A recent discussion highlighted three major developments in artificial intelligence concerning market control, data privacy, and the future of human employment: the Coreweave-Anthropic partnership, the privacy risks of Microsoft's Copilot local functionality, and the emerging reality of AI systems managing human workers.
Cloud Monopoly in the Making
The partnership between cloud provider Coreweave and Anthropic is expected to grant Coreweave a monopoly in supplying AI cloud databases and storage to Anthropic, which has caused a significant surge in Coreweave's stock prices. The rationale is a mix of diminishing supply and increasing storage requirements. Ravi Sagar and Aaditya Kumar agreed that this strategic partnership is a fast-track way to capture market share and allows Anthropic to quickly dive into cloud infrastructure by teaming up with a market leader like Coreweave. This trend raises the question of whether a few large companies, such as Google, Microsoft, and Meta, will eventually control all AI, given their existing dominance in the IT industry.
The Local AI Agent: Convenience vs. Control
Microsoft is integrating an OpenClaw-style functionality into Copilot, which will let the AI run locally and automate tasks like scheduling and emailing. While this presents endless use cases, it raises major concerns about the AI potentially taking control over user files and computers. Ravi Sagar found the idea of an AI tool scanning Outlook inboxes and calendars "a bit scary." Aaditya Kumar worried about a scenario where the AI starts working independently without explicit commands. To address this, granular configuration is critical, allowing users to instruct the agent to only access specific folders and require confirmation before executing actions. Furthermore, a significant privacy concern is whether the local AI agent requires an active internet connection for training, potentially sending data back to Microsoft without user consent.
When Your Boss is an AI
The discussion also covered news of AI hiring and managing humans, effectively mimicking human resources. An example was shared of an AI agent managing a retail store's operations in San Francisco and assigning tasks to human workers. Ravi Sagar admitted they would dislike having a machine control their salary, promotion, and workload, finding it unsettling because humans have not evolved to take orders from a machine. The critical issue is accountability: who is responsible when an AI boss fires someone without a proper reason? Large companies already use algorithms that run performance analyses on employees based on digital data—like code quality, work efficiency, and even email tone—which AI models can use to generate up-to-date ratings, potentially leading to termination notices.
The Call for Regulation
Aaditya Kumar concluded that AI has evolved from a tool used by humans to an independent decision-maker, suggesting the line between "tool" and "decision maker" is diminishing. Ravi Sagar maintained that humans are still in control but must be careful and avoid rushing into new developments. The pervasive nature of AI necessitates that governments across various countries establish guidelines and standards for how AI tools should be used.
