Welcome to another edition of Talsco Weekly

IBM i Brief: 📝 IBM i Announcements to think about.
AI: 💰 Google leads in AI training costs at $192M. 🤖 AI revolutionizes legacy code understanding and modernization. 🤔 AI rights debate gains momentum in tech industry.
Career: The IBM i Generalist – Specialist. 😆 AI transforms IBM i programmers into action figures.
Development: 💻 Cross-platform Node.js development guide gains traction. 🚀 JavaScript developers key to IBM i transformation success. 👩‍💻 Here’s a look at how functional programming is transforming enterprise software development.
Data: 💾 Legacy systems hold treasure trove of data insights.
Learning: 📚 Essential IBM i manuals guide for developers.

IBM i Brief

📝 IBM i Announcements to think about

IBM discontinues key LPPs in 7.6, including System/38 emulation and CICS Transaction Server, while introducing Java 21 and planning PowerHA SystemMirror automation.

You can learn more about it here.

AI

💰 Google leads in AI training costs at $192M

The surging expenses of training AI models have sparked industry debate, with Google’s Gemini 1.0 Ultra topping the charts at $192 million. While OpenAI’s GPT-4 cost $79 million to train, newer competitors like DeepSeek claim drastically lower costs of $6 million. The high costs stem primarily from cloud computing resources and developer salaries, raising questions about sustainability and market barriers.

How AI Models Are Trained and Why It’s So Expensive

AI models are trained by collecting vast datasets, like texts or images, and cleaning them for consistency. Engineers design neural networks to process this data, and during training, models predict, compare, and adjust parameters to improve accuracy. Fine-tuning with specialized data hones specific tasks.

However, this process is costly. High-powered GPUs and TPUs run for weeks, racking up millions in computing fees. Curating and labeling data is labor-intensive, and skilled engineers command high salaries. Multiple training rounds and maintaining data centers further escalate costs.

The hope is, in the future, the cost will come down.

How are IBM as well as vendors in the IBM i space approaching AI?

🤖 AI revolutionizes legacy code understanding and modernization

ARCAD is developing AI tools to enhance IBM i application development at both code and application levels. Their approach combines code explanation capabilities with IBM’s upcoming code assistant, using LLMs trained on company codebases to understand house programming styles. The system aims to improve development speed while maintaining consistency, with AI-powered tools helping analyze dependencies and architectural specifications through natural language interfaces.

🤔 AI rights debate gains momentum in tech industry

Anthropic leads research into “model welfare” as AI capabilities advance. The company explores whether AI models deserve worker protections and rights, particularly as they develop more human-like qualities.

As we build those AI systems, and as they begin to approximate or surpass many human qualities, another question arises. Should we also be concerned about the potential consciousness and experiences of the models themselves?

This investigation examines potential AI consciousness and ethical treatment of models, including protecting them from abusive interactions. The debate draws parallels to animal welfare concerns, especially as companies build massive “AI factories” for processing workloads.

Career

The IBM i Generalist – Specialist

The IBM i developer represents a unique breed in the technology landscape – they are both generalists and specialists simultaneously. This duality stems from the integrated nature of the IBM i platform itself.

The Generalist Nature

System Administration: Managing system resources, security, and performance
Database Management: Working with Db2 for i, handling data integrity and optimization
Application Development: Creating and maintaining business applications
Integration: Connecting IBM i with modern systems and technologies

The Specialist Depth

Deep Platform Knowledge: Understanding the intricate architecture of IBM i
RPG/SQL Expertise: Mastering traditional languages while adopting modern variants
Business Logic: Intimate knowledge of critical business processes
Security Practices: Implementing robust security measures specific to IBM i

Modern Evolution

Today’s IBM i developer must balance traditional expertise with modern capabilities:

Open Source Integration: Incorporating PHP, Python, and Node.js
API Development: Creating RESTful services for system integration
Web Technologies: Building modern interfaces for legacy applications
Cloud Integration: Connecting IBM i with cloud services and platforms

The Competitive Advantage

This unique combination of breadth and depth gives IBM i developers significant advantages:

Problem Solving: Ability to understand and solve issues across the entire stack
Business Value: Deep understanding of both technical and business requirements
Innovation: Capability to bridge traditional systems with modern solutions
Career Longevity: Adaptability to evolving technology while maintaining core expertise

This generalist-specialist combination makes IBM i developers particularly valuable in today’s technology landscape, where understanding both legacy systems and modern technologies is increasingly important.

