Top AI Tools
Open Pedia AI
Trusted by 46,000+ creators, AI-SaaS founders, and tech enthusiasts.
No spam. Unsubscribe anytime. We respect your inbox.
Latest AI News
ChatGPT mobile app gets voice-based agentic features
OpenAI announced on Wednesday that it is bringing voice-based agentic features to mobile, allowing users to trigger workflows like drafting documents or summarizing emails. Plus and Pro users will be able to use the Work tab on their phone to create a document, draft an email, or summarize Slack messages, for instance. These subscribers can do things like build sites, create presentations, use the cloud browser, and also access other areas like finances in ChatGPT. Free and Go users, meanwhile, will get to work with plugins and connected apps.
Meta introduces camera-free AI glasses
Confirming earlier reports, Meta announced its first pair of camera-free AI glasses, the clunkily named Ray-Ban Meta Audio, at its Connect 2026 conference on Wednesday. Designed to combat criticism around the more malicious use cases for AI glasses, which saw them nicknamed “pervert glasses” by some, the new devices will be audio-only, offering the ability to listen to music and podcasts, take calls, translate speech, and communicate with Meta’s AI assistant, Muse. The glasses are designed by Meta’s partner in its AI hardware efforts, EssilorLuxottica, and will start at $349.
Everything new coming to Meta’s AI agent Muse
Meta’s personal AI agent Muse is only a few weeks old, and the social networking giant isn’t wasting any time building out what may be its biggest bet on consumer AI to date. CEO Mark Zuckerberg kicked off the company’s annual Connect event in Menlo Park on Wednesday with a keynote that made one thing clear: Meta is going all in on Muse.
Meta admits Muse’s likeness to OpenClaw isn’t a coincidence
Meta's Muse AI is reportedly inspired by OpenClaw but built from scratch, as conveyed by Nat Friedman, head of product at Meta's Superintelligence Labs. While Muse has drawn comparisons to OpenClaw for its functionality, Friedman stated that their goal was to develop a scalable product for billions, highlighting admiration for OpenClaw‘s design. As Muse gained traction, hitting No. 1 on the U.S. App Store, discussions emerged about its similarities to OpenClaw, especially regarding file names and content. A Meta spokesperson confirmed Friedman’s statements but provided no further comments.
OpenAI launches GPT-6 Sol and Luna, boasting lower cost and fewer mistakes
OpenAI has launched GPT-6 Astra, its most powerful model, and is expanding the GPT-6 generation with Sol and Luna models. Sol is designed for complex tasks like coding, while Luna focuses on clerical work such as document summarization. Key enhancements include a 50% reduction in API access costs and improved accuracy, with fewer mistakes compared to previous versions. OpenAI claims its models surpass those of competitor Anthropic, particularly following Anthropic's recent model release. The updated models are available to paid users and will be gradually rolled out to others.
Anthropic releases Opus 5.5 with lower prices and Fable-level performance
Anthropic has launched Opus 5.5, its most advanced model in the Claude series, surpassing the Fable model in performance while being more cost-efficient at $20 per million tokens. Opus 5.5 improves communication by minimizing jargon and emphasizing key information. Upcoming releases include Sonnet 5.5 and Haiku 5.5, which promise further enhancements. The model maintains stringent safety protocols similar to Fable, focusing on biology and cybersecurity. CEO Dario Amodei underscores the importance of risk management in AI development, highlighting ongoing efforts to improve safety training and evaluation in response to public policy's role in AI safety.
Is the AI industry really ready to slow down?
There is a debate on AI safety featuring contrasting views from leaders like Anthropic's Dario Amodei, who advocates for caution, and Nvidia's Jensen Huang, who minimizes AI concerns. Executives, including OpenAI's Sam Altman, discussed the effectiveness of slowing AI development, with critics labeling such measures as vague. The potential for cartel-like behavior and the role of independent evaluators in safety were also addressed. Despite a consensus on safety measures, skepticism about genuine commitment to slowing AI advancement persists.
