OpenAI, the artificial intelligence research organization behind the popular language models GPT-3 and GPT-4, is facing significant challenges in developing its next major model, GPT-5. According to a recent report in The Wall Street Journal, the development of GPT-5, code-named Orion, is running behind schedule, with results that don't yet justify the enormous costs incurred.

This news comes on the heels of an earlier report in The Information, which suggested that OpenAI is exploring new strategies as GPT-5 may not represent as big a leap forward as previous models. The WSJ story provides additional insight into the 18-month development of GPT-5, revealing that the organization has completed at least two large training runs aimed at improving the model by training it on enormous quantities of data.

However, the initial training run went slower than expected, hinting that a larger run would be both time-consuming and costly. While GPT-5 can reportedly perform better than its predecessors, it hasn't yet advanced enough to justify the cost of keeping the model running. This raises concerns about the feasibility and effectiveness of OpenAI's approach to developing its next-generation language model.

In a departure from its traditional approach, OpenAI has also hired people to create fresh data by writing code or solving math problems. Additionally, the organization is using synthetic data created by another of its models, o1. This shift in strategy may indicate that OpenAI is struggling to find innovative ways to improve its models, or that it's exploring alternative approaches to overcome the challenges it's facing.

OpenAI did not immediately respond to a request for comment, but the company previously stated that it would not be releasing a model code-named Orion this year. This lack of transparency and communication from OpenAI only adds to the uncertainty surrounding the development of GPT-5.

The implications of OpenAI's struggles with GPT-5 are far-reaching. The organization's language models have been instrumental in driving innovation in areas like natural language processing, chatbots, and content generation. If GPT-5 fails to deliver significant advancements, it could have a ripple effect throughout the AI research community and the industries that rely on these technologies.

As the AI landscape continues to evolve, the challenges faced by OpenAI serve as a reminder that even the most advanced organizations can encounter setbacks. It will be interesting to see how OpenAI adapts to these challenges and whether it can overcome the hurdles to deliver a groundbreaking language model that meets the expectations of the AI community.