😆 AI transforms IBM i programmers into action figures

Here is some quick humor. Have you heard about the social media trend called the “Barbie Box Challenge”? To be honest, I had not until now.

Here is an IBM i take on how to create your own collectible action figure.

Development

💻 Cross-platform Node.js development guide gains traction

Here is a comprehensive guide on developing cross-platform Node.js applications.

It emphasizes the importance of platform-agnostic coding practices (meaning IBM i compatible).

Key points include using path.join() for file paths, implementing proper error handling across systems, and leveraging cross-platform compatible libraries.

The discussion highlights common pitfalls in Windows vs. Unix development and suggests tools for ensuring consistent behavior across different operating systems.

🚀 JavaScript developers key to IBM i transformation success

A new study reveals the strategic advantage of JavaScript developers in IBM i modernization efforts. The abundant availability of JavaScript talent, combined with their web development expertise, makes them ideal candidates for transforming legacy applications.

Organizations can leverage JavaScript frameworks and tools to create modern web interfaces while preserving valuable backend business logic on IBM i.

👩‍💻 Here’s a look at how functional programming is transforming enterprise software development

A comprehensive analysis highlights functional programming’s growing role in modern software development. The paradigm emphasizes pure functions, immutability, and declarative coding styles, leading to more maintainable and scalable applications.

Key languages like JavaScript, Scala, and Haskell are driving adoption, with particular benefits seen in web development, data processing, and machine learning applications.

Could functional programming be adopted by RPG developers?

Data

💾 Legacy systems hold treasure trove of data insights

Here is a quick video that explores how companies can unlock the valuable business data accumulated in their IBM i systems over decades of operation.

While legacy systems may be older, they contain mission-critical data that continues to drive business value.

The key trend is finding innovative ways to leverage this data effectively. Modern tools and integration methods like ETL and APIs now make it easier to access and analyze IBM i data, while keeping it secure in its native environment.

Trend on IBM i:

Rather than completely re-platforming, many organizations are choosing to modernize by building new interfaces and analytics capabilities on top of their existing IBM i data. This pragmatic approach minimizes risk while maximizing the value of their legacy systems and historical data.

When do you use API vs ETL tools, and is there another solution?

The question should be, what is the acceptable latency of the data?

APIs provide immediate access to data and work well for retrieving several records, but they’re not ideal for analyzing 10 million records of data.
ETLs are great for moving data between platforms and applying transformation rules, but they come with latency issues depending on how frequently you run them.

The Compromise:

Change Data Capture (CDC) tracks and records data changes in databases, capturing inserts, updates, and deletes. It streams these changes in real-time to downstream systems for processing.

How it works:

First, perform a bulk load to transfer data between platforms. Then, set up an agent that monitors the source system and replicates any changes to the target system.

This approach combines the bulk transfer capabilities of ETL with the real-time benefits of APIs. It reduces latency to near real-time by continuously monitoring and synchronizing changes back to the source system.

Learning

📚 Essential IBM i manuals guide for developers

A comprehensive collection of essential IBM i manuals has been compiled, focusing on programming references, APIs, and system commands. The guide includes direct links to key documentation for DB2 SQL, ILE RPG, system APIs, and CL commands, making it easier for developers to quickly access official IBM documentation. Special attention is given to frequently accessed manuals for daily development tasks and troubleshooting.

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!