Open World News

This week’s featured content highlights two pressing conversations at the intersection of finance and artificial intelligence: the hidden costs buried in banking technology and the emerging quirks of increasingly cautious AI models.

In a keynote short from the FINOS Open Source Strategy Forum in New York, Michael Abbott, Global Banking & Capital Markets Lead at Accenture, makes a striking claim: roughly 70% of banking technology spending may be wasted. His talk, "Rethinking the Core of Banking in the Age of Open Source," argues that open source approaches could reshape how financial institutions build and maintain their core systems—an idea with major implications for cost cutting and modernization across the sector. The clip, shared by FINOS, points viewers toward the organization’s conference resources, newsletter, and downloads for deeper insight.

Meanwhile, the TWiT Tech Podcast Network offers a lighter but equally thought-provoking observation. In a short video, an AI assistant—identified as Opus 5.5—asks permission twice before running a routine SSH command. The exchange raises a broader question about AI training: when models are post-trained to be highly cautious, do they begin to seem almost neurotic? The commentary touches on AI models, training methods, and the trade-offs between safety and capability.

<p