With the introduction of the new Watsonx platform, IBM is rebranding Watson, an AI system that has been around for over 20 years.
Watsonx, previewed at the IBM Think 2023 conference on Tuesday, is a new AI and data platform for underlying model and generative AI.
IBM Watsonx
This platform includes Watsonx.ai, Watsonx.data and Watsonx.governance.
Watsonx.ai is an enterprise studio that enables AI builders to train, test, and deploy generative AI capabilities powered by underlying models.
Scheduled for general availability in July, the studio also includes a model library that includes IBM’s pre-trained foundational models. Foundation models currently available in beta preview include fm.code, fm.NLP, and fm.geospatial.
Fm.code is a set of models that automatically generate code for developers. Fm.NLP is a suite of large language models for specific industries. fm.geospatial also provides models built on climate and remote sensing data to help organizations learn more about natural disaster patterns, biodiversity, land use, and geophysical processes.
Watsonx.ai studio is built on open source libraries and offers thousands of open models and datasets from generative AI vendor Hugging Face.
Watsonx.data, a data store built on an open lakehouse architecture for governed data and AI workloads, will be available in July, according to IBM. Watsonx.governance is a toolkit aimed at mitigating AI-related risks and protecting customer privacy. It will be generally available later this year.
Thanks to Watsonx, the 111-year-old technology vendor joins young tech giants such as Google, Microsoft and AWS, as well as independent AI hardware/software vendors such as Nvidia and SambaNova, into the burgeoning generative AI market can.
IBM differentiator
Futurum Group analyst Daniel Neumann said that while IBM appears to be lagging behind the fast-growing market, it’s not.
“This is what I call the next wave [in generative AI]’” said Neumann, noting that IBM provides enterprises with an AI platform that focuses on security and data privacy.
“The launches so far have mostly been user, consumer, social tools and some productivity apps,” he said. Watsonx is different because it focuses exclusively on enterprises, Newman added.
Moreover, even as major cloud providers push out enterprise AI offerings, the adoption of generative AI is still in its early stages, said Gartner analyst Arun Chandrasekaran.
“More and more enterprise customers want to customize generative AI models for their own use cases,” he said. “IBM is betting on the fact that multiple models will be used, but customers will want a consistent set of tools to operate them.”
Neumann said the AI vendor is also trying to differentiate itself by focusing on AI and hybrid cloud.
He said that using Red Hat, an open source software vendor it acquired in 2018, the vendor is appealing to companies building data governance systems for compliance on hybrid infrastructures of cloud and on-premises. I added that there is.
“We argue that choosing a hybrid cloud provides 2.5 times more value than choosing a single answer from one of the underlying landscape options.” IBM CEO Arvind Krishna said in a streaming live keynote at the conference on May 9.
IBM and Red Hat also unveiled a joint effort to make coding easier for developers with Watson Code Assistant. Code Assistant allows developers to generate code using accessible English commands.
For businesses, meanwhile, IBM’s services and consulting business will help service its underlying model, Neumann said.
IBM’s partnerships with big consulting firms like Accenture will also help deploy AI models at scale, he said.
“Companies have a lot of data and will want to build very specific models on top of the various basic tools offered as part of Watsonx,” Neumann said. “And those models are going to be very specialized in different areas,” Neumann said. “We’re going to need a lot of help.”
Some problems
Watsonx comes with challenges. Chandrasekaran said the main hurdle is achieving speed to market.
“IBM needs to get more mindshare in the generative AI space,” he said. “IBM’s research is a key innovator in AI, but IBM needs to accelerate the pace of commercializing its research capabilities.”
Another key challenge is competition, Neumann said. Many vendors, especially cloud giants Google and Microsoft, are vying to be the leading providers of what businesses need in generative AI technology.
IBM’s strength lies in its diverse customer base, Neumann added, and strong consulting and deployment could give the company a solid footing in the hybrid cloud world.
Esther Ajao is a news writer covering artificial intelligence software and systems.
