OpenAI Foundry: A New Platform for Running Advanced Machine Learning Models
OpenAI is a non-profit organization that conducts research in the field of artificial intelligence (AI) with the goal of creating friendly AI. They recently launched a new developer platform called Foundry, which enables customers to run their machine learning models on dedicated capacity. In simple terms, it allows users to use the latest AI models like GPT-3.5 with a larger context window, which can generate more accurate results.
Foundry is intended for customers who require a lot of processing power to run their AI models. It is designed for companies with larger workloads, allowing them to monitor their model’s performance with the same tools and dashboards that OpenAI uses. Customers also have control over the exact snapshot of their model, so they can choose to update to the latest model or not.
However, Foundry is not cheap. Rentals are based on 3-month or 1-year commitments, with prices starting at $78,000 for the lightweight version of GPT-3.5. This is because running the latest AI models can cost millions of dollars due to the high cost of processing power required to train and run them.
The launch of Foundry is part of OpenAI’s efforts to make their research more accessible to businesses and monetize their work. The company aims to generate $200 million in revenue in 2023, which is relatively small compared to the billions of dollars they have received in investments so far. Foundry will provide OpenAI with a new revenue stream by allowing businesses to access their latest AI models, while customers will benefit from the ability to use cutting-edge technology to enhance their operations.
