[ad_1]
OpenAI’s next major model may not represent as big a leap forward as its predecessors, according to New report In the information.
Staff who tested the new model, codenamed Orion, reportedly found that although its performance exceeds existing OpenAI models, there was less improvement than they saw in jumping from GPT-3 to GPT-4.
In other words, the rate of improvement appears to be slowing down. In fact, Orion might be no To be reliably better than previous models in some areas, such as programming.
In response, OpenAI created a core team to figure out how the company could continue to improve its models in the face of a dwindling supply of new training data. These new strategies reportedly include training Orion on synthetic data produced by the AI models, as well as doing further improved models during the post-training process.
OpenAI did not immediately respond to a request for comment. In response to previous reports about plans for its flagship model, the company said: “We have no plans to release a model codenamed Orion this year.”
[ad_2]