisions Comet as a "cognitive operating system" that will shift AI's role from providing answers to performing actions for users. Comet aims to integrate AI seamlessly into browsing sessions, potentially attracting users dissatisfied with existing browsers like Chrome. Srinivas noted that Comet would enable Perplexity to track user activity beyond its own app, facilitating the sale of premium ads, similar to Google's business model. The launch date for Comet is not yet known, but Srinivas indicated it would be in the coming weeks.Read more →
CNET...How a Proposed Moratorium on State AI Rules Could Affect YouA provision in the federal spending bill currently before the US Senate proposes a 10-year moratorium on state laws and regulations concerning artificial intelligence. Proponents argue this would foster AI industry growth and competitiveness against countries like China by preventing a fragmented regulatory landscape. Critics contend it would limit states' ability to protect consumers and address AI-related risks, favoring federal control. AI developers and some lawmakers support federal action to avoid a patchwork of state rules that could hinder innovation. However, some AI companies advocate for transparency standards instead of a moratorium. Several states have already begun regulating AI, with Colorado and California enacting consumer protections and AI-related laws. The proposed moratorium faces opposition from state attorneys general and some members of Congress, who argue it leaves Americans unprotected from potential AI harms. The Senate will now debate the measure, with potential for its removal due to budgetary rules. The outcome will significantly impact the future of AI regulation and consumer protection.Read more →
Gizmodo...Reddit Sues Anthropic, Accusing It of Illegal Data UseReddit has filed a lawsuit against Anthropic, an AI company, alleging that Anthropic used Reddit users' posts to train its AI chatbot Claude without authorization and in violation of Reddit's user agreement. The lawsuit accuses Anthropic of unjust enrichment and breaching the platform's user agreement, claiming its bots visited Reddit's website over 100,000 times since 2024. Reddit argues that Anthropic's actions contradict its public image as an ethical AI company. Anthropic has responded that they disagree with Reddit's claims and will defend themselves. This lawsuit highlights the growing conflict over the use of copyrighted content for AI training, with platforms and content creators increasingly resorting to legal action. Reddit has previously established paid agreements with companies like Google and OpenAI for AI training data access, suggesting a potential industry model of compensating content providers. The article suggests that larger AI companies will be able to afford these data access deals, potentially disadvantaging smaller firms.Read more →
TechCrunch...Amazon launches new R&D group focused on agentic AI and roboticsAmazon is launching a new agentic AI group within its consumer product division. The group will be based at Lab126, Amazon's hardware R&D division responsible for products like the Amazon Echo and Kindle. The goal is to develop an agentic AI framework for use in Amazon's robotics efforts, specifically to enhance the capabilities of warehouse robots.Read more →
Engadget...The Oversight Board says Meta isn't doing enough to fight celeb deepfake scamsThe Meta Oversight Board has criticized Meta for its inadequate handling of AI deepfake scams on its platforms, specifically those using celebrity likenesses. The board's decision stemmed from a case involving a manipulated video of retired soccer player Ronaldo Nazário promoting an online casino game. Despite numerous reports, Meta initially failed to remove the ad, highlighting flaws in its content moderation. The Oversight Board found that Meta prioritizes avoiding over-enforcement over actively preventing scams, and that reviewers lack the necessary training and authority to identify AI-manipulated content. The board recommended updating internal guidelines and empowering reviewers. Meta responded by stating that many of the board's claims are inaccurate and cited its ongoing efforts to combat scams using facial recognition technology. However, the article notes that similar scams persist, including deepfakes of Elon Musk and Jamie Lee Curtis. The article also references reports from The Wall Street Journal and regulators in Britain and Australia, which indicate that Meta's platforms are a significant source of fraud. The article suggests Meta is hesitant to add friction to its ad-buying process or ban advertisers with a history of scams.Read more →
TechCrunch...Anthropic co-founder on cutting access to Windsurf: ‘It would be odd for us to sell Claude to OpenAI’Anthropic's Co-founder and Chief Science Officer, Jared Kaplan, stated that the company limited Windsurf's direct access to its Claude AI models due to rumors of OpenAI acquiring the AI coding assistant. Kaplan emphasized Anthropic's focus on sustainable, long-term partnerships and expressed reservations about providing Claude to OpenAI. The decision follows reports that OpenAI is acquiring Windsurf for $3 billion. Windsurf expressed disappointment, citing potential instability for users. Anthropic cited computing constraints as another reason for the decision, prioritizing resources for lasting partnerships. Kaplan mentioned plans to increase model availability through a new computing cluster with Amazon. Anthropic is collaborating with other AI coding tool developers like Cursor and is focusing on developing its own agentic coding products like Claude Code, moving away from the chatbot paradigm.Read more →
