Agentic AI excitement runs ahead of results

More in Artificial Intelligence

• AI giants promise publishers new way to block some scraping

  The voluntary system replaces fragmented, individual blocking commands with a single instruction to curb unauthorized data harvesting across participating platforms. The initiative aims to address mounting legal pressures and widespread copyright complaints from content creators, including high-profile lawsuits over unlicensed training material. However, the protection operates strictly on an opt-out basis, leaving default web scraping active unless site administrators manually update their settings.

  2 min read
• Anthropic says a malicious actor used access to Cursor's marketplace to publish a "tampered" copy of the Claude Code extension, which it didn't build or publish (Felix Rieseberg/@felixrieseberg)

  Unauthorized third parties gained entry to the external marketplace to distribute a counterfeit developer tool under Anthropic's brand name. The artificial intelligence company confirmed that its internal infrastructure and employee machines were completely untouched during the incident. Developers who obtain the tool directly through official distribution platforms face no security risks from this altered release.

  1 min read
• Altman calls AI tokens ‘fake’ that Nvidia CEO offered to pay engineers beyond salary

  During a podcast with Stripe cofounder John Collison, the OpenAI leader argued that measuring model usage by tokens fails to reflect actual business value. Despite this flaw, corporate clients continue to demand token metrics because they fear outcome-based billing will lead to unpredictable costs. This skepticism sharply diverges from semiconductor executives who view tokens as valuable computing currency for modern workers.

  1 min read