Weekly AI News

Hey it’s Jul,
Greetings and welcome to the seventh edition of “Weekly AI News”!

This week, AI certainly isn’t doing things halfway. Between OpenAI trying to sell us specialised AI agents for the price of a car per month and Elon Musk struggling to convince the courts to let him keep OpenAI as his personal toy, the atmosphere remains electric.

Meanwhile, Amazon and Alibaba are betting on “hybrid” intelligence at low cost, and Apple reminds us that innovation takes time… way too much time (Siri, see you in 2027 👋).

Even McDonald’s is giving in to the AI craze to boost its 43,000 restaurants—proof that the future has made its way onto our plates. But let’s not forget: if Manus, the autonomous Chinese AI agent, can apply for jobs on Upwork on your behalf, it might soon be negotiating your salary too… 

Happy reading!

🇨🇳🤖 Chinese Startup Unveils Manus, a Fully Autonomous AI Agent

Manus claims to handle real-world tasks—from screening job candidates to designing graphics—entirely on its own. It reportedly outperformed ChatGPT and Google’s Gemini on GAIA benchmarks focused on agentic tasks, demonstrating advanced independence in web browsing, coding, and gig-economy platforms like Upwork.

The invite-only system also integrates property research and data analytics, highlighting a new wave of hands-off AI solutions that require minimal human oversight. The founders plan to open-source Manus’s underlying models later this year, hoping to accelerate broader advancements in fully autonomous AI.

💼🤖 OpenAI to Launch Premium AI Agents

OpenAI is developing specialised AI agents priced from $2,000 to $20,000 monthly, aimed at business professionals, software devs, and PhD-level research. Investor SoftBank has reportedly committed $3B for 2025. CEO Sam Altman predicts enterprise adoption of AI “workforce agents” will be game-changing, potentially driving a quarter of OpenAI’s revenue as companies pay for advanced automated expertise.

🎬🤖 Sora Video AI Headed to ChatGPT

OpenAI confirmed plans to integrate its Sora video-generation tool directly into ChatGPT’s interface. While the ChatGPT version will have fewer editing features than Sora’s standalone app, a future mobile app and faster “Sora Turbo” model are also in the works. OpenAI hopes this integration reignites interest after Sora’s underwhelming launch, putting video creation front and center for ChatGPT users.

🚀🤖 GPT-4.5 Preview Expands to All ChatGPT Plus Users

OpenAI rolled out its new GPT-4.5-Preview model to all ChatGPT Plus subscribers. Previously available only to Pro tier and developers via API, this model brings upgraded language understanding and generation capabilities. Early testers report more nuanced responses and improved reasoning, positioning GPT-4.5 as a stepping stone to future OpenAI releases.

🏫💡 OpenAI Launches $50M NextGenAI Academic Consortium

OpenAI announced NextGenAI, a $50M-funded initiative partnering with top institutions like Harvard and Oxford. The program provides grants, compute resources, and APIs to advance AI research in fields such as rare disease diagnostics and historical text digitisation. It follows ChatGPT Edu, OpenAI’s specialised GPT-4o for academia. This move underscores the company’s commitment to accelerating AI-driven innovation across universities.

🍎💻 ChatGPT on Mac Now Edits Code Directly

OpenAI’s ChatGPT app for macOS can now auto-edit code in Xcode, VS Code, and JetBrains. Users can enable “auto-apply” mode to have ChatGPT make changes without extra confirmation. Currently exclusive to ChatGPT Plus, Pro, and Team subscribers on macOS, the feature will roll out to Enterprise, Edu, and free plans soon. A Windows version is also in the pipeline.

📈💬 ChatGPT Weekly Active Users Double in Under 6 Months

A16z reports that ChatGPT soared from 200 million to 400 million weekly active users in less than half a year. The app initially drew consumer interest for novelty, but new integrations and use cases are driving sustained engagement. After launching as a research preview in late 2022, ChatGPT continues to break user-adoption milestones, fueling the broader generative AI trend.

💡🤫 Anthropic CEO Teases Future Release for Claude 4

Dario Amodei revealed Anthropic is “reserving” its Claude 4 models for larger leaps, expecting them to surpass top human coders by 2026. The company’s patient rollout strategy contrasts with competitors’ frequent updates. By waiting for more robust performance gains, Anthropic hopes Claude 4 will make a major impact in coding and advanced AI tasks upon its release.

🏭🔬 Anthropic Joins DOE’s ‘1,000 Scientist AI Jam’

Anthropic will participate in the Department of Energy’s initiative, where 1,000 scientists test AI tools for national security and advanced research. Claude will be evaluated on tasks from climate modeling to energy infrastructure security. This partnership highlights the growing role of private AI models in critical governmental and scientific arenas.

🏛️📜 Anthropic Submits AI Action Plan to the White House

Anthropic provided detailed recommendations calling for tighter export controls, national security stress-testing, and increased infrastructure. Their proposal aligns with broader efforts to regulate AI responsibly while maintaining innovation. By engaging directly with U.S. policymakers, Anthropic aims to shape a more secure and ethically guided AI ecosystem.

