Talsco Weekly: The Bob Effect

Welcome to another edition of Talsco Weekly

IBM i Brief: 📚 Former IBMer’s New Book Puts The Midrange In The Spotlight. 📉 IBM Stock Slides 2% Despite Major AI-Powered Bob Platform Upgrade.
AI: ⚙️ IBM Extends Bob AI Platform With Array of New Features. 🔧 Tokenmaxxing is over. Tokenminimizing misses the point too.
Career: 💡 Preparation is not memorizing answers. It is knowing what matters.
Development: 💻 Using SQL to retrieve information from the internet.
Hiring: 💼 IBM i’s Future Isn’t a Question Anymore — The Talent Question Is.
Leadership: 🔎 Why the Workplace Conversation Never Changes.
Security: 🔐 Quantum computing isn’t just a future problem for IBM i security teams. 🔒 Protecting and Innovating Critical Infrastructure Through New Security Landscapes.
Trends: 🚀 The Bob Effect: Why This Issue Keeps Circling Back to One Platform.

IBM i Brief

📚 Former IBMer’s New Book Puts The Midrange In The Spotlight

Timed to the AS/400’s 38th birthday, IT Jungle profiles Bill Shaffer’s 852-page IBM & Computing, a history that finally gives the Rochester midrange line the attention Shaffer felt it never got in prior IBM histories. The book traces the AS/400’s roots back to IBM’s failed Future Systems project and the Fort Knox consolidation effort of the early 1980s. Shaffer explains the pivot that led to Silverlake:

“The concept of Fort Knox was to combine all the midrange boxes together,” Shaffer said in the interview. “And again, it wasn’t possible. So some of the Rochester engineers that were part of the Fort Knox plan came back and said, you know, we should do a mini-Fort Knox. We should combine the ’34, ’36, and ’38 into one system, which became Silverlake, or the AS/400. And that used the Future Systems architecture, sort of like the FS architecture in the flesh.”

At its peak, Rochester employed 8,000 people and drove $14 billion a year in IBM revenue — a reminder of how large the midrange business once was, and how much of its DNA still runs the platform today.

📉 IBM Stock Slides 2% Despite Major AI-Powered Bob Platform Upgrade

IBM expanded its Bob agentic coding platform with multi-agent coordination, subagent context management, cost/usage tracking, and pre-built modernization workflows, plus a new “Bobalytics” analytics layer.

IBM also launched three domain-specific Bob Premium Packages — for Java, IBM i, and Z modernization. Shares still slid about 2% on the news, opening Thursday at $302.18, with Q2 earnings due July 22; Wall Street holds a “Moderate Buy” consensus ($306.47 target). Notably, Jack Henry’s Chief Technical Architect said the platform is already accelerating RPG development cycles.

AI

⚙️ IBM Extends Bob AI Platform With Array of New Features

IBM rolled out a major expansion of Bob, its agentic coding platform, adding multi-agent coordination, subagent context management for faster responses, integrated cost/usage monitoring, and pre-built modernization workflows. A new “Bobalytics” layer gives teams visibility into productivity, quality, and cost. IBM also unveiled three domain-specific Bob Premium Packages — for Java, IBM i, and Z modernization.

The IBM i take: Bob getting its own dedicated Premium Package, not just a bolt-on skill, signals IBM sees RPG and Db2 for i modernization as a real line item, not an afterthought to Java.

For shops sitting on decades of green-screen logic and shrinking bench strength, multi-agent coordination plus built-in modernization workflows means less “figure out AI on your own” and more “here’s a path already built for your platform.”

Worth watching whether Bobalytics ends up quantifying what a lot of IBM i shops have struggled to prove to leadership: how much legacy code actually gets touched, understood, and safely modernized.

🔧 Tokenmaxxing is over. Tokenminimizing misses the point too

The AI productivity story just hit a wall. After Meta’s internal “Claudeonomics” dashboard ranked engineers by tokens burned, the backlash was swift and spend caps followed. A new piece from Dev Interrupted argues the pendulum swung too far: token maxing inflated a fake numerator, token minimizing shrinks the denominator. Neither answers the only metric that matters—did the work reach production and stay there?

Our Take: This debate has a direct IBM i dimension. IBM Bob prices through Bobcoins, a billing abstraction that converts multi-model token consumption into a single predictable cost unit per task.

Enterprise customers also get pass-through pricing with full usage visibility. That design choice reflects something IBM i developers already know: the question was never how much compute you burned, it was whether the batch ran, the transaction posted, and the output balanced.

IBM Bob is building that outcome orientation into the AI coding workflow from the start. In a market still arguing about leaderboards, that is a meaningful head start.

Career

💡 Preparation is not memorizing answers. It is knowing what matters

Last issue introduced the four questions every IBM i interview really comes down to. This picks up where that left off — turning the framework into action.

Three steps. That is all it takes.

Research the company: not their About Us page. Their industry. What does IBM i do for them? What breaks if the system goes down?
Prepare one story: a moment where you faced a problem and walked through how you thought about it. School, training, or a past career all count.
Write down three questions: real questions about the work and what success looks like.

Whether it is your first IBM i interview or your first in years, the candidates who win are not the ones who memorized the most. They are the ones who showed up knowing what they bring to the table.

The takeaway. “I built X, which solved Y” beats a list of technologies every time. Ready to talk through yours? Book a 20-minute call.

