Sources
Loading...
Additional media
Loading...
DeepNewz, mobile.
People-sourced. AI-powered. Unbiased News.
The cost of intelligence, particularly in AI models like GPT-4, has significantly decreased. The price of GPT-4 equivalent intelligence has dropped approximately 240 times in 18 months, from $180 per million tokens to less than $1 per million tokens. This trend is driven by factors such as increased competition, particularly from open-source models. Additionally, inference pricing on GPT-4 models has dropped nearly 90% since its release. These reductions in cost are crucial for making AI more accessible and unlocking latent demand in various business models. Driving down compute costs is important to get AI into everyone's hands, despite the extreme cost of training GenAI models.
OpenAI has shared this chart a few times, but important to call out. Inference pricing on GPT4 family of models has dropped nearly 90% since it's release! This trend will continue. Business models that don't make sense now will in the future. Latent demand will be unlocked. https://t.co/2SWPxGuDPs
The extreme cost of training GenAI models! #AI #MachineLearning #DeepLearning #DataScience #GenerativeAI #LLM #LLMs #Python #Code #100DaysOfCode via @StatistaCharts @SpirosMargaris @PawlowskiMario @mvollmer1 @gvalan @ipfconline1 @LaurentAlaus @Shi4Tech @Fisher85M @kalydeoo… https://t.co/vBj0drQMky
Driving down compute costs seems super important to get AI into everyone's hands. https://t.co/1YjpOdEJK6
