OpenAI has announced that, thanks to advancements from its latest model, GPT-4.5, it can now rebuild the GPT-4 model with only five to ten engineers—down from the hundreds originally involved. GPT-4.5 is described as “10x smarter” than GPT-4 and is considered the company’s most potent release to date. This efficiency gain enables significantly faster and easier retraining of earlier models. OpenAI has secured a record-breaking $40 billion in new funding, with $30 billion from SoftBank, to expand its compute infrastructure further. CEO Sam Altman emphasized that OpenAI is no longer “compute-constrained,” shifting focus to data efficiency and algorithmic innovation. ​