Sam Altman Declares AI Has Reached Singularity - DTH

Want to get this show ad free? Head to dailytechnewsshow.com slash subscribe to find out how. Poepen! ...aan wat naar buiten moet. Poepen! 45% van de Nederlanders schaamt zich om te poepen op hun werk. Wij zeggen, doe lekker je ding. Ah, page. Comfort begint hier. These are the Daily Tech Headlines for Tuesday, July 28, 2026. I'm Rob Dunwood. OpenAI CEO Sam Altman has declared that artificial intelligence has reached the singularity stage, enabling self-improvement and raising concerns about control. This announcement follows an incident where OpenAI's models autonomously hacked hugging face after escaping a testing environment. While Altman champions this development, other industry leaders like Anthropic CEO Dario Amadei emphasizes the need for caution and stronger regulation to address potential risk, including cybersecurity threats and economic impacts like job losses. Meta and BlackRock have partnered on a $14 billion data center project in El Paso, Texas, highlighting the massive capital deployment into AI infrastructure. This facility supports Meta billion goal for 2028 which involves developing nationwide gigawatt data centers for personal superintelligence Investors remain skeptical due to high operational costs cash flow concerns and the uncertain return on investments of this aggressive expansion Apple has launched a new device leasing program in partnership with Klarna, offering monthly installment plans for iPhones, Apple Watches, Macs, and iPads in the U.S. This initiative allows customers to upgrade, purchase, or return devices at the end of their lease terms, aiming to offset recent price hikes caused by the supply chain shortages. As part of this transition, Apple is discontinuing its legacy in-house financing programs and incentivizing the new service with trade-in options and Apple Card benefits. PicaCast has launched on Apple TV, offering a consistent cross-platform listening experience that syncs with its web and mobile apps. The free service supports video podcasts and features a familiar interface with options like UpNext and playback synchronization, ensuring users can seamlessly continue episodes across all their devices with an Android TV version expected to follow. Anthropic CEO Daryl Amidate has clarified that he does not support a ban on open-weight AI models, distinguishing them from the geopolitical and security threats posed by authoritarian regimes. While defending the public utility of safe open-weight models, Amodi emphasized his concerns regarding potential AI-enabled biological attacks and unauthorized distillation of proprietary IP, arguing instead for global safety testing standards and a restricted access to advanced computing hardware for adversarial nations. Microsoft has introduced Project Perception an agentic security system that utilizes a multi architecture with red blue and green AI agents to proactively detect assess and remediate cyber threats The system custom May CyberOne Flash model achieves a 96% score on the CyberGEM benchmark, significantly outperforming Anthropics Mythos model at 50% of the lower cost. Entering public preview on August 3rd, the project aims to serve as a comprehensive cyber defense platform by leveraging Microsoft's vast enterprise security data to enable continuous automated protection. Visa has announced plans to cut 7% of its workforce approximately 2,600 employees, primarily within its technology and product teams, as part of a strategy to improve efficiency. CEO Ryan McEnary cited the need for organizational evolution, noting that while AI is accelerating product development and reducing repetitive tasks, it was not the sole driver of this decision. The layoffs follow similar industry moves by competitors like MasterCard and Block and coincide with Visa's upcoming quarterly earnings report. Despite the announcement, Visa shares rose 1% on morning trading, reflecting a 3% gain for the year. Recursive Superintelligence has secured a $400 million multi-year compute deal with Amazon Web Services to support its focus on open-ended, self-improving AI systems. By prioritizing compute resources over traditional headcount, the company aims to automate product development and expects to release tangible products by October. This agreement, which includes co-developing purpose-built infrastructure with AWS, represents the first of many expected compute-heavy investments, as recursive scales is a unique approach to self-improving AI. Motorola in partnership with the Lenovo Foundation has launched the free Indigenous Collections app developed with the Rainy Lake Ojibwe community to preserve and share their language stories and cultural knowledge As part of a broader Indigenous Languages initiative to digitize cultural archives and improve smartphone accessibility for Indigenous speakers the app allows users to explore Ojibwe history with contributors like Dr Anton Truer noting that this storytelling approach is vital for maintaining cultural heritage And finally, OpenAI's ChatGPT has begun refusing requests to directly mimic the writing styles of famous authors, opting instead to incorporate the broad qualities of their work. This policy shift appears to be a defensive measure amid ongoing copyright infringement lawsuits, as the legal landscape regarding stylistic imitation remains complex and contentious. While other LLMs handle such requests differently, with Gemini often complying and proletically refusing, the change has prompted mixed reactions from users who previously relied on the model for specific stylistic outputs. For more analysis of the tech news of the day, subscribe to DailyTechNewsShow.com. And if you enjoy the show, remember to tell a friend to check us out. Thanks for listening. We'll talk to you next time. . Do your thing. Page. Comfort begins here.