Bloomberg Tech→ original

Anthropic Computer Use: как одна кнопка стерла $285 миллиардов капитализации

Anthropic запустила функцию Computer Use для модели Claude 3.5 Sonnet. Теперь ИИ может двигать курсором, нажимать кнопки и заполнять формы в любых программах. Т

AI-processed from Bloomberg Tech; edited by Hamidun News
Anthropic Computer Use: как одна кнопка стерла $285 миллиардов капитализации
Source: Bloomberg Tech. Collage: Hamidun News.
◐ Listen to article

Tuesday was a bloody day on world markets, and the culprit wasn't the Federal Reserve or geopolitics, but a small button on the Anthropic website. When the company unveiled the Computer Use feature for its Claude 3.5 Sonnet model, the industry shuddered. We're used to neural networks writing text and drawing pictures, but Anthropic went further and gave AI "hands." Now the model can see the screen, move the cursor, click buttons, and enter text into any applications the way humans do. The market's reaction was immediate: shares of software and financial sector giants plummeted, erasing 285 billion dollars of market capitalization in mere hours.

To understand the scale of the disaster for traditional business, you need to look at who exactly suffered. Experian, RELX, and London Stock Exchange Group took the hit. These are companies that have spent decades building empires on providing complex data and specialized software. Their business model largely relies on thousands of people performing structured but monotonous tasks within their proprietary systems. Investors suddenly realized: if Claude can independently access any database, fill out forms, and compile reports, then an enormous layer of paid intermediary services becomes simply unnecessary. This isn't just a software update; it's a direct threat to the existence of entire industries.

Previously, automating work processes required companies to hire programmers to write complex API integrations. It was expensive, time-consuming, and often impossible due to the opacity of legacy systems. Anthropic elegantly bypassed this problem. Claude 3.5 Sonnet doesn't need permission from software developers to work within their programs. It simply "looks" at the interface and understands where to click. This transforms AI from a smart advisor into a full-fledged digital employee who can work in Excel, CRM systems, or banking terminals without any preparation. Traders saw in this the beginning of the end for the era of classic SaaS, where the value of a product lay in its user-friendly interface for humans.

The market panic highlighted a fundamental shift in how artificial intelligence is perceived. We've moved from the stage of "wow, it can tell jokes" to the stage of "damn, it can take my business." Bloomberg Intelligence analysts noted that the selloff began before American trading even opened, as soon as the description of the new feature appeared on Anthropic's blog. This shows that markets now monitor AI lab releases more carefully than they follow central bank reports. Any company whose added value consists of "moving data from one window to another" is now in the zone of maximum risk.

Interestingly, Anthropic positions Computer Use as a tool for developers in the beta testing stage. The model can still make mistakes, confuse icons, or "hallucinate" clicks. However, this didn't reassure investors. They understand that the pace of technological development is such that today's errors will become flawless algorithms tomorrow. If Claude can now fill out credit applications in Experian, then in six months it will be managing entire back-office departments. The only question is who will manage to adapt to this new world, and who will remain in history as a victim of the "October massacre of software stocks."

What happened is a clear signal to the entire technology sector. It used to be thought that AI threatened only creative professions or copywriters. Now it's become clear that companies with market capitalizations in the hundreds of billions of dollars, long considered a "safe haven," are under attack. Anthropic has shown that the interface is no longer a barrier for AI. This is the beginning of a new era of agentic intelligence, where software incapable of working directly with neural networks or replaceable by them is doomed to extinction. The market has voted with its dollars, and that vote sounds like a death sentence for the old school of IT services.

The bottom line: Anthropic has proven that there are no more "protected" sectors. If your business is based on people clicking buttons in a browser, it's time to urgently find a new strategy.

ZK
Hamidun News
AI news without noise. Daily editorial selection from 400+ sources. A product by Zhemal Khamidun, Head of AI at Alpina Digital.

Want to stop reading about AI and start using it?

AI News is a curated feed of AI/tech news. Hamidun Academy teaches you to use AI systematically in your work.

What do you think?
Loading comments…