Katya clicked and was taken to a page for another company, called Mercor, where she was instructed to interview on-camera with an AI named Melvin. “It just seemed like the sketchiest thing in the world,” Katya says. She closed the tab. But a few weeks later, still unemployed, she got a message inviting her to apply to Mercor. This time, she looked up the company. Mercor, it seemed, sold data to train AI, and she was being recruited to create that data. “My job is gone because of ChatGPT, and I was being invited to train the model to do the worst version of it imaginable,” she says. The idea depressed her. But her financial situation was increasingly dire, and she had to find a new place to live in a hurry, so she turned on her webcam and said “hello” to Melvin.
Nobody starts in the right place. You don’t begin with the correct tool and work sensibly within its constraints until you organically graduate to a more capable one. That is not how obsession works. Obsession works by taking whatever is available and pressing on it until it either breaks or reveals something. The machine’s limits become a map of the territory. You learn what computing actually costs by paying too much of it on hardware that can barely afford it.，更多细节参见使用 WeChat 網頁版
Two subtle ways agents can implicitly negatively affect the benchmark results but wouldn’t be considered cheating/gaming it are a) implementing a form of caching so the benchmark tests are not independent and b) launching benchmarks in parallel on the same system. I eventually added AGENTS.md rules to ideally prevent both. ↩︎，详情可参考谷歌
Opens in a new window