• Releases a family of IBM Granite models into open source, including its most capable and efficient Code LLMs that can out-perform larger code models on many industry benchmarks

• Jointly with Red Hat, launches InstructLab, a first-of-its-kind model alignment technique, to bring open-source community contributions directly into LLMs

• Unveils new vision and momentum for new data and automation capabilities; Accelerates generative AI infusion into IBM assistants, automation, infrastructure, resource management products and consulting services

• Strengthens collaborations with AWS, Adobe, Meta, Microsoft, Mistral AI, Palo Alto Networks, SAP, Salesforce, and SDAIA to expand capabilities and offer model choice, flexibility and governance through watsonx

At its annual THINK conference, IBM introduced significant enhancements to its watsonx platform, emphasizing its commitment to making artificial intelligence (AI) more open, cost-effective, and flexible for businesses. During his keynote speech, CEO Arvind Krishna detailed IBM’s strategy to invest in and contribute to the open-source AI community.

1. Open-Source AI Innovations

• Granite Models: IBM has open-sourced its advanced Granite language and code models, which range from 3 billion to 34 billion parameters. These models are designed for tasks like code generation, bug fixing, and application modernization. Available under Apache 2.0 licenses on Hugging Face and GitHub, these models have been optimized for performance across various programming languages.
• InstructLab: In collaboration with Red Hat, IBM launched InstructLab, enabling continuous development of AI models through incremental contributions. This initiative allows businesses to tailor AI models to their specific needs, fostering a more collaborative and open innovation environment.

2. New watsonx Assistants

• watsonx Code Assistant for Enterprise Java Applications: Launching in October 2024, this tool aids in transforming Java applications.
• watsonx Assistant for Z: Available in June 2024, this assistant enhances user interactions with IBM Z systems.
• watsonx Code Assistant for Z Service: Also available in June 2024, this service provides code explanation and documentation through natural language.

3. Enhanced Hardware and AI Deployment

• NVIDIA GPU Offerings: IBM is expanding its GPU offerings to include NVIDIA L40S and L4 Tensor Core GPUs. This support extends to Red Hat Enterprise Linux AI (RHEL AI) and OpenShift AI.
• Quick AI Deployment: IBM is leveraging deployable architectures for watsonx to enable swift AI deployment, equipped with robust security and compliance capabilities.

4. New Generative AI Data Products

• IBM Data Product Hub: Launching in June 2024, this hub will help organizations manage and optimize their data for AI workloads.
• Data Gate for watsonx: Another June 2024 launch, this capability will enhance data governance and observation.

5. AI-Powered Automation

• IBM Concert: Set to debut in June 2024, this generative AI-powered tool will serve as the nerve center for an enterprise’s technology and operations. It integrates with existing systems to provide predictive insights and suggest solutions for application management issues.

6. Expanding Ecosystem and Partnerships

• Third-Party Integrations: IBM continues to expand its ecosystem, bringing third-party models onto watsonx and enhancing collaborations with major tech companies like AWS, Adobe, Meta, Microsoft, and more.
• Strategic Partnerships: New partnerships, such as with Mistral AI and expanded collaborations with Palo Alto Networks and Salesforce, aim to deliver advanced AI capabilities across various platforms and industries.

7. Client and Partner Success Stories

• Broadridge Financial: Leveraging generative AI for enhanced developer productivity and application sustainability.
• Casper Labs: Utilizing blockchain-enabled governance for responsible AI scaling.
• Department for Work and Pensions: Transforming service delivery with AI, focusing on human-centered values and reusable components.
• Dun & Bradstreet: Co-creating AI solutions for procurement and business process optimization.

IBM's updates to the watsonx platform and its continued investment in open-source AI highlight the company's commitment to driving innovation, enhancing AI accessibility, and fostering collaboration across the tech community.