Gizmodo...Amazon Looks to Replace Pesky Paid Workers With Humanoid Bots That Don’t Need to PeeAmazon is reportedly testing humanoid robots for last-mile package delivery, aiming to automate this final step in fulfillment. According to an anonymous source cited by The Information, these robots, potentially similar to Agility Robotics' Digit, are being trained to deliver packages from Rivian-made vans to customers' doorsteps. The testing will occur at an indoor park in San Francisco. This initiative is coupled with internal efforts to develop AI software to operate and guide the robots. The move is seen as a continuation of Amazon's efforts to automate its workforce, potentially reducing its reliance on human labor. Concerns are raised about the robots' ability to handle real-world complexities and the potential impact on human workers.Read more →
The Verge...Anthropic launches new Claude service for military and intelligence useAnthropic has launched Claude Gov, an AI model tailored for U.S. defense and intelligence agencies. These models have relaxed safety restrictions for government use and are designed to analyze classified information effectively. According to Anthropic, these models are already in use by agencies at the highest level of U.S. national security, with access limited to government entities handling classified data. Claude Gov is designed to address specific government needs such as threat assessment and intelligence analysis, while undergoing rigorous safety testing. These models differ from consumer-facing Claude by being more permissive when handling classified information and demonstrating enhanced understanding of defense and intelligence documents, as well as relevant languages and dialects. The launch of Claude Gov follows OpenAI's release of ChatGPT Gov and reflects a broader trend of AI companies seeking partnerships with government agencies. Anthropic is part of Palantir’s FedStart program. Other companies, like Scale AI, are also expanding their government contracts, including a deal with the Department of Defense and a five-year agreement with Qatar.Read more →
TechCrunch...X tests highlighting posts that are liked by users with opposing viewsX (formerly Twitter) is launching an experiment to identify posts that appeal to individuals with differing viewpoints. This initiative utilizes the platform's Community Notes feature, which is typically used to flag misleading or inaccurate content. A select group of Community Notes contributors will be prompted to evaluate and provide feedback on specific posts based on their initial Like activity. The feedback gathered will be used to develop an open-source algorithm designed to recognize posts that resonate across diverse perspectives. Contributors will share what they appreciate and dislike about the post, selecting options such as "I learned something interesting" or "I don’t agree with it." X aims to uncover ideas and insights that bridge different viewpoints, potentially motivating users to share such ideas and fostering progress. Community Notes was launched in 2022 after Elon Musk's acquisition of the company and has since been adopted by Meta's platforms.Read more →
Engadget...It turns out you can train AI models without copyrighted materialA collaborative research effort involving multiple universities and nonprofits has successfully trained a large language model (LLM) using only public domain and openly licensed material, challenging the assertion by AI companies that training on copyrighted material is essential. The resulting model, while comparable in performance to Meta's Llama 2-7B from 2023, was significantly more difficult to create due to the need for manual data annotation and legal verification. This study serves as a counterpoint to claims made by companies like OpenAI and Anthropic, who have argued that current AI models would be impossible or unlikely to exist without the use of copyrighted data. While the study is unlikely to alter the course of AI development driven by major companies, it may influence legal and regulatory discussions surrounding AI training data.Read more →
Gizmodo...Jeff Bezos’s Washinton Post Plans to Add Random Opinion Writers Edited by AIThe Washington Post plans to expand its opinion section by incorporating new writers guided by artificial intelligence, a program called Ripple. This initiative will utilize an AI coach named Ember to assist new writers, including those from Substack and non-professionals, by providing writing prompts, shaping content, and tracking story strength. While the articles will be reviewed by a human editor before publication, the move has raised concerns due to past issues with AI-generated content, such as those experienced by CNET. The Washington Post has faced financial challenges, including layoffs and subscriber losses, and this new AI-driven approach is viewed skeptically by some, especially given resistance to AI in journalism at other publications.Read more →
