OpenAI's GPT-5 reportedly falling short of expectations | TechCrunch
Briefly

OpenAI's latest model, GPT-5, code-named Orion, is facing development delays, with some results not yet justifying significant costs and efforts put into it.
Reports suggest GPT-5 is progressing slowly, having undergone multiple training runs that highlight the challenges of managing development costs while seeking better performance.
OpenAI's strategy for GPT-5 involves not only using existing data but also employing new hires to generate original data, as well as utilizing synthetic data.
Despite being said to perform better than its predecessors, GPT-5's advancements remain insufficient to validate the expenses involved in its development and operation.
Read at TechCrunch
[
|
]