In my opinion, generative AI is in high demand with the rise of consumer-focused artificial intelligence (AI) offerings. The development of large language models designed to facilitate generative AI has ushered in a new era of innovation, bringing the benefits of AI and machine learning to new audiences.

While this technology can be used to produce consumer-facing applications that can write poetry or create a meal plan, generative AI, and the foundation models technology that generative AI is built upon, will have a more profound impact on the enterprise. According to the 2023 IBM Institute for Business Value CEO Study, 75% of CEOs surveyed believe that competitive advantage will depend on who has the most advanced generative AI, and 50% of respondents are already integrating the technology into products and services.

But how can a company make the most of these advancements and maximize potential benefits?

Enterprise data is often the key to unlocking value from AI. Mission-critical data and applications, such as data related to revenue generation, logistics, or regulatory compliance, can provide invaluable insights when AI is applied. By applying AI techniques, such as traditional or generative AI for model training, fine-tuning, and inferencing, companies have the opportunity to  gain insights from their mission-critical data and applications. Notably, many of these mission-critical applications are built on the IBM Z platform.

Leveraging enterprise data can help businesses accelerate their AI journey. For example, IBM clients in the financial services industry have reported that they are addressing the use case of real-time fraud detection using inferencing methods with IBM z16. In the future, we may also see the transformation of developer roles and increased productivity across all roles with AI’s help. AI for developers may help increase their productivity and efficiency by automating routine programming tasks.

Accelerating Decision Velocity with AI on IBM Z

With the rise of instant payments, the dramatic increase in online transactions, increasing fraud, and new regulations, there’s no shortage of challenges within the financial services industry. One use case that we see some banks addressing with AI is analyzing transactions for fraud.

Based on my conversations with clients, key considerations for all industries with mission-critical workload include applying AI without slowing down applications, accessing results of AI models before transactions time out, and scoring every single transaction — not just a sample of the transactions.

Developed over five years, IBM z16 is designed to address these challenges. For example, clients reported that running deep-learning models at scale in real time on all transactions was not practical, meaning fraud detection models were typically only run on a subset of financial transactions. IBM z16 brings together AI inferencing, the IBM Telum Processor, and the high-volume transaction processing IBM is known for. IBM z16 is designed to score business transactions at scale with low latency. For consumers, this could mean reducing the time and energy required to handle fraudulent transactions on their credit card. For both merchants and card issuers, this could mean a reduction in revenue loss as consumers could avoid the frustration associated with false declines.

The IBM Integrated Accelerator for AI on the Telum Chip in IBM z16 is optimized for low latency and high throughput and is designed to enable clients to leverage state-of-the-art AI in workloads where they could not before. AI inference tasks on IBM Z are designed to take advantage of zIIPs to reduce the overall operational costs of AI on IBM Z. Some key examples of AI processes that leverage zIIPs are: AI models deployed in Watson Machine Learning of z/OS, all the open-source AI frameworks and tooling (such as TensorFlow, Pytorch, Python packages, etc.) running inside z/OS Container Extensions (zCX), and SQL Data Insights that utilizes zIIPs for executing semantic queries.

Accelerating Developer and System Programmer Productivity

In the future, AI will revolutionize the way we work, impacting various roles, including sysprogs as well as front-end and back-end developers. This transformation follows a pattern that is similar to earlier technologies that have reshaped industries. Over time, mainframe programmers have adapted and progressed with advancements in technology. Today’s mainframe programming has become notably faster, capable of significantly larger projects, and leveraging state of the art technologies compared to the past.

AI can help developers, sysprogs, and sysadmins become more productive, enabling them to focus on more challenging coding problems and removing repetitive work. For example, sysadmins and sysprogs are able to utilize IBM Watson Code Assistant to generate Ansible Playbook code scripts from plain English for streamlining system automation tasks. Later in 2023, developers should be able to leverage technologies, such as IBM Watson Code Assistant to generate code starting from simple natural language.

IBM is also planning to release a new version of IBM z/OS at the end of September 2023, designed to build the foundation for enabling intelligent systems administration guidance and automation that learns and improves. IBM z/OS 3.1 is being built to enable AI framework support, which is intended to augment z/OS with intelligence that optimizes IT processes, simplifies management, improves performance and reduces special skills requirements. With this, it’s intended that users will be able to leverage AI in z/OS to start their journey to systems self-management.

With businesses growing, the demand for developers will rise. AI can help developers focus on innovation and meeting business needs, while AI-powered tools can handle the mundane tasks, such as code maintenance and systems upgrades.

To learn more about how AI could impact enterprises today and in the future, please explore the exciting use cases, products, and technical resources at Journey to AI on Z.

Statements regarding IBM’s future direction and intent are subject to change or withdrawal without notice and represent goals and objectives only.