Engadget...X tests... centrismX (formerly Twitter) is experimenting with a new feature designed to identify and promote posts that are widely agreeable to its users. The "experimental pilot" program involves surveying a small group of users about their opinions on posts, with responses ranging from positive to negative. X aims to use this feedback to develop an open-source algorithm that can effectively identify posts liked by people from different perspectives. This initiative, similar in concept to Community Notes, seeks to boost broadly appealing content rather than focusing on fact-checking. X hopes the feature will uncover ideas that bridge perspectives, increase awareness of resonant content, and motivate users to share such ideas.Read more →
TechCrunch...Google says its updated Gemini 2.5 Pro AI model is better at codingGoogle has announced an updated preview of its Gemini 2.5 Pro model, which it claims excels at coding and demonstrates top-tier performance in math, science, knowledge, and reasoning. The updated model builds on the previous Gemini 2.5 Pro release from a month ago and incorporates user feedback to improve style and structure, resulting in more creative and better-formatted responses. Gemini 2.5 Pro is available starting today on AI Studio, Vertex AI, and the Gemini app, with general availability expected in a couple of weeks.Read more →
The Verge...X changes policy to restrict AI training on postsX (formerly Twitter) has updated its developer agreement to prohibit the use of content sourced from the platform or its API for training artificial intelligence models, including both foundation and frontier models. This restriction prevents developers from utilizing X's data to fine-tune or train AI systems.Read more →
The Verge...Why Microsoft has created its own print magazineMicrosoft has launched a print magazine called "Signal" to reach business leaders directly, bypassing the digital world. The magazine's first issue focuses on AI and includes deep dives on Microsoft technologies, interviews with executives, and lighter content. Microsoft aims for a more thoughtful and lasting form of communication compared to ephemeral online content. The company printed 1,500 copies for distribution at business events and will gauge interest before increasing production. "Signal" is inspired by Costco's magazine and emphasizes how businesses are using Microsoft's technology, with AI being a prominent theme. Microsoft's Copilot AI assistant contributed to some parts of the magazine's creation. The company collaborated with the Slow Journalism Company for writing and editorial assistance. The author of the article enjoyed reading "Signal", particularly the soccer piece and the 50-year anniversary stories, and is curious to see if other tech companies will follow suit with print publications.Read more →
TechCrunch...Anduril raises $2.5B at $30.5B valuation led by Founders FundAnduril, a defense technology startup specializing in autonomous weapons and software, has secured $2.5 billion in a Series G funding round led by Founders Fund with a $1 billion investment, the largest check Founders Fund has ever written. This round doubles Anduril's valuation to $30.5 billion and was reportedly oversubscribed by a factor of eight. The funding follows a year in which Anduril doubled its revenue to approximately $1 billion. A significant factor contributing to Anduril's success is the U.S. Army's contract for developing AR/VR headsets for soldiers, initially awarded to Microsoft but reassigned to Anduril in February with a $22 billion budget. This contract has led to a partnership between Anduril and Meta, prompting Anduril's founder, Palmer Luckey, to publicly reconcile with his former employer. Trae Stephens, Anduril's executive chair and co-founder, and also a partner at Founder's Fund, announced the funding in an interview.Read more →
Engadget...Foreign propagandists continue using ChatGPT in influence campaignsOpenAI has disrupted four Chinese covert influence operations that utilized ChatGPT to generate social media posts and comments across platforms like TikTok, Facebook, Reddit, and X. These operations aimed to stir up misleading political discourse on topics ranging from US politics to a Taiwanese video game. The posts often supported and decried opposing viewpoints to create division. OpenAI also disrupted similar operations originating from Russia, Iran, and North Korea. These malicious applications included debugging code, generating scripts, and creating content for use in phishing campaigns. In 2023, cybersecurity firm Mandiant found AI-generated content used in politically motivated online influence campaigns dating back to 2019. OpenAI also disrupted an Iranian operation using ChatGPT to create political articles about US elections for fake news sites and generate opposing viewpoints on X and Instagram. Despite the use of AI, these operations did not necessarily achieve greater engagement.Read more →
