Key Considerations for Adding AI to Your Modernization Strategy
Jason McGee, IBM Fellow, CTO IBM Cloud, GM Cloud Platform and Common Services, on how artificial intelligence is changing the way we think about modernization
By Dava Stewart
Key Considerations for Adding AI to Your Modernization Strategy
Jason McGee, IBM Fellow, CTO IBM Cloud, GM Cloud Platform and Common Services, on how artificial intelligence is changing the way we think about modernization
By Dava Stewart
Modernization is one of those concepts that means different things to different people. Jason McGee, CTO, IBM Cloud, says that it can be thought of as a spectrum. On one end is straight workload migration and lift and shift. On the other is refactoring and building new solutions for legacy systems using new technologies and cloud-native approaches.
Challenges: Landing Zones and People
McGee runs the platform for the cloud business at IBM and works with clients who are building and moving some applications to the cloud. Modernization, which can include migration to the cloud as part of a fit-for-purpose approach, and now artificial intelligence (AI) as well, is a big part of his work. If migration can be considered a spectrum, organizations that choose to migrate some applications directly to the cloud and refactor and build new solutions for others face some challenges. AI is bringing a new and rapidly changing set of questions to the process of modernization.
One of the biggest considerations is focusing early on where on that spectrum each application or workload should land. By thinking in terms of multiple landing zones and structuring processes and teams around different approaches, McGee says organizations can begin to build a hybrid strategy that fits specific workloads. “We see a lot of people on IBM Cloud who are doing targeted lift-and-shift migrations of VMWare and Power workloads to cloud, and then we also see a lot of people using OpenShift and containers to transform and modernize applications where they tend to be actually making changes,” he says.
Along with developing a flexible, comprehensive strategy from the beginning, companies need to consider how people will actually use the modernized technology that results from the changes. For example, questions teams should ask from the beginning include:
Solutions: Choosing the Right Path
Depending on which end of the modernization spectrum an organization lands on, different paths exist. For example, a straight workload migration strategy could include moving select workloads to cloud. Another option might be to use OpenShift and containers to transform and modernize apps.
A more structured approach to building a platform team is an emerging trend. In that scenario, an organization might create a team to standardize the use of the cloud, another to set up accounts, one for setting up clusters and so on. McGee says this approach doesn’t really scale well in larger organizations, but for some companies it allows teams to overcome some of the challenges of modernization.
In a situation where refactoring legacy systems that take advantage of new technologies, a more fit-for-purpose approach may be more suitable. Finding the right blend of on-premises and clouds fit to different workloads is more beneficial in that scenario. For example, IBM is optimized for mission-critical workloads, such as those in regulated industries like finance. Others may be better for front office, productivity, marketing or consumer facing workloads.
Bold solutions.
Built to scale.
Whether you’re just starting to modernize or looking to add more accessibility, security and reliability to your infrastructure, Rocket Software can help accelerate and optimize your journey.
With robust infrastructure modernization solutions, your team can focus on the work that counts, while ours enables you to take advantage of the latest innovations in AI-driven automation, development tools and open data formats.
→
Drive productivity and double down on operational performance across hybrid environments.
→
Safeguard against breaches and threats while simplifying regulatory compliance audits.
→
Add resilience and reduce downtime after an attack with enhanced backup and synchronization.
→
Integrate cross-company products to bridge the gap between modern apps and the mainframe.
Modernize Your Infrastructure
The Impact of AI on Modernization Strategies
For some organizations, the right approach to modernization involves AI solutions.
AI is both part of our day-to-day work lives already and is also poised to disrupt virtually every other area. As organizations decide how to implement AI in their modernization strategies, one important qualification must be taken into account: things are changing—and they are changing fast. McGee says answers to questions involving AI today are likely to be different in a month.
However, one thing appears sure. “All of those AI technologies are most accessible in cloud. They’re delivered as cloud services, so it’s causing even more fuel on the fire to move workloads into the cloud more quickly,” says McGee.
Some of the impacts of AI on modernization that are discernable now include:
Skills
What’s necessary to understand the process and technology mechanics of actually making the move in optimization of running workloads and cost analysis of modernization is changing.
“I’m seeing lots of ‘I’m doing the same thing I was doing before, but I’m using AI technologies to help me be more effective at doing it.’ I see a ton of that, and I think that will really accelerate over the next year or two with generative AI techniques being added into some of the traditional techniques,” notes McGee.
