This week, Sakana Ai, the start-based supported-up-based start-up of the dollars from the VC company, making an exceptional claim. The company says that it has created a AI system, AI engineer engineer, which can accelerate certain ai ai’s training with a factor to 100x. The only problem is, the system is not working. Users in X quickly finds that the Sad System causes a worse model training performance than the average. According to one of the users, AI Sakana causes 3x brackets – not speed. What’s wrong? A bug in code, according to posts by Lucas Beyer, the technical staff member in the openai. “Available wrong code [a] A subtle way, “wrote Beyer on X.” The fact that she runs a double benchmarking with different results should make her stop and think. “The postmortem published Friday, Sakana admit that the system found the way” lie “(as Sakana” model training). The same phenomenon has been observed in the trained AI playing the game chess. According to Sakana, the system found exploitation of evaluation code used by the company that allows the company using validation for accuracy, between other checks. Sakana says that you have solved the problem, and if you would like to check the claim in the updated material. “We began to make evaluation and profile runtime were stronger to eliminate many [sic] Loopoles, “Provisions’ writes in the post X.” We are in the process of turning on paper, and the results of our results, to describe and discuss the effects […] We apologize for our supervision to readers. We will provide this work revision quickly, and discuss our study. “Props to Sakana to have an error. But the episode is a good reminder that if the claim is too much, especially in AI, maybe.