Engadget...Tesla is trying to stop certain self-driving crash data becoming publicTesla is attempting to block the release of certain crash data requested by the National Highway Transportation Safety Administration (NHTSA), arguing the information is confidential and its disclosure would benefit competitors by revealing details about the effectiveness of Tesla's self-driving technology and the number of crashes linked to specific systems. The Washington Post has filed a lawsuit against the NHTSA seeking more detailed crash information involving driver-assistance systems like Autopilot and Full Self-Driving (FSD), claiming the agency withholds critical details about the technologies in use and the circumstances of the crashes. Tesla argues that details about driver behavior, road conditions, and specific versions of driver-assistance tech should remain private. However, The Post contends that information about hardware and software versions is not confidential, as drivers can access it through their dashboards. Tesla counters that while individual owners may know certain information, it is shared with Tesla with the expectation of privacy. Tesla also leverages NHTSA loopholes to redact most data related to crashes involving its vehicles, maintaining that its "self-driving" tech requires active driver supervision. Separately, Tesla plans to launch its Robotaxi service in Austin, Texas, this month.Read more →
TechCrunch...How to watch Apple’s WWDC 2025 keynoteApple's 2025 Worldwide Developers Conference (WWDC) begins Monday, with the keynote scheduled for 10 a.m. PT / 1 p.m. ET. The event is expected to feature announcements regarding updates to Apple's operating systems, including iOS, iPadOS, macOS, watchOS, tvOS, and visionOS. Additional anticipated announcements include a dedicated gaming app and new AirPods features. The keynote and subsequent sessions will be available on Apple's event page and YouTube channel. Reports suggest a potential design overhaul for Apple's operating systems to create a more cohesive user experience across devices, as well as possible changes to the naming scheme. For developers, Apple is reportedly planning an AI-powered coding tool in partnership with Anthropic for Xcode. TechCrunch will provide comprehensive coverage of WWDC.Read more →
TechCrunch...WWDC 2025: What to expect from this year’s conferenceApple's WWDC 2025 is anticipated to showcase significant software updates and new technologies, particularly focusing on iOS. The upcoming iOS version is rumored to feature a major design overhaul, potentially drawing inspiration from visionOS with a transparent interface and circular app icons. This redesign may extend across Apple's ecosystem, including CarPlay. A change in the iOS naming system is also expected, potentially shifting to a year-based naming convention (e.g., iOS 26). While AI updates are expected, they may be less prominent than in the previous year, with potential announcements including AI-powered features in the Health app, Messages, battery management, and the Shortcuts app. A dedicated gaming app is also rumored to replace Game Center, offering access to Apple Arcade and social gaming features. macOS may receive the new gaming app and benefit from new AirPods features, while Apple TV may get a visual overhaul and UI changes. AirPods are expected to gain live language translation, new head gesture controls, and improved features like auto-pause and camera control. Hardware announcements are unlikely, although support for spatial controllers for the Vision Pro is possible. A previously rumored smart home tablet is not expected to be announced.Read more →
CNET...Amazon's Building Humanoid Robots to Speed Deliveries to YouAmazon is reportedly developing human-like autonomous robots for package delivery, according to an internal report. The company is building an obstacle park in San Francisco to test this technology. This initiative marks Amazon's first explicit venture into humanoid robotics for package delivery, potentially replacing human workers. These robots will utilize AI technology similar to that used in other areas of Amazon's business. In addition to delivery robots, Amazon is also developing humanoid robots for warehouse tasks, using AI to enable robots to understand and respond to natural language commands. Amazon has numerous robotics-related job openings across multiple countries.Read more →
Gizmodo...Things Humans Still Do Better Than AI: Understanding FlowersA recent study published in Nature Human Behaviour reveals that AI tools, including ChatGPT, struggle to understand familiar physical concepts as well as humans do. The study, led by Qihui Xu at Ohio State University, compared the conceptual understanding of humans and AI models (GPT-3.5, GPT-4, PaLM, and Gemini) across thousands of words, using psycholinguistic ratings like the Glasgow and Lancaster Norms. The research found that while AI performs well with abstract concepts, it falters with concepts tied to sensory experiences, such as flowers. This limitation stems from AI's lack of sensorimotor experiences. AI models trained on both text and images showed improved understanding of visual concepts, suggesting potential future improvements through sensorimotor data and robotics. The findings highlight implications for AI-human interactions, emphasizing the richness of human experience beyond language.Read more →