🤔💻 Claude 3.7 vs. 3.5: Which Is Better for Coding?

Anthropic’s Claude 3.7 garners mixed feedback: while it shows stronger autonomy, some devs say it overcomplicates code or ignores prompts. Others appreciate its problem-solving abilities when given structured instructions. The debate underscores a shift in AI design toward autonomous solutions — and reminds us that “newer” isn’t always “better.”

💰🚀 Anthropic Hits $61.5B Valuation in New Funding Round

Anthropic raised $3.5B in a Series E, marking a $61.5B post-money valuation. Lightspeed Venture Partners led the round, with Salesforce Ventures and others joining. The new capital will fund bigger compute resources, safety research, and global expansion. Anthropic’s Claude 3.7 Sonnet and Claude Code remain central to its strategy, powering partnerships like Amazon’s Alexa+.

📄🔎 Mistral Launches OCR for AI-Ready Document Processing

Mistral AI has unveiled a powerful API that extracts data from complex documents with speed and accuracy. It supports images, equations, and tables, converting them to markdown for AI workflows. Mistral OCR processes up to 2,000 pages per minute, outperforming rivals like Google’s Document AI and GPT-4o. This tech promises to transform heavy-document industries, making archives truly AI-accessible.

🤖🏢 Meta Plans Major Expansion of AI for Businesses

Meta aims to offer AI solutions to “hundreds of millions” of businesses, particularly small firms lacking AI teams. Features could include business agents that manage tasks like customer engagement. As AI use spreads, Meta wants to ensure small ventures can access advanced tools, mirroring broader trends of AI democratisation across the corporate world.

🏥🤖 Microsoft Launches AI Copilot for Healthcare Tasks

Microsoft’s new Dragon Copilot helps clinicians by automating notes, referrals, and more, building on its Nuance acquisition. The system listens during doctor-patient visits to generate documentation, reportedly saving five minutes per patient. Initially available in the U.S. and Canada, the Copilot aims to reduce clerical burdens and combat burnout among healthcare workers.

🛍️🧠 Amazon Debuts Hybrid Reasoning AI Model

Amazon is working on “Nova,” a hybrid reasoning system that blends quick responses with deeper multi-step logic. The goal: deliver top-tier performance without sky-high costs, challenging OpenAI, Anthropic, and Google. With AGI-focused leadership at the helm, the project signifies Amazon’s bid to become a leading AI innovator alongside its partnership with Anthropic.

☁️🦾 AWS Establishes Dedicated AI Agent Division

Amazon Web Services formed a new internal group to develop enterprise-focused AI agents. It unifies teams working on Amazon Q, an AI chatbot and assistant unveiled at re:Invent. Spearheaded by longtime exec Swami Sivasubramanian, this division bets big on agentic AI for tasks like customer service and financial forecasting, potentially unlocking a multi-billion-dollar opportunity.

🌐👀 Cohere For AI Unveils Aya Vision Multilingual Model

Aya Vision, available in two sizes (8B and 32B), excels at image understanding and is fluent in 23 languages. It competes with much larger rivals, outpacing them in visual tasks while also translating and interpreting images worldwide. Released under a non-commercial license, Aya Vision advances global accessibility in AI-driven image analysis.

🔎🤖 Google Search Debuts ‘AI Mode’ for Conversational Results

Google’s new AI Mode transforms search into a dynamic conversation powered by Gemini 2.0. It simultaneously queries diverse sources to deliver detailed responses with cited links. With upgraded AI Overviews also on offer, Google aims to keep pace with ChatGPT-like services, blending familiar search mechanics with next-gen AI dialogue.

🏭🤖 Larry Page Launches Dynatomics for AI-Driven Product Designs

Google’s co-founder Larry Page is reportedly starting Dynatomics, an AI firm focused on factory-ready product designs. Drawing on LLMs, the startup aims to streamline manufacturing planning, from prototypes to assembly lines. It signals Page’s continued interest in disruptive AI ventures beyond Google’s main ecosystem.

🤖❤️ Tavus Debuts Emotionally Aware AI Avatars

Tavus’ new Conversational Video Interface incorporates three AI models—Phoenix-3, Raven-0, Sparrow-0—to capture emotional signals and respond with natural animations. Avatars maintain eye contact, react to user moods, and speak at realistic intervals. This leap in “emotional intelligence” hints at a future where AI-driven customer service and digital assistants feel more human than ever.

🌐🚀 Opera Introduces Browser Operator AI at MWC

Opera’s Browser Operator uses voice commands to autonomously perform tasks like opening sites, filling forms, or collecting data—no cloud dependency needed. Users can interrupt or guide the AI at any time, prioritising privacy and real-time control. Currently in preview, it’s set to launch widely as part of Opera’s broader AI roadmap.

