News/AI Models
AWS launches OpenAI’s first open-weight models in 6 years
AWS has launched day-of-launch availability of two new open-weight models from OpenAI on Amazon Bedrock and Amazon SageMaker, breaking Microsoft's traditional exclusivity with OpenAI. This marks the first time OpenAI has released open-weight models since GPT-2 in 2019, allowing AWS customers to fine-tune the models for specific use cases without directly interacting with OpenAI. What you should know: The two new models—gpt-oss-120b and gpt-oss-20b—represent OpenAI's first open-weight releases in six years. Open-weight models have visible parameters that allow AWS customers to fine-tune them for specific use cases, though the underlying training data isn't visible like in fully open-source models. OpenAI...
read Aug 12, 2025AI transforms K-12 education as 2.5M teachers give into it, save hours weekly
AI is rapidly transforming K-12 education, with nearly a third of teachers now using the technology weekly and millions accessing specialized platforms like MagicSchool AI. This shift represents a fundamental change in how both students and educators approach learning, creating new opportunities for personalized instruction while raising concerns about academic integrity and over-reliance on artificial intelligence. What you should know: The current high school senior class represents the last generation to experience pre-ChatGPT education, having started freshman year just months before the chatbot's release. Students have evolved beyond simple copy-pasting, now using multiple AI models and asking chatbots to introduce...
read Aug 11, 2025Why most AI pilots fail to scale beyond proof-of-concept
Artificial intelligence pilots generate excitement across enterprises, but most never escape the experimental phase. While hackathons produce impressive demos and leadership presentations showcase promising prototypes, the majority of these initiatives quietly stall in organizational silos, never achieving meaningful scale or business impact. The pattern repeats across industries—from financial services to manufacturing to healthcare. Companies excel at experimentation but struggle with the transition from proof-of-concept to operational reality. The gap between pilot and platform represents one of the most significant challenges facing enterprise AI adoption today. However, some organizations successfully navigate this transition. The difference isn't just technological capability—it's a fundamental...
read Aug 11, 2025On-premises GPU servers cost same as 6-9 months of cloud
A new analysis reveals that on-premises GPU servers cost roughly the same as six to nine months of equivalent cloud capacity, yet most AI executives remain unaware of this stark mathematical reality. This hidden cost structure means companies could save hundreds of thousands of dollars over three to five years by reconsidering their cloud-first AI infrastructure strategies. The big picture: While cloud computing promised flexible, pay-as-you-go scaling, AI workloads break these traditional assumptions in ways that make cloud economics misleading for sustained GPU-intensive operations. Key cost comparisons: The financial gap between cloud and on-premises becomes stark when examined closely. A...
read Aug 11, 2025AI model training could consume 4 gigawatts by 2030
AI model training could consume more than 4 gigawatts of power by 2030—enough to power entire cities—as energy demands for frontier AI development continue doubling annually, according to a new report from Epoch AI, a research institute investigating AI trajectory, and the Electric Power Research Institute (EPRI), an independent nonprofit. This exponential growth in power consumption poses significant challenges for utility companies and could derail tech giants' climate commitments, even as companies explore distributed training and flexible power solutions to manage the unprecedented energy demands. What you should know: Recent AI training runs like Elon Musk's Grok AI already require...
read Aug 11, 2025TD Securities launches AI assistant for equity teams using OpenAI tech
TD Securities has launched the TD AI Virtual Assistant, powered by OpenAI's GPT models and developed with Layer 6, a Canadian AI company acquired by TD Bank in 2018, to deliver real-time equity insights to its institutional sales, trading, and research teams. The assistant, which went live on July 8, represents a significant step forward for AI adoption in the highly regulated financial services sector, where banks have traditionally taken a cautious approach to implementing AI agents. What you should know: The TD AI Virtual Assistant serves as a sophisticated knowledge management system designed specifically for front-office equity professionals. The...