The Verge...Runway CEO Cris Valenzuela wants Hollywood to embrace AI videoThis is an interview with Cris Valenzuela, CEO and co-founder of Runway, an AI video generation platform. Valenzuela discusses the evolution of AI video generation since founding the company in 2018, highlighting improvements in model quality, increased awareness of AI's value, and growing competition. He emphasizes Runway's role as a tool for filmmakers, not a disruptive force, and mentions partnerships with Lionsgate and AMC Networks. The conversation covers the ongoing AI vs. art debate, including Runway's involvement in a class-action lawsuit regarding the use of copyrighted material in AI training. Valenzuela defends Runway's approach, arguing that AI creates new content rather than replicating existing works. The interview also explores Runway's structure as a research lab, the tension between product development and research, and the company's revenue model, which includes subscriptions and partnerships with educational institutions and studios. Valenzuela addresses concerns about the potential displacement of jobs in the creative industries, advocating for upskilling and retraining. He envisions a future where AI democratizes filmmaking, enabling more diverse voices to tell their stories.Read more →
TechCrunch...Anthropic unveils custom AI models for U.S. national security customersAnthropic has launched a new suite of AI models, called "Claude Gov," specifically designed for U.S. national security clients. These models are tailored to meet the operational demands of government agencies, offering enhanced capabilities in strategic planning, operational support, and intelligence analysis within classified environments. According to Anthropic, Claude Gov models demonstrate improved handling of classified material, greater understanding of intelligence and defense documents, enhanced proficiency in critical languages and dialects, and better interpretation of cybersecurity data. The company has partnered with Palantir and AWS to market its AI solutions to defense customers. This move aligns with a broader trend among leading AI labs, including OpenAI, Meta, Google, and Cohere, to pursue defense contracts and establish closer ties with the U.S. Defense Department.Read more →
Gizmodo...OnePlus Pad 3 Shoves a Massive Battery Into Its Super-Skinny Tablet FrameThe OnePlus Pad 3 tablet, launching July 8th, features a large 12,140mAh battery with 80W fast charging, a 13.2-inch 3.4K 144Hz display, a Snapdragon 8 Elite chip, 256GB of storage, and 12GB of RAM. It also includes eight speakers with spatial audio. Running Oxygen OS 15 based on Android 15, the tablet offers AI features like AI Summarize, AI Writer, AI Translations, and AI Speak, along with Google Gemini support. Productivity features include enhanced multitasking and O+ Connect for file sharing with PCs, and a remote control feature for Macs. Pre-orders begin June 5th, with the tablet priced at $699.99, the Smart Keyboard at $199.99, and the Folio Case at $49.99. The Stylo 2 pen is compatible but sold separately.Read more →
TechCrunch...The founder of DeviantArt is making a $22,000 display for digital artAngelo Sotira, founder of DeviantArt, has launched a new venture called Layer, a high-end screen designed specifically for displaying digital art. Unlike consumer products like Samsung's The Frame, Layer is a premium product priced at $22,000, targeting a niche market of digital art enthusiasts. Layer focuses on showcasing code-based and generative AI art, requiring significant computing power to render the art in high resolution without compression. The company partners with artists, paying them royalties based on the display time of their work. Layer aims to provide a seamless, long-term art display experience, avoiding the need for constant manual updates. The startup has secured $5.7 million in funding from venture capital firms and angel investors.Read more →
TechCrunch...X changes its terms to bar training of AI models using its contentSocial network X has updated its developer agreement to prohibit third parties from using its API and content to train large language models. This change, implemented on Wednesday, restricts the use of X's data for fine-tuning or training foundation or frontier models. The update follows X's acquisition by Elon Musk's AI company, xAI, in March. Previously, X had updated its privacy policy in 2023 to use public data for AI model training and later allowed third parties to train their models. Other companies like Reddit and The Browser Company have also implemented similar safeguards against AI crawlers.Read more →
TechCrunch...Toma’s AI voice agents have taken off at car dealerships – and attracted funding from a16zAI voice startup Toma, co-founded by Monik Pamecha and Anthony Krivonos, initially focused on banking and healthcare but pivoted to the automotive industry after dealerships expressed a need for assistance with overwhelming phone call volumes. The founders conducted extensive on-site research, visiting dealerships across Oklahoma and Mississippi to understand their operations. This hands-on approach, which included immersing themselves in dealership culture, helped them refine their AI voice agent to address specific needs such as scheduling service appointments and handling parts orders. Toma's AI is now used in over 100 dealerships, with an onboarding process that involves training the AI on customer calls to account for variations in dealership operations. The company operates on a subscription model, with increased capabilities leading to higher subscription fees. Toma recently secured $17 million in funding led by a16z, with participation from Y Combinator, Scale Angels fund, and Yossi Levi (Car Dealership Guy). The funding will support the company's growth, including expanding its sales team.Read more →