💸💡 Alibaba Releases Low-Cost QwQ-32B AI Model

QwQ-32B employs scaled-up reinforcement learning to match larger competitors on coding, math, and complex reasoning—all at 90% lower cost. It charges just $0.20 per million tokens, undercutting models like DeepSeek-R1. Open-sourced under Apache 2.0, QwQ-32B continues China’s rise in affordable, high-performance AI solutions.

📱💥 Samsung Unveils $300 Galaxy A Phones With New AI Features

Samsung’s latest Galaxy A lineup brings AI-based search and photo editing to budget devices. Priced at $300, the phones aim to compete directly with Apple’s $599 iPhone 16e. By integrating AI features more widely, Samsung targets consumers who want accessible devices without sacrificing modern capabilities.

🍏⏳ Apple Faces Significant Delays in Revamping Siri

A complete overhaul of Siri’s fragmented architecture is pushed to 2027, per insider reports. Apple planned to unify legacy code and new AI systems, but struggled with engineering complexities and chip shortages. With rivals launching advanced voice assistants, Apple’s measured pace highlights the risks of falling behind in the AI race.

📲🤖 Deutsche Telekom Teams with Perplexity for AI Phone

T-Mobile’s parent company is developing an AI Phone integrating Perplexity Assistant for on-lock-screen queries and tasks. The device will also tap Google Cloud AI, ElevenLabs, and Picsart, aiming to shift from app-driven to AI-driven experiences. Expected to launch under $1K, it’s a bold move toward proactive mobile intelligence.

🎙️😮 Sesame Aims to Cross the ‘Uncanny Valley’ With New AI Voice Tech

Brendan Iribe’s startup Sesame demonstrated a voice AI that responds with natural tone and real-time emotional awareness. By adjusting pace, pitch, and personality mid-conversation, it delivers an unsettlingly lifelike experience. Early testers reported feeling “freaked out” by the realism, underscoring the leap in AI-driven conversational presence. Try it here.

💼🛍️ Salesforce Launches AgentExchange, the World’s First AI Agent Store

Salesforce’s new marketplace for advanced AI agents aims to help businesses tackle complex, automated tasks. The company predicts these agentic solutions will add $6 trillion to the global economy by 2030. By offering curated, enterprise-ready AI, Salesforce positions itself at the forefront of agent-centric digital transformation.

⚖️🚫 Judge Declines to Block OpenAI’s For-Profit Conversion

A federal judge refused Elon Musk’s request to halt OpenAI’s shift into a for-profit model, instead scheduling a trial for the fall. Musk argues the move breaks prior agreements and harms charitable aims. However, the judge noted Musk’s own $97B bid for OpenAI undercuts claims of “irreparable harm.” Two of Musk’s claims were dismissed, but the court did acknowledge public funds deserve protection.

🇺🇸💰 Trump and TSMC Unveil $100B U.S. Chip Factory Initiative

Taiwan Semiconductor Manufacturing Company will invest $100B to build three new U.S. plants and two packaging facilities for AI chips. The move could create tens of thousands of high-paying jobs. This largest-ever foreign direct investment in America aims to bolster domestic chip production for AI, advanced computing, and other tech industries.

📞🔊 Global Call Center Giant Uses AI to ‘Neutralise’ Indian Accents

France-based Teleperformance, the world’s largest call center operator, teamed up with Sanas AI to remove or soften Indian accents in real time. The company claims it improves customer satisfaction, though critics worry about cultural implications. Teleperformance has exclusive rights to Sanas’ “accent translation” tech, spotlighting AI’s growing role in outsourced customer service.

📈⚙️ Report: Using Generative AI Boosts Worker Productivity by 33%

A Federal Reserve Bank study found that employees integrate generative AI into regular workflows, not just as an occasional tool. Productivity jumped by a third each hour they used AI. As AI adoption rises, organisations may see significant efficiency gains and are likely to invest more heavily in these transformative technologies.

🍔🏪 McDonald’s to Roll Out AI Upgrades at 43,000 Locations

The fast-food giant is deploying AI-powered drive-throughs, connected kitchen equipment, and predictive order systems worldwide. By automating routine tasks and anticipating equipment failures, McDonald’s aims to cut wait times and reduce stress for employees. The transformation illustrates how AI is reshaping even the most established global brands.

Community

Join AI Whisperer Community!

Ready to take your AI journey to the next level? Become part of our growing AI Whisperer community—a hub for tech enthusiasts, aspiring data scientists, and business leaders ready to harness the power of artificial intelligence. Inside, you’ll find:

• Community: Like-minded individuals keen to grow together

• Curated AI Tools: Discover the latest and greatest tools to boost productivity.

• Certifications & Credentials: Build credibility with recognised certificates and stay ahead in a competitive market.

Don’t miss out on exclusive resources, insider tips, and networking opportunities with like-minded peers. Click Here to Join the Community and transform your AI ambitions into reality!

That's it for this week!
Until next time, stay curious and keep exploring the ever-evolving world of AI!
Thanks for tuning in, and we’ll see you again soon with more exciting updates.

Jul