|
|
|
25.04.26 - 14:12
|
What the Fed Can Do About Anthropic′s Latest System (Bloomberg)
|
|
|
Anthropic says its newest AI system can detect serious cybersecurity risks, but the new technology is so powerful that the company is holding back public release. The nearly autonomous system can detect banking software vulnerabilities and spawn sub-agents that operate without human oversight at unprecedented speed. Treasury Secretary Scott Bessent and Fed Chair Jay Powell took the unusual step of calling in major bank leaders to discuss the risks raised by the new model. Experts weigh in on whether regulation and open-source tools can keep the financial system safe. (Source: Bloomberg)...
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
25.04.26 - 11:01
|
Understanding the Most Viral Chart in Artificial Intelligence | Odd Lots (Bloomberg)
|
|
|
METR, which stands for Model Evaluation and Threat Researc, is focused on understanding the degree to which AI models can engage in autonomous, complex tasks. METR see this is as a particularly important benchmark, given the risk that AI could one day be engaged in recursive self improvement, taking humans out of the loop. But how do you really gauge a model's ability to do complex problems. And what is being measured for exactly? On this episode we speak with METR's President Chris Painter as well as Joel Becker, a member of the technical staff who works on evaluation methods for the organization. We discuss both the mechanics and the philosophy of METR's work, and what it means when we see a a chart showing that Clause Opus 4.6 can do a task that would take a human nearly 12 hours.
(Source: Bloomberg)...
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|