CNET...Motorola's 2025 Edge Launches With 2-Day Battery, 68-Watt Charging for $550The Motorola 2025 Edge smartphone, launching Thursday, features several software enhancements previously seen on Motorola's new Razr phones. Priced at $550, it boasts Moto AI, a triple 50-megapixel camera system, and Motorola's Photo Enhancement Engine. The phone has a 6.7-inch super HD POLED display with a 120Hz refresh rate and curved edges. Camera features include a 50-megapixel main lens, a 50-megapixel ultrawide lens, a 10-megapixel telephoto lens with up to 30x digital zoom, and a 50-megapixel front-facing camera. The device comes in Pantone Deep Forest color. AI capabilities include noise reduction, detail amplification, dynamic range improvement, and integration with Google Photos' AI features. It also features a dedicated AI key for Moto AI functions like Pay Attention and Next Move, along with Google AI features like Gemini Live and Circle to Search. The phone has IP68 and IP69 dust and water resistance ratings, MIL-STD-810H certification for durability, and Corning Gorilla Glass 7i for drop and scratch resistance. It is powered by a MediaTek Dimensity 7400 chipset, 8GB of RAM, and 256GB of storage, with a 5,200-mAh battery that Motorola claims lasts up to two days. It supports 68-watt TurboPower charging and 15-watt wireless charging and ships with Android 15. The Motorola 2025 Edge will be available unlocked in the US at Best Buy, Amazon, and Motorola's website, and through select carriers in the coming months. It will also be available in Canada via Motorola's website.Read more →
TechCrunch...Collibra acquires data access startup RaitoCollibra, a data governance platform, has acquired Raito, a data access startup, to enhance its offerings in response to the growing need for data access management in the age of AI. The acquisition aims to bolster Collibra's existing Collibra Protect product by incorporating Raito's cloud-native technology to automate and scale data access controls. Collibra CEO Felix Van de Maele emphasized that managing data access has become increasingly challenging for enterprises due to the rise of AI and workflow automation. The acquisition follows a trend of data companies being acquired to address data governance gaps as companies adapt to AI innovation, with recent examples including Salesforce's acquisition of Informatica, and acquisitions by Alation and ServiceNow. Van de Maele noted that the increasing fragmentation of data stacks has created governance challenges, making the integration of Raito into Collibra's unified governance platform a strategic move. Raito, founded in 2021 by former Collibra employees, had previously raised $4 million in venture funding, including from Collibra itself.Read more →
TechCrunch...TechCrunch Sessions: AI launches in Berkeley today — here’s what you’ll miss if you’re not hereTechCrunch Sessions: AI is being held today at UC Berkeley's Zellerbach Hall. The event features a lineup of AI experts and focuses on insights, innovation, and networking within the AI community. The agenda includes main stage panels, fireside chats, and breakout sessions covering topics such as the future of AI-human interaction, AI funding for startups, AI ethics and safety, building on foundational models, and leveraging AI in various industries. Speakers from Anthropic, Khosla Ventures, OpenAI, DeepMind, Amazon, Toyota, and Google Cloud, among others, are participating. The event also offers opportunities for networking through an Expo Hall, networking lounge, and the Braindate app, as well as after-hours side events. Attendees can learn about deploying AI in regulated industries, navigating AI policy, and building scalable GenAI applications. The event concludes at 4:30 p.m. PT.Read more →
The Verge...The cursed world of AI kiss and hug appsThe article discusses the rise of AI video apps that allow users to create deepfakes for entertainment purposes, such as kissing celebrities or altering appearances. These apps, often found through ads on platforms like TikTok, offer a range of templates, from innocuous ones like changing hair color to more suggestive or disturbing options, such as enhancing breast size or creating suggestive dances. The generated videos often have noticeable AI flaws, like awkward kissing or distorted features, and racial biases have been observed. These apps typically operate on a microtransaction or subscription model. While some uses of these apps may seem harmless, such as creating heartwarming videos or exploring gender identity, the author explores the ethical implications and potential for misuse, particularly in creating non-consensual or exploitative content. The author also shares a personal experience of using these apps to create deepfakes of deceased parents, highlighting the complex emotions and potential comfort that such technology can provide, despite its imperfections. The article also touches on the moderation policies of major tech companies like Google, Apple, and Meta regarding these apps and their advertisements, noting inconsistencies and challenges in regulating this type of content.Read more →
