Acceleration Economy analysts Kenny Mullican, CIO of Paragon Films, and I spent this week at IBM Think 2024, engaging with company executives, ecosystem partners, and customers. Below, we offer 10 big takeaways from a wide range of strategic developments shared during the gathering.
1. All In on AI
IBM Chairman and CEO Arvind Krishna, in his keynote address, positioned AI as revolutionary technology, comparing it to such developments as the steam engine, electricity, and the Internet. It truly has the capability to change everything about the way we do business and certainly the way we interact with technology. IBM is going all in, and Krishna conveyed that energy. IBM highlighted partner and client momentum its AI and data platform, IBM watsonx, is having since announcing watsonx at last year’s Think conference.
2. Watsonx Assistants Showcase Innovation
IBM introduced a new class of watsonx assistants and enhanced automation capabilities to boost AI-driven productivity at scale. Powered by Granite Large Language Models (LLMs), the assistants are designed to help clients achieve higher productivity gains across various domains:
- watsonx Code Assistant for Enterprise Java Applications to help businesses simplify and accelerate their Java application lifecycle with automation and GenAI
- watsonx Orchestrate Assistant Builder to help clients create their own AI assistants across domains
- watsonx Assistant for Z to transform user interaction with mainframe systems and facilitate rapid knowledge transfer
3. AI Without the Heavy Lifting
Watsonx Orchestrate rounds out the watsonx offerings in a critical way. Equipped with guidelines to streamline the automation process, Orchestrate is part of IBM’s larger strategy to continue adding what you might call “AI middleware” offerings: presets, frameworks, and templates that customers can use to rapidly test and deploy automation without the heavy lifting of hiring AI talent, training models, and wiring together various systems.
4. Better BI
IBM learned from customers that the powerful Cognos Analytics platform has a learning curve that keeps many business decision-makers from being able to quickly get answers from their data to support real-time business decisions. In response to customer feedback, they’ve put a big focus on using GenAI to overcome that challenge. The goal: users should be able to query their data using natural language and get a response in under five seconds, and actionable insights within one minute. This is a clear indicator of where we’re heading with GenAI: AI serving as the user interface.
5. Addressing App Lifecycle Management
IBM Concert, previewed at Think this week, leverages GenAI to provide insights into your application stack, revealing connections, dependencies, gaps, and opportunities within your architecture. Utilizing watsonx technology, Concert offers real-time data and dependency mapping. This capability helps you identify operational challenges, diagnose root causes, anticipate potential issues, and proactively implement recommended actions and automations.
6. Tapping ‘Dark Data’ For Maximum LLM Impact
Multiple Think 2024 presentations featured a slide with an interesting fact: Nearly all publicly available data is represented in LLMs because it’s housed on the Internet. Yet less than one percent of data housed within enterprises is used in training models. This represents an enormous opportunity for value creation – if properly accessed and handled, this “dark data” can create extremely powerful, relevant models.
Recognizing this opportunity, IBM announced InstructLab, a new way to rapidly update LLMs based on untapped or new organizational data. The InstructLab model alignment technique is somewhat in between the process of Retrieval Augment Generation (RAG) – which doesn’t involve changing model weights but inference on custom datasets – and actual fine-tuning, which involves re-training the model’s weights and biases. InstructLab is an innovative approach; it’s exciting to see customers and partners applying it and their results.
7. AI at Scale: Moving out of the Sandbox
One major theme IBM and Krishna emphasized is that AI shouldn’t only be tested or piloted on a small scale. Now is the time to deploy AI across your company, across your customer base, to various aspects of the business. This idea is summed up in the phrase “AI at scale” which was repeated many times in the past few days. IBM is trying – and succeeding – in shifting the market’s mindset from, “AI is useful to make my operations leaner or save on costs” to “AI can drive new revenue, allow us to gain market share, and expand product offerings.” It’s a more aggressive approach and it’s resonating.
8. Is Quantum Real? 3 Trillion Experiments Prove It
Krishna talked about quantum computing as real technology that is being used today, not just something we’ll see in the future. IBM has 70 quantum computers deployed around the globe, with over 3 trillion experiments already conducted. IBM predicts that within three to five years, we’ll be able to solve problems with quantum that could never be solved with classic computing tech. We’re eager to see if that comes true, but three to five years is not long to wait!
9. Partnerships Across the Spectrum
In the opening keynote, Krishna shared the stage with various partners to talk about their relationships and customer benefits. The first was a representative from SDAIA, the AI research arm of the Saudi Arabian government. With IBM, SDAIA built the world’s highest-performing LLM in Arabic, covering over 20 dialects and unlocking enormous opportunities for Arabic-speaking countries to apply LLMs. Krishna also welcomed Adobe CEO Shantanu Narayen and Citi co-CIO Shadman Zafar to talk about various ways the companies are collaborating to scale GenAI and benefit customers
In the past two years, IBM has done an incredible job lowering the barriers for partners to access its product ecosystems from startups, SMBs, and enterprises. Through IBM’s Partner Plus program, organizations of all sizes and kinds can seamlessly interface with IBM. Think, which started with Partner Plus Day, has been a great exhibition of that with opportunities to walk the show floor engaging IBM experts on areas for collaboration, suggestions on how its tech can be applied, and beyond. IBM made it accessible and innovation-friendly, especially for startup founders like myself.
10. The New IBM
Based on conversations with leaders from IBM, partners, startups co-creating with IBM, and people who have worked for the company for decades, there is one common thing that everyone seems to be rowing in the same direction on: the IBM of today is not the IBM of even three years ago, let alone five or 10 years ago.
Since Krishna became CEO, the company has fundamentally changed its market position and products, and its momentum is palpable. They’ve become a leader in business-facing applications leveraging cutting-edge AI through watsonx.
They’ve pioneered an ecosystem model of co-creation, which is, in turn, reshaping how many other large tech companies build AI applications as evidenced by news this week with Adobe, AWS, Microsoft, SAP, Salesforce, and more. Perhaps most importantly, they have defined what the new IBM can look like beyond mainframes and a computer playing Jeopardy: one with the ambition and technology to stake out a dominant role in the AI era.
Stay tuned as Acceleration Economy tracks IBM’s continued AI innovation and partner initiatives throughout 2024.