Watsonx.ai Positions IBM in Generative AI for the Enterprise
Discover more from AI in Business
Watsonx.ai Positions IBM in Generative AI for the Enterprise
Building on a heritage in AI earned by winning Jeopardy!, IBM offers foundation models targeted to an enterprise market it has long served
By John P. Desmond, Editor, AI in Business
With the introduction of its Watsonx platform in the spring, IBM embarked on a strategy of providing a hybrid cloud environment and a generative AI platform that centers on enterprise, data and governance.
The idea is to allow AI builders to train, test and deploy generative AI capabilities powered by foundation models. The studio includes a model library populated with IBM's trained foundation models, including models that generate code for developers and large language models for specific industries. Geospatial models are aimed at helping organizations know more about natural disaster patterns, biodiversity, land use and geophysical processes.
The Watsonx.ai studio will build on open source libraries and offer thousands of open models and data sets from generative AI vendor Hugging Face.
Daniel Newman, an analyst at Futurum Research, called the announcement “the next wave” in generative AI, for its target market of enterprises who need security and data privacy in their AI platforms, according to an account in TechTarget. Watsonx is differentiated by its focus on enterprises, he said, noting, “What we’ve seen mostly launched so far have been tools for users, consumers and social and a few productivity apps.”
Gartner analyst Arun Chandrasekaran noted that generative AI adoption is still early stage and with its focus on cloud and on-premise hybrid infrastructures, IBM may have an advantage. “More enterprise clients want to customize the generative AI models to align with their use cases,” he stated to TechTarget. “IBM is betting on the fact that there will be numerous models used, but clients would look for a consistent set of tools to operationalize them.”
In comments at the announcement of Watsonx in May, IBM CEO Arvind Krishna stated, “We assert that a hybrid cloud choice gives you two and a half times more value than picking a singular answer from one of those underlying landscape choices.”
Newman at Futurum sees that IBM's relationships with enterprise customers and the consulting firms serving them will help it to gain traction. “Companies have tons of data, and they’re going to want to build very specific models on top of the different foundational tools being offered as part of Watsonx,” he stated. “And those models are going to become specific to different parts of the business, and they’re going to need a lot of help.”
In an account of the Watsonx May press conference from CNBC, CEO Krishna stated, “We allow an enterprise to use their own code to adapt the model to how they want to run their playbooks and their code. Then they can deploy it for themselves without any danger of their code leaking.”
Clients and collaborators at that time included SAP, NASA, Wix and PyTorch. Areas of AI integration into the enterprise where IBM will concentrate include customer care, procurement, cybersecurity and elements of supply chain and IT operations, to replace “more repetitive, back-office processes,” he stated.
“We see this easily taking anywhere from 30 to 50 percent of that volume of tasks and being able to do them with really as much or better proficiency than even people can do,” Krishna stated. “That lot, we see getting embraced right away starting this year, and getting to full fruition over the next three to five years.”
Data Governance Tools Seen as Differentiator
The data governance tools of Watsonx.ai are a differentiator, helping to make information more accessible for regulators and third parties, according to a recent account in Forbes. “This is important because of new European, Latin American, and US government legislation currently being formulated,” stated Paul Smith-Goodson, principal analyst, Moor Insights & Strategy, author of the account. “Data governance helps make processes easier for stakeholders and allows clients to manage regulatory changes.”
To help guide its customers in the creation of AI models to help run their businesses, IBM Consulting has created a Center of Excellence to support generative AI efforts, and has staffed it with over 1,000 specialists. IBM’s Watsonx.ai models are trained on IBM’s enterprise-focused data lake, a central repository of structured and unstructured data, using its custom-designed, cloud-native AI supercomputer, Vela.
“Foundation models are the building block for various downstream natural language processing (NLP) tasks,” stated Smith-Goodson. “They can be fine-tuned or adapted for specific applications, such as generative AI models.”
He sees IBM’s focus on governance as critical to enabling generative AI to scale across an organization. “You certainly would not want to scale a model with the potential to produce unpredictable outcomes,” he stated. “A model that generated false information or hallucinations would be detrimental to a business.”
In a historical context, one observer sees four waves of AI. “And despite Big Blue’s best efforts with its Watson stack of software, famously first used to play the Jeopardy! game show and beat human experts back in February 2011, it has not really been much of a commercial success in the enterprise for the first two waves,” stated Timothy Prickett Morgan, co-editor of The Next Platform, in an account in a recent issue.
The first AI wave in this view was based on expert systems; the second, on machine learning in the 2010s; the third, deep learning and neural networks; and the fourth, large language models.
“Foundation models enable generative AI,” stated Sriram Raghavan, vice president of AI strategy and roadmaps at IBM Research, in the Next Platform account. “The reason we are bringing Watsonx together is we want a single, enterprise-grade hybrid environment in which clients can do best of breed of both machine learning and foundation models. That’s what this platform is about.”
The Watsonx.ai stack is API-compatible with the Hugging Face transformer library, which is written in Python, for both AI training and AI inference, Next Platform notes. “This Hugging Face API support is critical because the number of models being written atop Hugging Face is exploding, and this API is becoming a kind of portability standard,” allowing for a consistent model to be deployed internally or on any cloud.
Giving customers flexibility in building their AI platforms is a priority for IBM. “Lots of customers want to take a base model and add 100,000 documents to get their own base model to adapt to use cases,” Raghavan stated. “We want to be able to let them do it with the cheapest cost possible,”
With the stack, data and governance pieces, “This might finally be the Watson AI stack that Big Blue can sell,” article author Morgan stated.
IBM Watsonx Enhancements Include Granite Series Models
IBM this week announced enhancements to Watsonx, preceding its TechXchange Conference to be held in Las Vegas next week. Among them is the Granite series models, which will include a comprehensive list of the sources of data and filtering steps performed to produce the training data for the models. These are planned for third quarter availability.
For assessing model risk, IBM is offering a tech preview for Watsonx governance. Stakeholders will be able to view relevant metrics in dashboards of their enterprise-wide AI workflows, with approval stages shown, so humans are proved to be engaged at the appropriate points.
“We are here to support clients through the entire AI lifecycle,” stated Dinesh Nirmal, senior VP products for IBM Software. “IBM is collaborating with customers to help them scale AI in a secure, trustworthy way.”
An early testimonial was offered by Matrix Holograms, a US company offering an individualized tutoring experience to students and US public schools and low-income countries via AI-powered holograms.
Read the source articles and information in TechTarget, from CNBC, in Forbes. in The Next Platform and from IBM.
(Write to the editor here; tell him what you want to read about in AI in Business.)
