OpenAI has launched GPT-4.5 with a cost of $75 per million tokens, markedly higher than previous models. This price reflects both the model's advanced capabilities and its extensive resource requirements, leading to a shortage of GPUs that has delayed rollout. CEO Sam Altman acknowledged the difficulty in predicting growth surges that strain infrastructure, prompting OpenAI to invest heavily in additional GPUs. Despite the expensive model, Altman cautioned users about its limitations, stating it is not designed primarily for reasoning tasks, highlighting ongoing concerns in AI development, such as hallucinations and cybersecurity.
The introduction of GPT-4.5 highlights OpenAI's revenue strategy, reflecting the significant resource demands and market competition for advanced AI hardware.
OpenAI's rollout of GPT-4.5 is hindered by a shortage of GPUs, forcing the company to stagger access and manage expectations around its capabilities.
Altman acknowledged that the company's growth surges have led to GPU shortages, illustrating the logistical challenges faced in scaling AI technology effectively.
Despite high costs for GPT-4.5 usage, Altman cautioned users on its limitations, noting that it does not excel in reasoning tasks.
Collection
[
|
...
]