
The All-in-One AI Pulse: Latest Tools, Trends, and Updates in your Inbox
"AI will reach human levels by around 2029. Follow that out further to, say, 2045 we will have multiplied the intelligence, the human biological machine intelligence of our civilization a billion-fold.”
— Ray Kurzweil
Welcome to another edition of the Be10x AI Pulse.
Meta finished its expensive hiring spree and immediately dissolved entire AI teams, Google claimed Gemini uses just "five drops of water" per query (researchers disagree), and Elon Musk declared Grok 5 has "a shot at being true AGI." From corporate chaos to environmental spin to bold predictions, AI companies are moving fast and breaking things, sometimes quite literally.
Let's dive into the developments that are reshaping the industry from the inside out.

META

Source: FilmMagic/Getty Images
Meta is undergoing a sweeping restructure of its artificial intelligence operations, dissolving its entire AGI Foundations division and reorganizing into four new units under the leadership of Alexandr Wang. The company has also imposed a hiring freeze across its AI division after completing an expensive talent acquisition spree.
Wang distributed an internal memo outlining the new organizational structure, which includes dedicated teams for research, training, products, and infrastructure. Most division heads now report directly to him, including Chief Scientist Yann LeCun, who previously operated with more independence through the FAIR research lab. The restructure also creates a mysterious "TBD Lab" focused on exploring "omni" models and frontier AI research.
The hiring freeze represents a dramatic shift from Meta's aggressive summer recruiting campaign, during which the company poached high-profile researchers from competitors with multi-million dollar offers. Any new hires now require Wang's personal approval, signaling a move from expansion to consolidation as the newly assembled team focuses on execution over acquisition.

Source: Google Blog
Google published new research detailing Gemini's environmental footprint, claiming each chatbot query consumes energy equivalent to watching TV for nine seconds and uses only five drops of water. The company says Gemini became 33 times more energy efficient while cutting carbon emissions by 44 times over the past year.
According to Google's analysis, a single Gemini text request consumes 0.24 wh of energy — slightly less than the 0.34 wh that Sam Altman previously revealed for ChatGPT queries. The study positions Google's AI as increasingly sustainable even as capabilities expand, presenting what appears to be significant environmental progress.
However, researchers have criticized the methodology for excluding water consumed by power plants that generate electricity for data centers, which represents the majority of actual water usage in AI operations. The selective accounting may be painting an artificially optimistic picture of AI's environmental impact, highlighting the need for industry-wide standardized measurements to provide accurate assessments of the technology's true ecological footprint.
XAI & ELON MUSK

Source: X/Twitter
Elon Musk made several bold AI predictions during a series of X posts, revealing that xAI's Grok 5 model will begin training in September and claiming it "has a shot at being true AGI." He also asserted that the upcoming Grok Imagine video generator will surpass Google's VEO 3 model "in every respect, with no exceptions."
Beyond technical capabilities, Musk addressed AI's societal implications, suggesting that artificial intelligence will actually increase birth rates because it will be "programmed that way." The comments came during a broader discussion about declining global birth rates and AI's role in addressing demographic challenges.
The AGI declaration is particularly significant given the ongoing competition between major AI labs to achieve artificial general intelligence. With OpenAI also hinting that its models are approaching AGI-level capabilities, Musk's claim sets up what could become the next major rivalry between former business partners, especially since there's no universally agreed-upon definition of what constitutes true AGI.
MICROSOFT

Source: Microsoft
Microsoft is rolling out a new COPILOT function that brings AI assistance directly into Excel cells, allowing users to generate summaries, classify data, and create tables using natural language prompts. The feature integrates seamlessly with existing formulas and automatically updates results as underlying data changes.
Powered by OpenAI's GPT-4.1 Mini model, the COPILOT function operates entirely within Excel without accessing external web data or company documents, keeping inputs confidential. Users can simply type natural language requests like "summarize sales trends" or "classify customer feedback" and watch AI generate relevant insights in real-time.
Microsoft has cautioned against using the feature in high-stakes scenarios due to potential inaccuracies and current capacity limitations. The tool is currently available to Microsoft 365 Beta Channel users, with a broader rollout planned for Frontier program web users. The development represents a significant step toward making AI accessible to the millions of professionals who rely on spreadsheets daily.


Google's enhanced search experience that combines traditional results with AI-powered conversations. The new agentic abilities can handle complex tasks like making restaurant reservations, booking appointments, and multi-step research queries, transforming search from information retrieval into action execution.

Runway's beta tool for creating interactive text-based gaming experiences through natural language prompts. Build, explore, and play games that generate storylines, characters, and scenarios dynamically, offering endless possibilities for interactive entertainment and creative storytelling.

When a railway employee's son from Chennai, India, set out to create an AI that could provide accurate answers with verifiable sources, he had no idea his company would become a "unicorn" with a valuation of over $1 billion.
The video explores the journey of Arvin Srinivas, co-founder of the AI company Perplexity. It highlights how his vision, born from the frustrations he faced during his own academic research, led to the creation of a tool that is now challenging major tech giants. The video also showcases how his success proves that a brilliant idea and hard work can disrupt an entire industry.

Learn to use AI Tools & ChatGPT. Create any kind of Presentations under 10 secs, do any kind of IT work under 10 mins, become top 1% of excel users who can use the functions and save your job and 90% of your time on daily basis.
✅ Become a highly paid prompt engineer (Average Salary 17LPA)
✅ No prior Technical or AI knowledge required
✅ Save upto 2 Hours Everyday
✅ Grow your salary upto 3X

This week, we found a few interesting things for you to check out:
Cohere launched Command Reasoning, a new enterprise model that outperforms competitors like gpt-oss and DeepSeek R1 on agentic benchmarks, targeting business automation use cases.
ByteDance released Seed-OSS, an open-source reasoning model family with 500k+ token context windows and strong benchmark performance for long-document processing.
Google and the GSA announced a new deal offering Gemini to government agencies at just $0.50 per agency to accelerate federal AI adoption across departments.
Chinese firms are moving away from Nvidia's H20 chips toward domestic alternatives after being "insulted" by comments from U.S. Commerce Secretary Howard Lutnick.
Thankyou for reading!
We love crafting every week’s issue for you, and would only want to serve the best content to you — so, if you could take out 2 minutes to tell us how we did, it would help us improve the Be10x AI Pulse experience for you! :D
See you later,
The Be10x Editorial Team