read Aug 11, 2025Claude gets 4 new personalization features to remember your work style
Anthropic has rolled out a comprehensive personalization upgrade for Claude, its AI assistant, marking a significant shift from generic chatbot interactions toward truly adaptive AI partnerships. The company, which competes directly with OpenAI's ChatGPT in the enterprise AI market, introduced four major features designed to make Claude remember your preferences, maintain project continuity, and adapt its communication style to match your specific needs. This upgrade addresses one of the most persistent frustrations with AI assistants: the need to repeatedly provide context and preferences in every conversation. Instead of treating each interaction as a blank slate, Claude now builds understanding over...
read Aug 11, 2025Still not getting it: Apple delays Siri overhaul to 2026 amid App Intents struggles
Apple's ambitious plan to transform Siri into a truly intelligent assistant is facing significant technical hurdles, with the company now targeting spring 2026 for a comprehensive overhaul that could finally deliver on promises made nearly two years ago. According to Bloomberg's Mark Gurman, a respected Apple analyst who frequently reports on the company's internal developments, Apple is making progress on what it calls an "overhauled" Siri experience. However, engineers are struggling with a critical component called App Intents—a feature designed to let Siri take meaningful actions across different applications on behalf of users. The stakes couldn't be higher for Apple....
read Aug 11, 2025Biotech startup Tahoe raises $30M for AI cancer drug discovery platform
Biotech startup Tahoe Therapeutics has raised $30 million in Series A funding led by Amplify Partners, bringing its total funding to $42 million and valuing the company at $120 million. The Palo Alto-based company has developed breakthrough technology for generating massive biological datasets needed to train AI models that can simulate living cells, positioning it to accelerate cancer drug discovery through digital cell modeling. What you should know: Tahoe's proprietary Mosaic platform can generate unprecedented amounts of single-cell data by testing multiple patient cell types simultaneously, rather than conventional one-patient-at-a-time approaches. In February 2024, the company released Tahoe-100M, a dataset...
read Aug 11, 2025Survey reveals AI agents can control computers but create massive security risks
Researchers from Zhejiang University and OPPO AI Center have published the most comprehensive survey to date of "OS Agents"—AI systems that can autonomously control computers, mobile phones, and web browsers by directly interacting with their interfaces. The 30-page academic review, accepted for publication at the Association for Computational Linguistics conference, comes as major tech companies including OpenAI, Anthropic, Apple, and Google race to deploy AI agents capable of performing complex digital tasks, while highlighting significant security vulnerabilities that most organizations aren't prepared to address. The big picture: This technology represents a fundamental shift toward AI systems that can genuinely understand...
read Aug 11, 2025Financial firms deploy AI to auto-generate fraud rules as UK losses hit $1.3B
Financial services firms are increasingly deploying AI-powered fraud detection systems that automatically generate and optimize rules based on historical data patterns, replacing traditional manual rule creation. This shift comes as fraud losses in the UK reached £1.1 billion in 2024, with confirmed fraud cases rising 14% to 3.13 million, driven by more sophisticated AI-enabled attacks including deepfakes and synthetic identities. The scale of the problem: Fraud has become the most common crime in the UK, accounting for 41% of all crime in England and Wales, with financial services firms facing particular challenges. Q1 2024 saw 8,374 consumer complaints about fraud...
read Aug 11, 2025Evasive though persuasive: Study finds AI reasoning models produce fluent nonsense instead of logic
University of Arizona researchers have found that large language models using "chain of thought" reasoning are fundamentally flawed at logical inference, functioning more like "sophisticated simulators of reasoning-like text" than true reasoners. The study reveals that these AI systems, which the industry increasingly relies on for complex problem-solving, fail catastrophically when asked to generalize beyond their training data, producing what researchers call "fluent nonsense" with a deceptively convincing appearance of logical thinking. The big picture: The research challenges the AI industry's growing confidence in reasoning models by demonstrating that apparent performance improvements are "largely a brittle mirage" that becomes fragile...
