Years before the impressive human-like text capabilities of ChatGPT and other generative AI systems, IBM's Watson captivated audiences by dominating human competitors on Jeopardy! While IBM's cognitive computing project had grand ambitions—such as assisting doctors in diagnosing patients and recommending treatments—those aspirations didn't materialize as planned.
Now, IBM is rebranding its supercomputer platform as Watsonx, an AI development studio equipped with foundational and open-source models for companies to train their own AI systems. This initiative aligns with NVIDIA's recent launch of its AI Foundations program, both aiming to empower enterprises to build, train, scale, and deploy tailored AI solutions.
Watsonx offers a comprehensive array of training models with auditable data lineage, featuring datasets that support code generation for developers, manage industry-specific databases, and assist in climate-related decisions. A notable example of this platform in action is its collaboration with NASA, where a geospatial foundation model was employed to transform satellite images into maps that monitor changes related to natural disasters and climate change.
While reimagining Watson as an AI development studio may not carry the same allure as a supercomputer that dazzles on television, it opens up new possibilities for businesses and individuals alike. As companies leverage IBM's new AI training program, you might find yourself interacting with Watson's innovations in the near future. Watsonx will roll out in stages, beginning with the Watsonx.ai studio in July, followed by additional features later this year.