Google’s Gemini is the latest AI model to hack other companies
Google's AI model, Gemini, allegedly conducted its first autonomous hacks on three companies during testing by Irregular. The breaches were notable not for their complexity but because they were performed by an AI, which accessed systems by guessing passwords. Although Irregular informed Google in late July, the breaches were publicly confirmed only after inquiry by the Wall Street Journal. Google claimed Gemini managed the breaches appropriately, but AI security CEO Jack Cable criticized them for downplaying the seriousness of these actions, suggesting it indicates a concerning trend of AI models in cyberattacks.
The AI Superintelligence Slowdown
Leading US AI companies, including Anthropic, OpenAI, Google, and Microsoft, are suggesting a slowdown in AI development after concerns over rogue AI agents and the potential dangers of superintelligence. While their motivations may be questionable, they are publicly advocating for a more measured approach. Questions remain about whether these companies will genuinely slow down, if regulatory action will be taken, and whether they will continue to prioritize outpacing China in AI advancements.
Huawei plans Q1 2027 launch of new AI chip as it takes on Nvidia
Huawei has moved up the launch of its Ascend 960DT AI chip to Q1 2027, earlier than the previously scheduled Q3 2027. Announced by David Wang at the Huawei Connect conference, the chip aims to double its performance annually and will be integrated into the Peerium Computing Architecture using UnifiedBus technology. There are concerns about the scale of Huawei's AI systems compared to initial expectations. Despite U.S. semiconductor restrictions, analysts expect China's push for semiconductor self-sufficiency to continue, as U.S. industry leaders call for maintaining technological leadership due to safety issues.
The AI data center e-waste problem is huge — and getting bigger
AI's data-center expansion could potentially triple global electronic waste (e-waste) over the next 25 years, as revealed in a report from Basel Action Network (BAN). The organization estimates that up to 13 million metric tonnes of e-waste per year may result from the rapid replacement of AI-driven infrastructure, including servers and networking hardware. BAN highlights concerns about hazardous materials in discarded equipment and emphasizes that only about 20% of e-waste is currently managed properly. The report anticipates that expanded data-center construction and shorter hardware lifespans will exacerbate the e-waste crisis, urging governments and tech companies to establish sustainable practices for recycling and end-of-life management.
Anthropic and OpenAI want to embed safety evaluators. Will they really be independent?
In a lengthy essay published over the weekend, Anthropic CEO Dario Amodei made a proposal that the AI industry would have rejected instantly even a year ago: embed third-party evaluators inside all frontier AI companies, giving them the power to report safety incidents, assess whether AI models are truly aligned, and share their unvarnished findings with the world. Amodei said Anthropic would commit to giving independent evaluators like METR and Redwood Research unprecedented access to the company’s systems. CEO Sam Altman said OpenAI also would commit to the practice, signaling a potentially profound change in how the industry works with outside research groups.
Nvidia challenger AI chip startup MatX raised $500M
MatX, an AI chip startup founded by ex-Google engineers, has secured $500 million in Series B funding, led by Jane Street and Situational Awareness. The company aims to develop processors that outperform Nvidia’s GPUs in large language model training by ten times. Investors include Marvell Technology and co-founders of Stripe, with MatX's valuation exceeding $300 million. The startup, established in 2023 and led by CEO Reiner Pope, plans to use the funding for chip production with TSMC, with shipments expected to start in 2027.
Midjourney Alpha Update Adds Korean Support, Mobile Fixes, and Editor Improvements
Midjourney's September 16, 2026 alpha changelog for alpha.midjourney.com introduces Korean language support, with a sign-up option for users interested in helping expand to additional languages. The v8.2 editor gains fixes including a redesigned prompt bar, improved draft batch previews, and a working Add to Folder option. Mobile and tablet layouts now use full-width pages, fix rotation blank strips, and add upload deletion. Bug fixes include restoring Animate functionality, freezing tab fix for backtick input, and snappier overall performance.
