IBM recently unveiled updates and enhancements to its watsonx platform at the annual THINK conference. The focus was on making artificial intelligence (AI) more accessible, cost-effective, and flexible for businesses. This includes the open-sourcing of the Granite models, a new family of high-performance code and language models. These models, available on platforms like Hugging Face and GitHub, can range from 3 billion to 34 billion parameters and excel in tasks such as code generation and error correction.
In partnership with Red Hat, IBM also introduced InstructLab, a capability that allows developers to continuously develop base models through incremental contributions. This approach is similar to open-source software development and enables businesses to create customized AI models that directly address their specific needs.
At the conference, IBM CEO Arvind Krishna emphasized the company’s commitment to open source AI innovation, drawing parallels to the success of Linux and OpenShift. IBM also introduced IBM Concert, an AI-powered tool that provides generative insights into customers’ application portfolios. This tool helps identify and address potential issues before they escalate, ultimately improving efficiency and proactivity.
Additional updates announced at the conference include Watsonx Code Assistant for Java applications, Watsonx Assistant for Z, and enhancements to Watsonx Orchestrate. These improvements aim to streamline AI adoption and usage for businesses.
IBM’s Global AI Adoption Index revealed that while many large companies have implemented AI, there are still barriers hindering the full utilization of AI models. To address these challenges, IBM introduced new AI-powered generative data products, such as Data Product Hub and Data Gate for watsonx, to help organizations manage complex data for AI workloads.
Expanding its capabilities, IBM announced support for new NVIDIA GPUs to enhance AI and mission-critical workloads. The partnership with industry leaders like AWS, Adobe, Meta, Microsoft, and others aims to further improve AI capabilities and model options through the watsonx platform.
In conclusion, IBM is dedicated to advancing AI through open source contributions and strategic partnerships to drive enterprise AI adoption at scale. This approach aims to unlock economic benefits across various industries.
Article Source
https://itbrief.com.au/story/ibm-unveils-major-watson-ai-open-source-innovations