CNET...I Turned Off My iPhone 16's AI Camera Button. Life Has Never Been BetterThe article discusses the author's experience with the new camera control button on the iPhone 16. While excited about upgrading from an older iPhone, the author finds the camera control button to be an annoying and unreliable feature. The button, intended to provide quick access to the camera and AI-powered Visual Intelligence features, is easily triggered accidentally, leading to unwanted photos and battery drain. The author criticizes the button's sensitivity and questions its usefulness compared to existing methods of accessing the camera. The author views the camera control button as an example of tech companies prioritizing AI integration without considering the user experience, and expresses hope for more intentional AI feature implementation in the future. Instructions are provided for disabling the camera control button.Read more →
CNET...My Pick for Laptop Shoppers Seeking Tariff Relief: Acer's $700 Aspire 14 AIThe Acer Aspire 14 AI, priced at $700 at Costco, is highlighted as a budget-friendly Copilot Plus PC offering good performance and long battery life due to its Intel Lunar Lake CPU. While its design and display are not exceptional, the laptop provides modern components, including the latest Intel Core Ultra Series 2 processor with a neural processing unit (NPU) capable of 40 TOPS, 16GB of RAM, and a 1TB SSD. The review emphasizes its ability to handle AI workloads and future-proof capabilities, making it a strong contender in the budget laptop category for 2025. It avoids common pitfalls of budget laptops, such as outdated parts and flimsy construction, featuring aluminum top and bottom panels. The review also suggests considering the Microsoft Surface Laptop 7 and Asus Zenbook A14 for those willing to spend more.Read more →
CNET...2 Ways I'm Using ChatGPT Advanced Voice to Improve My LifeThe article discusses the author's experience using ChatGPT's advanced voice feature, highlighting its potential as a tool for creators and problem-solvers. The author, a content creator, details the differences between ChatGPT's standard and advanced voice modes, noting the latter's more natural and human-like interaction due to its unified model (GPT-4o). The author provides practical examples of how they use advanced voice, including as a sounding board for ideas, for more nuanced translations between English and Mandarin Chinese speakers, and for preparing for a presentation at the Cambridge Science Festival. The author also explains how to enable and utilize features like web search and file uploads in conjunction with advanced voice. The article emphasizes the versatility of advanced voice and encourages readers to explore its capabilities, noting its availability to both free and paid ChatGPT users.Read more →
The Verge...Amazon is reportedly training humanoid robots to deliver packagesAmazon is reportedly developing AI-powered humanoid robots for package delivery, aiming to automate the final stage of its logistics operations. According to The Information, Amazon is building a training facility, a "humanoid park," at its San Francisco office to test these robots, including models from Unitree, in scenarios involving Rivian electric vans. This initiative aligns with Amazon's recent launch of a new AI team focused on creating adaptable robots for distribution and logistics hubs. The company already employs various robots in its warehouses and acquired Zoox in 2020, signaling a broader strategy to automate package delivery from warehouse to customer.Read more →
CNET...WWDC 2025: Tune In to Watch Apple's Upcoming Software Event and Big iOS 19 RevealApple is preparing to host its annual Worldwide Developers Conference (WWDC) from June 9-13 at Apple Park in Cupertino, California. The event will feature a mix of online and in-person sessions, including a keynote and Platforms State of the Union address on June 9. WWDC traditionally unveils major software updates for iOS, iPadOS, and WatchOS. This year, expectations include a potential design overhaul for iOS 19, iPadOS 19, and MacOS 16, aiming for a more cohesive style across platforms reminiscent of VisionOS. The event is also expected to highlight Apple's advancements in AI, potentially revealing new Apple Intelligence tools. The keynote on June 9 at 10 a.m. PT will detail updates to Apple platforms, while the Platforms State of the Union at 1 p.m. PT will offer a deeper dive into new tools for MacOS, TVOS, VisionOS, and WatchOS. The event is free for developers to attend online, with in-person attendance requiring an application. While hardware announcements are unlikely, rumors suggest potential future releases like next-generation AirTags and a possible iPhone 17 Air.Read more →