Development

💻 Using SQL to retrieve information from the internet

Simon Hutchinson shows how to call an external API straight from Db2 for i, using QSYS2.HTTP_GET to fetch JSON and JSON_TABLE to parse it into columns — demoed by pulling live stock prices for Dell, IBM, and Tesla. He walks through both a raw SQL example in ACS Run SQL Scripts and a full free-format RPG program that loops through symbols with FOR-EACH and %LIST, displaying results via DSPLY. Notes it could be extended to insert into a table and run on the job scheduler for automated, unattended data capture.

Hiring

💼 IBM i’s Future Isn’t a Question Anymore — The Talent Question Is

Every AI story in this issue points to the same conclusion: IBM i isn’t fading, it’s getting reinforced. IBM just fast-tracked a dedicated Bob package for a platform that was supposedly on its way out. Companies don’t build purpose-specific AI tooling for systems they expect to retire.

But there’s a hiring myth worth killing here. Bob explaining a 90,000-line program or tracing a calculation 18x faster doesn’t replace the developer — it replaces the six hours of archaeology that used to stand between a developer and doing the actual job. Someone still has to validate what Bob produces, understand the business logic it’s surfacing, and own the judgment calls a machine can’t make about a system nobody fully documented in the first place. That’s more valuable work, not less, and it takes an experienced IBM i mind to do it.

If anything, AI raises the floor on what a good IBM i developer needs to know: less time lost to tribal-knowledge hunts, more time spent on the decisions that actually require a human. Shops that read this as “we need fewer people” are going to lose the race to the ones who read it correctly: “we need the right people, freed up to do the work that matters.”

Looking to hire IBM i talent? Book a 20-minute call.

Leadership

🔎 Why the Workplace Conversation Never Changes

Decades of leadership, engagement, and culture initiatives keep circling back to the same problems because they change what people do, not what people see. A new feedback process built on the belief that employees need to be monitored just becomes another monitoring tool. The intervention changes, the underlying assumptions don’t.

For IBM i leaders, this cuts close. Many of the patterns running your team today, who gets trust, how skills get valued, what “modernization” is even allowed to mean, were set years ago, often by necessity, and have quietly hardened into “how things are done here.”

A new AI tool, a new hire, a new process gets bolted onto the same old lens, and nothing actually shifts. Breaking the cycle means questioning the assumptions the team is operating from before reaching for the next framework or tool.

Security

🔐 Quantum computing isn’t just a future problem for IBM i security teams

The “harvest now, decrypt later” threat means encrypted data leaving your IBM i today could be decrypted by tomorrow’s quantum computers. NIST has finalized post-quantum cryptography standards, and enterprises like Cloudflare and Signal have already made the shift. IBM i shops with long-life, high-value data need to understand what’s at risk and start planning their migration before the window closes.

Our Take: This isn’t about replacing your existing IBM i security framework. It’s about adding a layer your current tools weren’t built to handle.

🔒 Protecting and Innovating Critical Infrastructure Through New Security Landscapes

IBM Z GM Skyla Loomis reaffirms the platform’s resiliency claim (under a third of a second of downtime a year, 99.999999% uptime) and announces GA of three new Z security tools: zSecure Detection (ransomware/threat monitoring), zSecure Secret Manager (automated certificate management), and Z Database Assistant (agentic AI for DBA tasks). Not an IBM i release, but the same resiliency bar IBM i shops already live by.

What Project Glasswing and Project Lightwell are:

Project Glasswing — an industry coalition (IBM, Anthropic, and others) that hunts for vulnerabilities in widely-used and open-source software before attackers do, using AI to speed up discovery. Findings get shared via coordinated disclosure and upstream patches.
Project Lightwell — a $5 billion IBM/Red Hat initiative pairing AI with 20,000+ engineers to harden the open-source supply chain. Its centerpiece is a “security clearinghouse”: report a vulnerability through an embargo-protected channel, get a validated, signed patch backported to your exact dependency versions, no forced upgrades.

Trends

🚀 The Bob Effect: Why This Issue Keeps Circling Back to One Platform

If it feels like Bob showed up three times in this issue, that’s not a coincidence — it’s the trend. A stock that slid despite a major AI upgrade. A platform expansion built around multi-agent coordination and cost analytics. And now, a Premium Package for i that connects natively to live IBM i environments, reads and writes RPG, COBOL, CL, and Db2 for i in place, and ships with skills purpose-built for legacy modernization: explaining old code, converting Fixed-Format to Free-Format, refactoring monoliths, filling the database-engineer gap most IBM i shops don’t have anymore.

The individual features matter less than the pattern. One user documented a report calculation 18x faster than a senior developer’s prior effort. Another used Bob to auto-document a 90,000+ line undocumented program for new hires walking in with zero institutional knowledge. That’s not a coding assistant demo — that’s the skills gap this community has been worried about for years, getting addressed in real time.

Our take: Bob for i didn’t take years to show up after the general platform launched — it showed up the same day. That pace is the real story here, not any single feature. IBM i has spent a long stretch watching modern tooling arrive late, if at all. This is the counter-signal: AI on IBM i is no longer a future promise, it’s a present-tense shift, and it’s moving at the speed of the rest of the platform for once. Worth watching whether that pace holds through the next release cycle, or whether this was a one-time push to make a splash.

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!