Google's WebMCP moves the web closer to becoming a structured database for AI agents
In the future, AI agents won’t just search the web; they’ll browse it, shop on it, and complete tasks on their own. At least that’s Big AI’s vision, and Google’s WebMCP wants to turn websites into standardized interfaces for these agents. For website operators who depend on human visitors, that could be a serious problem.
Read full article about: xAI's founder exodus reportedly tied to safety concerns and frustration over Grok's failure to catch up
xAI has lost half of its founders in recent weeks and months. Elon Musk said on X that some departures were part of a restructuring where "unfortunately we had to part with some people" to "improve speed of execution."
But former employees tell a different story. One ex-employee told The Verge that many people at the company had grown disillusioned with Grok's focus on NSFW content and its lack of safety standards. A second former employee backed that up: "There is zero safety whatsoever in the company." According to the source, Musk deliberately pushed to make the model less restricted, viewing safety measures as censorship. Among other things, Grok had generated sexualized images of children.
You survive by shutting up and doing what Elon wants.
Another common complaint is that xAI is "stuck in the catch-up phase" without shipping anything fundamentally new compared to OpenAI or Anthropic, even though they're all trying to do the same thing anyway. Several people who left are now using money from the SpaceX merger to start their own companies, including AI infrastructure startup Nuraline.
Comment
Source: The Verge | Musk via X
Ad
Anthropic recruits ex-Google data center veterans to build its own AI infrastructure empire
Anthropic is discussing building at least 10 gigawatts of data center capacity worth hundreds of billions of dollars, recruiting ex-Google managers and lining up Google as a financial backer to make it happen.
An AI agent got its code rejected so it wrote a hit piece about the developer
After a volunteer developer rejected its code, an autonomous AI agent independently researched his background and published a hit piece attacking his character. The incident at Matplotlib shows how theoretical AI safety risks are becoming real.
Read full article about: Anthropic raises $30 billion, pushing valuation to $380 billion
Anthropic has closed a $30 billion Series G funding round, bringing the AI company's post-money valuation to $380 billion.
The round was led by GIC, Singapore's sovereign wealth fund, and U.S. investment firm Coatue. D. E. Shaw Ventures, Dragoneer, Founders Fund, ICONIQ, and MGX, an Abu Dhabi-based technology investment fund, joined as co-leads. Microsoft and Nvidia also participated, building on previously announced strategic partnerships. Anthropic says it will use the capital for research, product development, and infrastructure expansion.
Anthropic reports annualized revenue of $14 billion, having grown more than tenfold in each of the past three years. Claude Code, the company's coding tool, now accounts for over $2.5 billion in annualized revenue on its own.
One notable detail about how companies are using AI: more than 500 customers spend over $1 million per year on Claude, according to Anthropic, and eight of the ten largest Fortune companies are among its users.
Ad
Read full article about: Microsoft AI CEO: "Most" white-collar tasks will be automated in 18 months
Microsoft AI CEO Mustafa Suleyman predicts the end of traditional white-collar work in 18 months.
"I think that we're going to have a human-level performance on most, if not all, professional tasks," Suleyman says in an interview with the Financial Times. "So white-collar work, where you're sitting down at a computer, either being a lawyer or an accountant or a project manager or a marketing person — most of those tasks will be fully automated by an AI within the next 12 to 18 months."
Suleyman leads Microsoft's AI division, which has invested billions in OpenAI and Anthropic and operates Copilot, one of the most widely used AI work tools. He describes the shift as already underway: In software engineering, developers are already using "AI-assisted coding for the vast majority of their code production."
Anthropic CEO Dario Amodei even predicted that half of entry-level office jobs could disappear within one to five years. He is already observing that fewer junior and mid-level employees are needed. AI could be better than humans in many areas within one to two years, while the labor market adapts with a delay.
Suleyman's boss, Microsoft CEO Satya Nadella, on the other hand, sees more of a shift where existing cognitive tasks might be automated, but new, more demanding tasks would emerge.
OpenAI has yet another new coding model and this time it's really fast
OpenAI’s new GPT-5.3-Codex-Spark is a smaller coding model that runs on Cerebras chips and pushes over 1,000 tokens per second. It’s the company’s first model built specifically for real-time programming.
Ad
Read full article about: OpenAI is retiring GPT-4o and three other legacy models tomorrow, likely for good
OpenAI is dropping several older AI models from ChatGPT on February 13, 2026: GPT-4o, GPT-4.1, GPT-4.1 mini, and o4-mini. The models will stick around in the API for now. The company says it comes down to usage: only 0.1 percent of users still pick GPT-4o on any given day.
There's a reason OpenAI is being so careful about GPT-4o specifically: the model has a complicated past. OpenAI already killed it once back in August 2025, only to bring it back for paying subscribers after users pushed back hard. Some people had grown genuinely attached to the model, which was known for its complacent, people-pleasing communication style. OpenAI addresses this head-on at the end of the post:
We know that losing access to GPT‑4o will feel frustrating for some users, and we didn’t make this decision lightly. Retiring models is never easy, but it allows us to focus on improving the models most people use today.
OpenAI
OpenAI points to GPT-5.1 and GPT-5.2 as improved successors that incorporate feedback from GPT-4o users. People can now tweak ChatGPT's tone and style, things like warmth and enthusiasm. But that probably won't be enough for the GPT-4o faithful.