read Aug 11, 2025House Democrats push CMS to cancel AI-driven Medicare approval program
More than a dozen House Democrats have pressed Centers for Medicare & Medicaid Services (CMS) Administrator Mehmet Oz to cancel a planned prior authorization pilot program that would expand AI-driven approval requirements to traditional Medicare. The program, set to begin testing in six states this January, represents a significant shift for traditional Medicare, which has historically operated without extensive prior authorization requirements. What you should know: The pilot program incorporates artificial intelligence to help make healthcare coverage decisions, marking a departure from traditional Medicare's historically streamlined approach. Representatives Suzan DelBene of Washington and Ami Bera of California led the Democratic...
read Aug 11, 2025Reddit blocks Internet Archive to protect $200M AI licensing deals
Reddit has blocked the Internet Archive from indexing its content after discovering AI companies were circumventing Reddit's scraping restrictions by harvesting data from archived pages instead. The move effectively eliminates a key resource for researchers and users who relied on archived Reddit content to track deleted posts and preserve community discussions. What you should know: The Internet Archive's Wayback Machine can now only capture screenshots of Reddit's homepage, not individual threads, profiles, or comments. Previously, the Wayback Machine served as a comprehensive backup of Reddit content, documenting everything from deleted posts to user activity across various subreddits. Moving forward, the...
read Aug 11, 2025Dim future: Microsoft Lens app shutting down beginning in fall as users directed to Copilot
Microsoft is shutting down its popular Lens PDF scanning app after nearly a decade, directing users toward its AI-powered Microsoft 365 Copilot instead. The retirement affects over 92 million users who relied on the free, straightforward document scanning tool that converted printed and handwritten notes into PDFs and other file formats. What you should know: Microsoft Lens will be completely phased out over a four-month period starting this fall. The app will be retired from iOS and Android devices on September 15, 2025, and removed from app stores on November 15. Scanning functionality will be disabled on December 15, 2025,...
read Aug 11, 2025Sitch dating app uses AI matchmaker to set up $160 blind dates in NYC, SF and LA
A new dating app called Sitch uses an AI matchmaker to set up users on blind dates, charging up to $160 for eight introductions. Instead of browsing profiles or messaging matches directly, users answer questions about their preferences and values, then ask the AI about potential dates before deciding whether to meet in person. How it works: The AI-powered matchmaking process replaces traditional swiping with conversational screening. Users first answer questions from the AI "matchmaker" about their interests, ideal dates, and personal values. The AI compiles this information into categories like "Non negotiables," "Red flags," and "Nice to Haves." When...
read Aug 11, 2025Cleveland Clinic partners with Piramidal to monitor brain health in real time
The Cleveland Clinic is partnering with San Francisco-based startup Piramidal to develop an AI foundation model that monitors patients' brain health in intensive care units using electroencephalogram (EEG) data. The system aims to interpret continuous streams of brain wave data and flag abnormalities in seconds, potentially transforming how doctors detect neurological issues in critically ill patients. Why this matters: Current EEG monitoring in ICUs requires manual review that can take two to four hours for a day's worth of data, with reports generated only every 12 to 24 hours—delays that could prove critical for patients experiencing seizures or declining brain...
read Aug 11, 2025BOLO: Cybercriminals use AI to create perfect fake government websites
Cybercriminals have discovered a powerful new weapon in their arsenal: generative artificial intelligence. Security researchers recently uncovered a sophisticated phishing campaign where hackers used AI tools to create nearly perfect replicas of Brazilian government websites, demonstrating how machine learning is making online fraud more convincing and harder to detect. The fake websites were so convincing that they could easily fool unsuspecting citizens seeking government services. This represents a concerning evolution in cybercrime, where AI democratizes the ability to create professional-looking scams that previously required significant technical expertise. The anatomy of AI-powered government impersonation Zscaler ThreatLabz, a cybersecurity research division of...
