Talsco Weekly: 🎙️ One IBM i Developer’s Take on AI

Welcome to another edition of Talsco Weekly

IBM i Brief: 📊 IBM’s Power Systems business is growing — but GenAI revenue hasn’t arrived yet. 🟢 IBM i doesn’t need the spotlight. It just works.
AI: 💡 Speed isn’t the goal. Better thinking is. 🏗️ Too many vendors may be the reason your AI strategy is stalling.
Career: 💬 A conversation with an IBM i developer left an impression.
Development: ⚡ Agentic AI is breaking GitHub’s infrastructure. 💸 SaaS pricing is becoming a trap for IT shops.
Data: 📡 IBM i shops can now stream live data to AI platforms — without the batch lag.
Leadership: ⚠️ What happens when new leadership misreads the platform.
Security: 🔒 AI agents introduce a security playbook IBM i shops don’t have yet.
Trends: 🤝 What does the future look like with AI in the Enterprise?

IBM i Brief

📊 IBM’s Power Systems business is growing — but GenAI revenue hasn’t arrived yet

IBM’s Infrastructure group posted $3.33 billion in Q1 2026 sales, up 15.3%, with Power Systems estimated at $205 million — a 19% jump driven by the Power11 upgrade cycle and pre-hike buying. IBM CEO Arvind Krishna outlined how on-platform AI inferencing is already transforming mainframe workloads, pointing to a blueprint IBM i shops are watching closely. The GenAI bump for Power Systems is expected. It just hasn’t hit the numbers yet.

🟢 IBM i doesn’t need the spotlight. It just works

One technology after another has promised to replace it. None have. A young professional working with the platform recently put it simply — when you first hear “AS/400” it sounds outdated, but the moment you work with it, something shifts. Layer after layer of capability, stability, and reliability. He spotted a green screen at a grocery store, transactions flowing seamlessly. IBM i isn’t legacy. It’s legendary.

AI

💡 Speed isn’t the goal. Better thinking is

AI can process, summarize, and generate at remarkable speed — but faster output doesn’t mean better reasoning. As we’ve noted before, an average developer with a great AI tool is still an average developer.

The real competitive advantage for IBM i developers isn’t in who adopts AI quickest. It’s in who uses it most deliberately.

Your platform expertise, your decades of business logic knowledge, your understanding of why the system works — that’s the direction AI needs. It can sharpen your thinking. It can’t replace the judgment behind it.

🏗️ Too many vendors may be the reason your AI strategy is stalling

Enterprises juggling 200+ suppliers are discovering that vendor sprawl — built up through years of well-intentioned point solutions — is now the primary bottleneck to scaling AI. Some Fortune 500 companies report 80% of vendors accounting for just 20% of spend. One global bank cut 50% in costs over five years by consolidating across 80 functions. Less complexity, faster AI execution.

What does this mean for IBM i?

There’s a push and pull reshaping the IBM i vendor landscape.

The push: outside vendors are moving in. Enterprise complexity and AI demand are making IBM i shops attractive targets for broader tech players offering integration, automation, and data solutions.

The pull: IBM i-native vendors — many who’ve coasted for years — are now being forced to innovate. AI isn’t optional anymore. It’s time to build.

Both forces are accelerating change.

Career

💬 A conversation with an IBM i developer left an impression

I speak with IBM i professionals every day. Some for the first time, some that I have known for years.

My call with this IBM i-er stood out. He’s one of the best of the best — an independent consultant who has always stayed on top of his game and his craft.

Sharp, forward-thinking, energetic.

He has recently embraced AI as a way to learn, a tool to help him ramp up quickly in areas he hasn’t been as strong. The excitement was unmistakable. As he put it,

“I have a lot of great work ahead of me. These are exciting times. And by no means does it mean abandoning IBM i. It simply reinforces the experience I already have.”

Development

⚡ Agentic AI is breaking GitHub’s infrastructure

GitHub’s CTO just published a candid post-mortem after two major incidents in one week. The root cause isn’t a bug — it’s volume. Since late 2025, agentic development workflows have exploded, pushing pull requests, API calls, and repository creation to record levels. GitHub scrambled its capacity plans from 10X to 30X.

This exponential growth does not stress one system at a time. A pull request can touch Git storage, mergeability checks, branch protection, GitHub Actions, search, notifications, permissions, webhooks, APIs, background jobs, caches, and databases.

For IBM i shops evaluating AI-assisted development, the lesson is clear: agentic workflows aren’t a future consideration. They’re already reshaping the infrastructure underneath your tools.

