OpenAI’s GPT-5 has reportedly failed to live up to expectations

Efforts to develop OpenAI’s next major model, GPT-5, are behind schedule, and the results still do not justify the huge costs. new report In The Wall Street Journal.

It resonates previous report In the information suggesting that OpenAI is looking for new strategies GPT-5 may not represent as big a leap forward as previous models. But the WSJ story contains additional details about the 18-month development of the GPT-5, codenamed Orion.

OpenAI has reportedly completed at least two large training runs, which aim to improve the model by training it on large amounts of data. The initial training run went slower than expected, signaling that a larger run would be both time-consuming and costly. Although GPT-5 may perform better than its predecessors, it is not yet advanced enough to justify the cost of keeping the model running.

The WSJ also reports that instead of relying solely on open data and licensing agreements, OpenAI has hired people to create new data by writing code or solving math problems. It also uses synthetic data created by another model, o1.

OpenAI did not immediately respond to a request for comment. The company said earlier It would not release a model codenamed Orion this year.