IBM shares AI strategy and tons of customer AI results

Under the leadership of CEO Arvind Krishna and the momentum of AI, IBM has truly transformed its business, technology and leadership. I should know; I worked there as an executive from 2000 to 2010. Now IBM is all about AI and hybrid cloud, including quantum computing.

Lucky analysts spent a day with IBM leadership last week, including the charismatic research director, Dario Gil, the hands-on experience of SVP of Software, Rob Thomas, and the leader of generic AI consulting, Matthew Candy, among other IBM Research executives. ,

IBM’s goal was to align analysts’ view of IBM with the new reality of the AI ​​business with a glimpse at the quantum field ahead of the IBM Quantum Summit in December. Although you’ll have to wait until I attend the summit to hear those details, I can now share some of the things I learned about AI.

IBM AI and Cloud Strategy

As everyone knows, AI starts in the cloud. And as IBM’s enterprise client base grows its AI efforts, its IBM Cloud plays an important role, with a large GPU farm for training and new IBM AI inference processors. IBM’s strategy is based on creating a library of tested and adapted foundation models, designed with a watchful eye to be adapted and annotated to filter out hate, abuse, and profanity (HAP). IBM then leverages these models with governance and data preparation tools directly to customers and makes them available through IBM Consulting clients.

As always, IBM’s Slideware is a bit busy, but as you’ll see, IBM has already done a lot of work to help customers on their journey to increased productivity, better results and lower costs through careful selection of AI models .

Three big opportunities, along with dozens of others.

While IBM shared more than two dozen use cases from its customer engagements, the company says three stand out in traction and ROI: digital labor, customer service, and app modernization.

The specific use cases were surprising. A year ago, most people barely knew what a foundation model was. Now, hundreds of IBM customers are fine-tuning the models and starting to deploy them with the full stack of hardware, data services, AI platforms, SDKs, and AI assistants. The three use cases mentioned above are low-hanging fruit that almost every enterprise can build and deploy. IBM has already benefited greatly, and almost everyone I spoke to during the break and dinner talked enthusiastically about how much time they save using AskHR, which automates many management tasks. Does.

With new foundation models, AI tools, and supporting applications, the WatsonX platform dramatically expands with apps to discover, tune, govern, code, orchestrate, and improve data quality (input) to drive significant business value. Is. Interestingly, the latter includes AI models that identify and filter hate, abuse and profanity (HAP) from input data using IBM’s proprietary inference accelerator, AIU, installed in the IBM Cloud. These new accelerators are fast and ultra-power efficient and are a product of IBM Research. While IBM has not publicly stated their intention to sell AIU outside of the IBM Cloud, I saw a startup (Neurality) with an IBM AIU in their booth at Supercomputing ’23. (More on neurality coming next week!)

The IBM Garage methodology enables business transformation from concept to user adoption while tracking value impact at each stage. This de-risks critical investments and enables long-term change through transparent measurement.

The circles shown represent the method by which the garage achieves this goal step by step. This year, IBM incorporated generative AI into these steps to further accelerate customers’ time to value.

1. co-production of (blue circles) – A combined team of diverse subject matter experts immersed in deep design thinking and research to uncover the true nature and value of the client’s opportunity. This phase establishes alignment on a “big idea” and creates a vision for the minimum viable product (MVP) and its value.

2. co-performance (middle red circle) shows a solution development cycle that uses DevOps and Lean practices to quickly launch and test an MVP. The goal is to validate and improve the value of the MVP in the market through iterative testing, measurement, and re-launching.

3. co-performance (green circle on the right) – Strengthens and scales the new culture of solutions and innovation across the enterprise. This phase is the time to broaden the feature set, stress test code, strengthen security and resiliency, deploy solutions broadly, and expand capabilities to continue driving change.

The results outlined on the right come from the IBM Garage Attachment. 102% ROI is unheard of in the industry, all on 67% faster results. Very good.

And IBM is just getting started. This month, IBM will GA watsonx.ai for on-premises deployment to enhance IBM Cloud. Next month, IBM will release a new watsonx.orchesrate with watsonx.governance and tuning studio, with over 1000 out-of-the-box skills. The two biggest barriers to enterprise adoption of AI are inadequate governance and the aforementioned HAPs, which IBM is aggressively addressing.

Well, I warned you earlier that IBM Slideware can be a little verbose, didn’t I? see below. Although this slide may take 20 minutes to read and understand, the bottom third is fascinating. This slide is only about recruitment and human resource management; IBM has similar slides for other areas of AI impact. This is just the tip of the iceberg. 85% first touch resolution? 66% reduction in human support for HR queries? 50% reduction in friction? seriously? Yes. IBM’s got this and more.

conclusion

While IBM’s original Watson was not successful, the rebirth of Watson as a comprehensive AI platform that enables successful service engagement is beyond “impressive”. When they first announced it, I yawned. Now I am amazed to see how WatsonX is transforming IBM into an AI powerhouse and transforming IBM customers’ businesses.

The analyst community, in general, is a fairly skeptical crowd. But I can tell you that everyone there left impressed. Yes, IBM still needs to work on how it will leverage technology like AI, and we all need to understand how AI will impact employment. One is everyday, the other is beyond my ken.

Well, I’ll be quiet now, at least until the Quantum Summit on December 4th.