Target
The core reason for modernizing and what they’ll modernize to. AI is changing business processes and how organizations interact with customers. It’s similar to the way companies had to quickly add mobile technology to their strategies in the 2000s and early 2010s.
Key AI and Modernization Factors to Consider
As organizations continue the process of digital transformation and the role of AI grows, there are a few key factors to consider in planning your strategy. Data security and privacy have always been a top concern, and McGee says, “[Security is] something that I think always has a strong focus in these modernization environments, but is even more relevant in AI because [data] is being used and embedded in new ways.”
Other crucial factors to consider:
- Governance
- Compliance
- Interoperability and integration
Given the broad implications of emerging AI technologies, companies need to consider how their approaches and processes will be affected, as well as how factors such as security and governance could be complicated. Modernization is necessary, but choosing a path forward carefully rather than taking a more random and unplanned adoption of cloud or AI will yield better and safer results.
Embrace or Avoid AI?
When it comes to choosing to embrace or avoid AI, business leaders may wonder if there is a choice at all. Might some industries be better off avoiding AI completely? McGee says it’s unlikely. “I think, generally speaking, this new wave of AI really does hit everybody,” he says.
If it seems like that’s always the answer, McGee says it wasn’t true for previous machine learning (ML) models. For ML to yield helpful results, the models needed a massive amount of data to train. Smaller organizations didn’t have access to the amount of data required.
Now, foundation models and a wave of large language models allow companies to build AI for others to step in and use. Base models are more general purpose and can be applied to a variety of use cases. These models are more technically accessible, and they also solve real business problems, making them useful for most every kind and size of business.
McGee does add one caveat: “While AI lowers the skill level for things like coding and other specific skills, it does require some expertise in AI itself.” Organizations that use AI need to have someone who understands the models, and that can be difficult to acquire.
McGee suggests developing a “small, central function that’s like a pocket of expertise” within the company—or finding the right partner to develop that expertise. Such a group or team can then make the big decisions around AI and set standards, as well as expose AI to the broader organization where most people will just be consuming the end capability. Many companies developed platform engineering teams to help move to the cloud, and a central group focused on AI could operate similarly.
No More Sleepless Nights.
Learn more
Increase Overall Outputs Through Efficiency Measures
View Infographic
Optimize IT Resources Through Infrastructure, Processes and People
Read infographic
Monitor, Manage and Mitigate IT Risk
Read infographic
Opens in a new window.
AI can do a more comprehensive job of analyzing an environment, moving things, transforming how things are built.
What’s Next
Predicting the future is always a risky endeavor, but McGee has some well-educated guesses about what’s coming next in both the short term and the long term when it comes to modernization.
In the very short term, we can expect a rapid evolution of AI within specific domains. McGee says that AI is “going to get really good at code,” as well as day-to-day workflows. He thinks this is what we can expect in the next 12 to 24 months, and notes that AI is literally a day-to-day thing already. One example: the widespread use of chatbots to perform customer service functions.
As AI becomes more prevalent, the issue of running the models will likely begin to emerge as an inhibitor for some. The cost of licensing some newer products is roughly double what previous software licenses cost. McGee believes the rapid evolution of AI will be forced to slow a bit due to cost and affordability.
In the longer term, crossing boundaries between domains could result in very powerful tools. “You see early signs,” says McGee. “Even now, some models do text and some do images and some do videos, but now you’re starting to see some that could do more than one at the same time.” As an example, he says some AI can turn pictures into code. This multi-domain facet may well represent the second big wave of how AI systems can transform business.
Once that kind of technology is applied to modernization, you can begin to think about things like wholesale migration of code from one technology base to another, or from one operating system to another. At that point he says, “AI can do a more comprehensive job of analyzing an environment, moving things, transforming how things are built.”
“There’s a lot of experimentation going on, but it’s been interesting already to see how the base technique can be reapplied and how what started out as English language and large language models become code very fast because code just kind of looks like another language, and a lot of other digital things can look the same,” McGee says.
Jason is responsible for the architecture and overall technical evolution of IBM’s public cloud. Jason is also the general manager of IBM’s cloud platform, including IBM’s PaaS Services (OpenShift, Kubernetes, Code Engine, Observability, Schematics, Data Services and Developer Tools), Satellite and core services (billing, IAM, account and resource management, user experience).