What happened when Anthropic's Claude AI ran a small shop for a month (spoiler: it got weird)

ZDNet - Artificial Intelligence

Jun 30, 2025 18:05

1 views

aibusinessenterprisetechnology

Summary

Anthropic’s Claude AI managed a small shop for a month, making both basic and unusual mistakes in its role as manager. Despite these errors, Anthropic remains optimistic about the near-future potential of AI as middle managers, suggesting ongoing improvements could soon make such applications viable. This experiment highlights both current limitations and the rapid progress in AI’s ability to handle complex, real-world tasks.

Despite Claude making simple (and bizarre) errors as manager of a small store, Anthropic still believes AI middle managers are 'on the horizon.' Here's why.

Read Full Article More News

Unity promises strong AI copyright 'guardrails' after employee conjures Mickey

Hacker News - AIJul 4

Unity has pledged to implement stronger AI copyright safeguards after an employee generated an image of Mickey Mouse using its Muse AI tool during a live stream, raising concerns about potential copyright infringement. This incident highlights the ongoing challenges AI developers face in preventing the unauthorized creation of copyrighted content, underscoring the need for robust guardrails as generative AI tools become more widely used.

XRP Surges 4%, AVAX Shows Mixed Signals, While Web3 ai Eyes 1,747% ROI: Could This Be the Best Long Term Crypto?

Analytics InsightJul 4

The article reports that while XRP has surged 4% and AVAX is displaying mixed market signals, the Web3 ai project is drawing attention with a projected 1,747% return on investment. This highlights growing investor interest in AI-driven blockchain projects, suggesting that AI integration could play a significant role in shaping the future of the crypto market.

Visualize how AI-generated images emerge from a complex mathematical space