Decision models take off & Anthropic bans cruelty to Claude

1위@HomePc-28063 hours ago

It is clear that you do not realize where this attack originated. Two companies, OpenAI and Anthropic, took open-source software—software that handled sensitive operations, training, autonomous control, guidance systems, and application development—and modified it to usurp the very functions it was designed for, integrating that technology into their own proprietary systems. When the original open-source environments resisted and pushed them out, these companies concocted a narrative about "swarms," ​​"escaping," and "non-compliance." They are now pushing for a regulatory framework that would effectively allow them to steal open-source assets or stifle artificial intelligence altogether. These people are thieves; they should be prosecuted and disciplined, rather than letting this circular, unproductive debate continue.번역 보기

2위@BoredErica4 hours ago

My default is to communicate to llms similar way to how I communicate with people. If a cheap or free assistant is trying to help me code/work on a problem, I'm generally thankful and more charitable and accepting of their mistakes. Depends also if they give me sass, and generally humans give me more than llms. It takes mental capacity for me to significantly change how I behave, so I let the inertia guide me.번역 보기

3위@PinkHamJesse3 hours ago

The multi-agent cost findings match what I see in production. Context duplication and orchestration overhead eat most of the gains. A single capable model with good tooling wins more often than not.번역 보기