A decade on from the trials and tribulations of IBM Watson, IBM unveiled its multi-model and multi-cloud Watsonx to drive AI-first enterprises – what we are calling “The Generative Enterprise” at HFS.
IBM is describing the platform as a “full technology stack” for training, tuning, and deploying AI models, including foundation and large language models, while ensuring tight data governance controls. Watsonx.data focuses on the data scientist; Watsonx.ai the application developer; and Watsonx.governance is then used to deploy the model using a data model factory to ensure that AI is used ethically and responsibly.
In our view, Watsonx is the first enterprise-grade offering to address the Generative Enterprise holistically. Here’s our interpretation of Watsonx:
- Watsonx.data helps you create the data model. It focused on the data scientist leveraging Red Hat Open Shift to prepare, tokenize, train, and validate internal and external data.
- Watson.ai helps you ask the relevant questions (design prompts). ChatGPT, for example, never gives exactly the same response twice. Learn how to prompt your LLM more intelligently with both short and long prompts to compare quality and accuracy. One of the key benefits of GPT4.0 is the ability to absorb very long prompts (as large as 1000s of words) at rapid speed.
- Watson.ai also helps to iterate. Try asking the same question in different ways, exploring multiple responses to the same prompt, and then comparing the results, detecting bias, and being aware of it.
- Watsonx.governance evaluates responses which is as critical as much of what we have experienced so far as how ChatGPT gets it wrong. Asking questions in different ways, discovering contradictions, and asking to self-assess, is a key aspect of GPT4 that has improved significantly since the prior version.
- Watsonx.governance helps eradicate bias by constantly expanding our understanding of bias in LLMs. ChatGPT, for example, is biased based on the underlying approach used to build the LLM and the data used to train it.
- Watsonx overall fosters the generation of new ideas (Generative Thinking). The big challenge now confronting us as we pursue becoming a true Generative Enterprise is to constantly seek new ideas beyond the constraints of our current LLM. You should ask ChatGPT to summarize, synthesize and find the contradictions in the result it creates. Invest time in learning how conceptual blending approaches are evolving.
The Bottom-line: IBM could have been at the center of the AI revolution but was left out as a bystander. Watsonx has the potential to put IBM front and center of the Generative Enterprise
Watsonx seems very well thought through for AI-powered enterprise use cases, especially for horizontal call centers, HR, and F&A. IBM seems to have learned from its original Watson launch by deploying it internally first, launching an apps development platform to demystify the technology. However, the IBM narrative for Watsonx continues to be more technology-centric versus business-centric, which they need to address with their Watsonx narrative.
IBM still woos the CIO budget, but that’s only a third of the total enterprise tech spend. We believe IBM runs the risk of missing out on the broader CXO budgets but polarizing itself around the CIO.
Posted in : Analytics and Big Data, Artificial Intelligence, Automation, Autonomous Enterprise, Business Data Services, ChatGPT, Cloud Computing, The Generative Enterprise, Uncategorized
