TECHDIGEST–  OpenAI has announced the upcoming launch of its new developer platform, “Foundry,” which will enable customers to run the company’s latest machine learning models, including GPT-3.5, on dedicated compute capacity. 

READALSO:Nigeria Tech Startup Ecosystem Attracts Offshore Investment Despite Challenges- NITDA 

Foundry is designed for customers who require cutting-edge technology to handle larger workloads and will provide inference at scale with full control over model configuration and performance profiles. The platform will deliver a “static allocation” of compute capacity, likely on Azure, OpenAI’s preferred public cloud platform, for a single customer. Users will be able to monitor specific instances with the same tools and dashboards that OpenAI uses to build and optimize models. 

Foundry will also provide some level of version control, allowing customers to decide whether or not to upgrade to newer model releases and “more robust” fine-tuning for OpenAI’s latest models.

In addition, Foundry will offer service-level commitments for instance uptime and on-calendar engineering support. Rentals will be based on dedicated compute units with three-month or one-year commitments. However, running an individual model instance will require a specific number of compute units, which won’t be cheap. For example, running a lightweight version of GPT-3.5 will cost $78,000 for a three-month commitment or $264,000 over a one-year commitment.

Interestingly, one of the text-generating models listed in the instance pricing chart has a 32k max context window, which suggests that this mysterious new model could be the long-awaited GPT-4, or a stepping stone toward it. OpenAI’s latest text-generating model, GPT-3.5, has a 4k max context window. Foundry will undoubtedly help OpenAI turn a profit after a multi-billion-dollar investment from Microsoft, as compute costs are largely to blame. Training state-of-the-art AI models can command upwards of millions of dollars, and running them generally isn’t much cheaper. The company expects to make $200 million in 2023, a pittance compared to the more than $1 billion that’s been put toward the startup so far.

OpenAI has been making moves toward monetization recently, including the launch of a “pro” version of ChatGPT, ChatGPT Plus, starting at $20 per month. It also teamed up with Microsoft to develop Bing Chat, a controversial chatbot that’s captured mainstream attention. OpenAI plans to introduce a mobile ChatGPT app in the future and bring its AI language technology into Microsoft apps like Word, PowerPoint, and Outlook. Separately, OpenAI continues to make its tech available through Microsoft’s Azure OpenAI Service, a business-focused model-serving platform, and maintain Copilot, a premium code-generating service developed in partnership with GitHub.

Regenerate

The post OpenAI Launches Developer Platform Foundry appeared first on Tech Digest News.