The Evolution of Expertise — and What Comes Next.
For decades, computing has moved information, knowledge and decision-making closer to the people who need it. From centralized systems and expert rules to intelligent automation and AI, each generation has expanded what technology can do.
From a desk full of terminals to intelligent systems.
The journey began with a practical problem: important information lived in different computers, on different networks, using different architectures and communications. Getting to the information was difficult. Understanding what to do with it was harder.
The interface kept changing. The direction remained the same: move information, expertise and action closer to the point of need.
Many systems. One workstation.
Telecommunications desktops could contain several terminals: DEC asynchronous ASCII, IBM 3270 mainframe access, emerging UNIX systems and early PCs with little or no LAN connectivity. NCR UNIX technology could emulate multiple host communications and present character-based sessions on a single workstation. The first step was bringing the systems together.
From access to intelligence.
The next generation captured defined call-center logic and applied it to data from multiple systems. A simple “no dial tone” report could trigger billing checks, known-trouble checks, line testing, interpretation of results, customer questions and dispatch decisions—simultaneously or in sequence as the data required.
The architecture changed. The knowledge survived.
The concept was rewritten in Java as the Intelligent Customer Advocate System. Rules could also continue monitoring known troubles after the customer interaction ended. When a common problem cleared, related dispatches could be removed from the queue—fix one problem and eliminate many unnecessary truck rolls.
Modernize, then freeze.
Y2K opened technology budgets across telecommunications. Hardware and software were replaced or upgraded, then environments were frozen as the millennium approached. When the date rolled over successfully, much of the modernization money had already been spent.
Integrate. Surround. Reuse.
With large legacy investments already in place, the business case increasingly favored surrounding and enhancing existing systems rather than replacing everything. As telephone companies merged, established automation could be upgraded and expanded as a common standard.
The market changed.
The traditional copper and landline business was declining. SMExpert was formed to carry the principles of intelligent automation into another generation. A major organization committed to a trial, but the environment proved unfavorable and the project did not proceed to an actual trial. The commercial effort ended with retirement. The evolution of the idea did not.
Access to data isn't the same as expertise.
A subject matter expert knows which information matters, where to find it, how pieces relate, what a result means and what should happen next. Expert systems attempted to capture that reasoning and make it repeatable.
Integration gave the system access to information. Expertise gave the information meaning. Automation turned that meaning into action.
Put intelligence as close to the point of need as it makes sense.
Across industries, computing has repeatedly moved outward—from specialists and centralized systems toward the people doing the work and ultimately toward the customer, patient or individual.
Healthcare
Central data center → major departments such as laboratory and pharmacy → nurses and physicians → patient portals and personal devices → AI-assisted understanding and action.
Telecommunications
Specialized back offices → integrated call centers → field technicians → web self-service → intelligent systems capable of reasoning across increasingly complex networks and services.
The Principle
Move information, capability and expertise outward when doing so improves the process. Keep expert judgment where risk, complexity or economics say it belongs.
AI changes the scale of what is possible.
Traditional expert systems required people to identify knowledge and explicitly encode rules. Today's AI can work across vast amounts of structured and unstructured information, interact in natural language and assist with reasoning that would have been impractical to program rule by rule.
More Knowledge
Documents, conversations, historical cases, databases, images and operational information can contribute to a much broader knowledge environment.
Natural Interaction
The interface can shift from knowing which application and screen to use toward simply describing the problem or desired outcome.
From Assist to Act
With appropriate authority and safeguards, intelligent systems can move from explaining and recommending toward carrying out portions of a process.
AI changes the possibilities dramatically. The fundamentals—expertise, trustworthy information, integration, process and value—still matter.
The business case has always mattered.
Technology makes something possible. A business case determines whether it is worth doing. From eliminating unnecessary dispatches to extending scarce expertise, the useful question is not simply “Can we automate it?” but “What measurable value does automation create?”
The winners won't necessarily be the organizations with the most AI. Value will come from finding the right problems to solve with it.
Blue sky. What comes next?
This is deliberately the speculative part of the story: not promises, but questions and predictions worth watching as AI moves from an extraordinary new tool toward infrastructure woven into everyday work.
The Interface Fades
People increasingly describe the outcome they want instead of navigating a collection of applications, screens and menus.
Expertise Becomes Ubiquitous
More people will have immediate access to specialized knowledge, changing when and why a human specialist must enter a process.
Self-Service Becomes Intelligent Service
Instead of giving people portals and expecting them to understand the process, intelligent systems increasingly understand what the person is trying to accomplish.
Legacy Systems Live Longer
AI may make it economically attractive to surround older systems with intelligent layers rather than replace every underlying platform.
Authority Becomes the Big Question
The issue moves beyond who can see the information toward what an intelligent system should be permitted to decide and do.
ROI Separates Novelty From Value
AI will enable countless technically impressive ideas. The enduring ones will solve worthwhile problems at an acceptable cost and risk.
The future is wide open.
What should we do with it?
Five decades watching computing move closer to the problem.
SMeXpert reflects experience spanning healthcare computing beginning in the 1970s and telecommunications technology from the 1980s forward—through centralized systems, UNIX, expert systems, Java, integration, automation, self-service and the emergence of artificial intelligence.
This site is not a solicitation for consulting work. It is a continuing exploration of how technology captures and extends human expertise, how those capabilities move outward, and how strong business cases and measurable ROI turn interesting technology into useful technology.
“The technology changed. The objective didn't.”
SMeXpert.com — Subject Matter Expertise, intelligent automation and the future of AI.