read Aug 11, 2025US judges adopt AI for legal research despite mounting errors
US federal judges are increasingly experimenting with generative AI to help with legal research, case summaries, and routine orders, despite recent high-profile mistakes where AI-generated errors went undetected in court rulings. The trend highlights a growing tension between judicial efficiency and accountability, as judges face fewer consequences than lawyers when AI mistakes slip through—yet their errors carry the force of law. The big picture: While lawyers have faced sanctions and embarrassment for submitting AI-generated briefs with fabricated cases, judges are now making similar mistakes with far greater consequences. In June, a Georgia appellate court judge issued an order relying on...
read Aug 9, 2025Condos with filters? Real estate agents use AI to fake property photos, sparking legal concerns
Real estate agents are increasingly using AI-generated images to enhance property listings, often creating misleading representations of homes that can dramatically differ from reality. The practice has sparked outrage from prospective buyers and raised legal concerns about false advertising in an industry already struggling with trust issues. What you should know: The Register, a UK technology publication, uncovered a property listing featuring an AI-generated image that showed structural elements and landscaping that didn't exist in the actual home. The manipulated photo included misaligned awnings, hedges that morphed into walls, and a flowerbed blocking a neighboring door that wasn't present in...
read Aug 8, 2025LangChain launches Open SWE, an AI agent for autonomous coding tasks
LangChain has launched Open SWE, an open-source asynchronous coding agent that operates in the cloud and integrates directly with GitHub repositories. The tool represents a significant evolution in AI-powered software development, allowing developers to delegate complex coding tasks that the agent can complete autonomously over extended periods. What you should know: Open SWE functions like an additional team member, capable of researching codebases, creating execution plans, writing code, running tests, and opening pull requests. The agent has already become a top contributor to LangChain's own projects, including LangGraph and its own repository. Users can get started in minutes with just...
read Aug 8, 2025OnePlus 13 gets AI Perfect Shot to fix blinks and bad expressions
OnePlus 13 users are receiving a software update that introduces AI Perfect Shot, the company's version of Google's Best Take feature for automatically fixing facial expressions in photos. The feature can recognize faces and replace unwanted expressions like blinks or turned heads with better alternatives from the same photo session, ensuring users always capture the perfect shot. What you should know: AI Perfect Shot works similarly to Google's Best Take by analyzing facial expressions and swapping in better alternatives from the same photo sequence.• The feature can fix common photo problems like poorly-timed blinks or people looking away from the...
read Aug 8, 2025WIRED reveals AI chatbot pricing is based on “vibes,” not economics
AI premium chatbot subscriptions like ChatGPT Pro and Claude Max cost around $200 per month, but their pricing appears to be based more on market positioning than actual economics. WIRED's investigation revealed that OpenAI CEO Sam Altman essentially set the $200 benchmark when launching ChatGPT Pro, and competitors simply followed suit—despite none of these companies claiming to profit from these premium tiers. What you should know: The $200 price point emerged through "vibes-based pricing" rather than careful financial analysis. OpenAI's ChatGPT Pro launched first at $200 monthly, with CEO Sam Altman openly acknowledging they weren't making money on the service...
read Aug 8, 2025Apple Intelligence users must wait until iOS 26 for GPT-5 integration
Apple Intelligence's ChatGPT integration will upgrade to GPT-5 with the release of iOS 26, iPadOS 26, and macOS Tahoe 26, according to Apple's confirmation to 9to5Mac. This means users will have to wait until Apple's next major software cycle to access OpenAI's newest AI model through Apple's ecosystem, even though GPT-5 became available to all ChatGPT users on Thursday. What you should know: Apple Intelligence currently relies on OpenAI's GPT-4o model for its ChatGPT integration features. The integration helps power certain Siri queries and works alongside Apple's Visual Intelligence feature, which functions similarly to Google Lens by identifying objects and...
read