Sam Altman Afraid Of GPT-5 👀, 97,000 Layoffs & Counting 🧑‍💼, Microsoft’s Copilot Mode 🪄

97,000 Layoffs & Counting: Why AI Isn’t Killing Jobs—It’s Rewriting Them

Salesforce CEO Marc Benioff recently questioned a popular narrative: “I keep asking CEOs what AI they’re using for these big layoffs, and I don’t think they know”. At Salesforce, AI handles 85 % of customer inquiries and qualifies leads 40 % faster, but rather than replacing people, it frees them for higher‑value work. Still, Benioff cautions that today’s AI lacks full accuracy — “AIs can’t fact check” — underscoring the need for human oversight.

Yet, as Aura points out, the deeper disruption isn’t mass elimination but misalignment. AI quietly absorbs core tasks — writing, summarizing, coordinating — leaving job descriptions unchanged while functions shift. This “silent automation” is driving reorganizations: over 97,000 tech roles were cut in early 2025, even as companies like Microsoft and Meta hired AI‑savvy talent.

The takeaway? AI is reorganizing work, not simply erasing it. Future‑proof careers will blend domain expertise with AI fluency, think prompt engineers, AI project managers, or data storytellers. For workers and solopreneurs alike, the path forward is clear: pair human judgment with smart use of AI to stay competitive.

💬 OpenAI Developments

Sam Altman is afraid of OpenAI's GPT-5 creation — "The Manhattan Project feels very fast, like there are no adults in the room"

OpenAI CEO Sam Altman recently admitted that during testing, the upcoming GPT‑5 model felt so fast and powerful that it made him feel “scared,” even likening its rapid development to the Manhattan Project—questioning “What have we done?” and warning that “there are no adults in the room” with respect to oversight . He emphasized that while GPT‑5 could mark a major leap toward AGI, its capabilities and pace of release raise serious concerns about governance, control, and the societal implications of deploying such transformative AI.

🚀 Tech Industry Moves

NVIDIA’s newly released Llama Nemotron Super v1.5 (49B parameters) tops the Artificial Analysis Intelligence Index in the ~70B open-model category, delivering best-in-class accuracy on advanced reasoning tasks like math, coding, science, function calling, and instruction following while offering up to 3× faster inference throughput and efficient deployment on a single NVIDIA H100 GPU . The model is trained using over 26 million high-quality synthetic examples with post‑training pipelines—reward‑aware preference optimization (RPO), direct preference optimization (DPO), and RLVR—for transparent, high-performance reasoning agent behavior, with full weights and datasets available via Hugging Face and NVIDIA Build.

Anthropic’s AI models are now the most widely adopted by businesses, capturing 32% of the enterprise AI market, while OpenAI holds 25%, according to a Menlo Ventures analysis referenced via TechCrunch . While OpenAI led the enterprise segment in 2023 with around 50% market share (compared to Anthropic’s 12%), Anthropic now dominates in corporate coding tools with 42% market share—double OpenAI’s 21%—driven in part by its Claude 3.5 Sonnet (June 2024) and Claude 3.7 Sonnet (February 2025) releases.

Anthropic's cofounder says 'dumb questions' are the key to unlocking breakthroughs in AI. At a recent Y Combinator event, Jared Kaplan—Anthropic’s cofounder and CSO—strongly advocated for asking simple, even “naive” or “dumb” questions (such as How much data is really needed for AI training?) to unlock foundational breakthroughs in the field, a mindset behind the discovery of scaling laws that reveal how model size, compute, and performance interrelate . He credits his physics training with this approach, noting that despite AI’s rapid progress, “a lot of the most basic questions haven’t been answered”—and by exploring them, Anthropic made strides toward more precise, predictive training dynamics.

Meta Platforms delivered a standout Q2 2025 performance, reporting revenue of approximately $47.5 billion (a 22% year-over-year increase) and earnings per share of $7.14, significantly above analysts’ projections . Much of the upside stemmed from AI-enhanced advertising tools boosting conversion rates, with CEO Mark Zuckerberg emphasizing the company’s strategic push toward “personal superintelligence”—including investment in new AI labs and smart glasses integration.

Google’s newly introduced AI Mode offers an interactive, chatbot-style search experience, designed to handle complex, multi-part queries through generative AI provided within the Search tab—moving away from traditional blue-link results toward conversational, synthesized answers . Marketers warn that this shift could disrupt established SEO and paid-search strategies, as brands must now optimize for presence and structured, conversational content that aligns with AI Mode’s evolving citation and output dynamics.

💡The Builder’s Brief

AWS cloud + AI slash IT costs

WebProNews reported that small businesses are using AWS to cut IT spending by up to 30 % by scaling resources and using automated tools. At the 2025 AWS Summit New York, Amazon announced Bedrock AgentCore features that let even small companies deploy AI agents to automate tasks like inventory management and customer service, reducing labor costs and errors. The article also notes that Amazon S3 Intelligent‑Tiering can reduce storage costs by 40 %+ and that AWS is investing $100 million in “agentic AI” tools for smarter data governance. Solopreneurs often operate on thin margins; learning to use scalable cloud services and AI agents helps them reduce overhead without hiring staff. AI‑driven automation can handle repetitive tasks such as inventory tracking or support, freeing founders to focus on growth.

AI adoption goes mainstream

The Times Square Chronicles summarized Stanford University’s 2025 AI Index. It found that U.S. businesses invested $109 billion in AI in 2024 and that nearly eight out of ten companies now use AI, up from just over half the year before. AI costs have fallen by over 280× in two years, making advanced models affordable for small firms. The report highlighted widespread adoption across industries, from hospitals to autonomous taxis, and warned that companies slow to adopt AI risk being left behind. Cheap and widely available AI means solopreneurs can now access tools previously reserved for large enterprises. Staying current with AI trends is vital to remain competitive.

Building a company of one

In Entrepreneur, Sarah Choudhary wrote about building a lean business without investors or co‑founders. She uses Notion as a virtual COO, Zapier to connect apps, Canva and ChatGPT for design and content, and tools like Stripe, Google Workspace and Calendly to handle payments, communications and scheduling. By automating admin tasks and hiring offshore experts for technical work, she focuses only on high‑impact tasks. This illustrates how solopreneurs can replicate the capabilities of a larger company by using AI and automation instead of staff, allowing them to scale on their own terms.

💡 AI Tool Deep Dive

Microsoft’s Copilot Mode, released in late July 2025, is more than a browser update, it’s an AI‑powered work companion. Copilot doesn’t just help you search; it summarizes articles, pulls insights from multiple tabs, drafts emails, and even generates action lists based on what you’re working on.

Picture this: you’re preparing a pitch. Instead of juggling endless tabs, Copilot gathers the key details, organizes them, and even suggests content for your presentation. It transforms browsing into a streamlined workflow hub, cutting hours of research and admin into minutes.

The takeaway? Copilot Mode is built for those wearing many hats, the ones juggling research, planning, and execution all at once. It brings organization, speed, and clarity to everyday online work, making it easier to focus on the tasks that actually move the needle.

👥 Connect & Feedback!

👉 Join Us:

📧 Advertise In Weekly AI News:

📧 Contact directly at [email protected]

😍 Share with your friends!