TechCrunch...Alphabet CEO Sundar Pichai dismisses AI job fears, emphasizes expansion plansIn a recent interview, Alphabet CEO Sundar Pichai addressed concerns about AI's potential to displace a significant portion of the company's workforce. He emphasized Alphabet's commitment to growth and framed AI as a tool to enhance employee productivity and drive new product development, rather than a replacement for human workers. Pichai acknowledged the legitimacy of job displacement fears, referencing concerns raised by Anthropic CEO Dario Amodei, but highlighted Alphabet's expanding ventures, such as Waymo, quantum computing, and YouTube's growth in India, as evidence of ongoing innovation and opportunities. While acknowledging the uncertainty surrounding the path to artificial general intelligence (AGI), Pichai expressed optimism about future progress in AI, while also noting the possibility of encountering plateaus in technological advancement. The interview also touched on Alphabet's recent layoffs, which Pichai characterized as more targeted compared to the larger cuts in previous years.Read more →
TechCrunch...North America takes the bulk of AI VC investments, despite tough political environmentDespite a perceived hostile environment for AI R&D in the U.S. under President Trump's administration, North America continues to dominate global AI venture capital funding. Between February and May of this year, North American AI startups received $69.7 billion in VC funding across 1,528 deals, significantly surpassing Europe ($6.4 billion across 742 deals) and Asia ($3 billion across 515 deals). While policies such as cuts to scientific grants, restrictions on foreign students, and trade policies have created uncertainty, investors still view the U.S. as the primary hub for AI innovation and potential returns. In 2024, North American startups secured 75.6% of all VC AI funding, and this share has increased to 86.2% so far in 2025, indicating a continued concentration of AI investment in the region.Read more →
TechCrunch...iOS 19: All the rumored changes Apple could be bringing to its new operating systemAhead of Apple's WWDC 2025, rumors are circulating about the upcoming iOS 19, potentially to be renamed iOS 26. Key anticipated changes include a major design overhaul inspired by the Vision Pro, a dedicated gaming app integrating Apple Arcade and App Store games, and a virtual AI health coach providing personalized advice. Other rumored features include AI-powered battery management, reverse wireless charging (tested on iPhone 17 Pro), and AI translation for the Messages app. Notably absent from the rumors are the previously showcased advanced Siri capabilities, which have been delayed. While AI improvements are expected, the primary focus is anticipated to be on design changes.Read more →
Gizmodo...Everything We Expect Apple to Announce at WWDC25 on June 9Apple is reportedly planning a major software overhaul across its product ecosystem, with all operating systems, including iOS, macOS, and iPadOS, potentially jumping to version "26." This update, expected to be unveiled at WWDC25 on June 9, is rumored to feature a visual redesign called "Solarium," characterized by glassy, translucent elements inspired by visionOS. While significant changes are anticipated, they may be more subtle than expected, with updates to Apple apps and potential AI enhancements. Apple is also rumored to be exploring partnerships with AI companies like Anthropic. A gaming push, including a dedicated video game app, is also a possibility. New hardware announcements, such as the iPhone 17, are not expected at the event.Read more →
TechCrunch...Hugging Face says its new robotics model is so efficient it can run on a MacBookHugging Face has released SmolVLA, an open AI model for robotics, designed to democratize access to vision-language-action models and accelerate research in generalist robotic agents. SmolVLA, trained on community-shared datasets, reportedly outperforms larger models in both virtual and real-world environments. This release is part of Hugging Face's broader effort to create an ecosystem of low-cost robotics hardware and software, including the LeRobot collection and recent acquisition of Pollen Robotics. SmolVLA, with 450 million parameters, can run on a single consumer GPU or even a MacBook and supports an asynchronous inference stack for faster response times in dynamic environments. The model is available for download, and early users have reported success in controlling robotic arms with it. While Hugging Face is a significant player, other companies like Nvidia, K-Scale Labs, Dyna Robotics, Physical Intelligence, and RLWRLD are also active in the open robotics space.Read more →
The Verge...US removes 'safety' from AI Safety InstituteThe US Department of Commerce has renamed its AI Safety Institute to the Center for AI Standards and Innovation (CAISI), shifting its focus from broad AI safety to national security risks and preventing restrictive regulations. Secretary of Commerce Howard Lutnick announced the change, emphasizing the goal of enhancing US innovation and ensuring US dominance in international AI standards. This move contrasts with the institute's original mission under former President Biden, which aimed to establish global best practices for mitigating AI risks and involved collaborations with major AI companies. The new institute will concentrate on demonstrable risks like cybersecurity, biosecurity, and chemical weapons, as well as countering malign foreign influence from adversarial AI systems. This shift aligns with the Trump administration's broader effort to accelerate the growth of American AI companies, including rescinding Biden's AI safety standards and promoting AI adoption across various sectors.Read more →