Breaking It Down:Git is version control software — the technology that tracks code changes. GitHub is a cloud platform built on top of Git that adds collaboration, automation, and now AI-powered development workflows. IBM i shops can run Git locally today without touching GitHub. But the agentic AI layer — where AI agents write code, open pull requests, and run pipelines automatically — that lives on GitHub’s platform. Local Git gets you in the game. GitHub gets you in the future.

💸 SaaS pricing is becoming a trap for IT shops

Disclosure: this one comes from a vendor. But strip away the pitch and the questions left standing are worth your time: SaaS pricing, lock-in, and who owns what you build.

SaaS costs rose 11.4% in 2025 — roughly five times general inflation — and 73% of vendors raised prices that year. The real danger isn’t the subscription itself; it’s building critical business applications on rented platforms. Cancel and you lose everything your team built. For IBM i shops running custom apps over their own data, the case for perpetual licensing over SaaS dependency has never been stronger.

Our Take: SaaS has its place. But there’s a bigger shift coming. Bain & Company finds that agentic AI is already automating tasks and replicating workflows that SaaS products once owned — and the financial markets are noticing. As of March 2026, the median valuation multiple for public SaaS companies has dropped to 3.4x, with revenue growth decelerating to 12.2% and forecasts pointing lower.

With AI maturing and the right governance and security in place, IBM i shops may soon find it easier than ever to build their own. The economics of dependency are changing. Shops that own their infrastructure and their data are quietly gaining an advantage most SaaS vendors haven’t priced in yet.

Data

📡 IBM i shops can now stream live data to AI platforms — without the batch lag

The journal is the pipeline. CData Sync has added change data capture (CDC) support for Db2 for i, tapping the database journal to detect and forward changes in near real time. No full table scans. Minimal production impact. Data flows continuously to cloud destinations like Snowflake, Postgres, and Amazon S3 — and from there, into AI and analytics pipelines.

What this means. If your IBM i is the system of record, it can now also be the live data source for AI. That’s a meaningful shift. The data advantage IBM i shops have always had just became easier to activate.

Leadership

⚠️ What Happens When New Leadership Misreads the Platform

A new CIO walks in.

They see green screens. They see RPG. They see something their resume has never touched.

They call it legacy. That conclusion has cost organizations millions of dollars and years of momentum.

To be fair, some of it is. Not because the platform aged out, but because “if it ain’t broke, don’t fix it” became the default strategy for years. Interfaces went unmodernized. Processes stayed frozen. That part is real.

But the platform’s capability never stopped advancing.

IBM i’s backward compatibility reads as stagnation from the outside. From the inside, it is the reason the data is clean, the integration is tight, and the business never went down. That is not a liability. That is the foundation AI-ready infrastructure gets built on.

The mistake is not asking questions. The mistake is arriving with the answer.

The leaders who get it right listen before they diagnose. They give a voice to the team that lives inside the platform every day. They can then help transform what is actually there into what the business needs next.

The developers who stayed current with every IBM i release, stepped outside the platform, and brought that knowledge back, those are not the people to sideline. They carry institutional wisdom you cannot hire or replace. Alienating them is not a modernization strategy. It is a liability.

A new leader who recognizes that changes the trajectory.

One who does not will spend the next decade finding out why.

Security

🔒 AI agents introduce a security playbook IBM i shops don’t have yet

AI agents can read your code, query your database, call APIs, and execute actions — autonomously. NIST is already flagging the risks: prompt injection, privilege escalation, and cascading failures across connected systems. IBM i shops have long operated with a “nobody targets us” mindset. That era is over. Adopting agentic AI without updated security governance isn’t bold. It’s exposed.

Breaking It Down: Most IBM i shops have two security priorities running in parallel right now — whether they know it or not.

First, get your house in order: audit your existing security policies, validate your configurations, and run penetration testing to confirm what’s actually enforced versus what’s assumed.

Second, start building an AI governance framework before agents start accessing your data through APIs and MCP connections.

These two tracks can’t run sequentially. The agents aren’t waiting.

Trends

🤝 What does the future look like with AI in the Enterprise?

AI is not going away. It’s here to stay and:

The most forward-looking teams are letting go of the old and embracing agentic ways of working. They’re collaborating with tools like Claude or Codex as agents take on more of the execution.

On the IBM i:

The leaders who will matter in the next two years aren’t the ones who master AI. They’re the ones who figure out how to lead teams alongside it, unlocking productivity while maintaining control.

Join

Contact us

If you are an RPG programmer looking to explore opportunities or a client who is looking for a talented IBM i professional, please contact us. We look forward to assisting you.

Do you know of someone who could benefit from Talsco Weekly? If so, please use the social media buttons to spread the word. Thank you!