IBM, accordingly, shifts from competitors and takes a risky decision to launch a family of artificial intelligence models as open-source products. This decision re-centers IBM’s differing business model for AI-based solutions, setting the platform apart from walled gardens and pay-per-use models borrowed from competitors like Microsoft.
Like Meta Platforms, IBM is taking an open-access approach with its new “Granite” family of AI models, allowing third-party firms to adapt these workshop-accelerating tools.
Lessons learned Here, work will be openly developed and shared, and although the models themselves will be open source, eighty percent of IBM’s business will be derived from the Watsonx version that allows for seamless integration and running of the customized models inside data centers.
Kumar further pointed out that global chief executives recognize “the strategic importance of owning generative AI models that create new content.” Krishna went a step further, stating that they are in the formative stage of generative AI models and fully focused on buyer advantage from competition but with safety and responsibility.
In this context, IBM will help the Saudi Data and Artificial Intelligence Authority train its “ALLaM” Arabic language model using the Watsonx model. This collaboration will improve IBM’s ability to support the linguistics of multiple dialects in Arabic.
This open-source approach differentiates IBM. This strategy aligns with IBM’s vision of promoting corporate responsibility in AI adoption. In this way, the applied tools allow organizations to customize the solutions according to their needs. IBM further supports the growth of language models that will help consider people with different backgrounds as equals able to cooperate with AI systems.
