AI as the missing piece in the “Exponential Web”

3/10/2026

Retroactive Improvement

Structured data gives your ideas a skeleton; APIs give them muscles; retroactive improvement gives them longevity.
-Boom Hot Take

Introduction

Disruptive Technology

There is no shortage of serious content about AI on LinkedIn. Precision architectures, transformer diagrams, model benchmarks, responsible governance frameworks — the platform is doing its job as a professional commons. And rightly so. Careers and reputations are at stake. The discourse reflects that gravity.

Use Every Tool You Have

Modern Web Information Standards

But I want to make a case for something different. Not less rigorous — just less guarded. I want to make a case for joyful inquiry: the kind of learning that happens when you poke at things that don't matter yet, in spaces where failure is free, and where the most instructive discovery is often a resounding, illuminating no.
I've been doing exactly that with modern digital information standards since the Y2K bug required new (human intensive) data interfaces. Before I took this job, I had already had the joy of pretending to be doing computer work in high school and experimenting with a (now defunct command prompt) to send messages to my classmates that frequently triggered disruptive adolescent laughter. This joy led me to pursue similar breakthroughs (Y2K Data Modernization) when I realized that I could still do "net send" in the corporate world.
However, that joy continued in 1999. I not only learned that my company still had "net send", but that the computer could make my job exponentially simple. Instead of sending that file to a digital printer and a team of data entry resources... I used a common (text delimited file) data format to migrate information to the Y2K compliant platform. Our job transitioned to quality assurance where we verified the accuracy of imported data and minimized human intensive (error-prone) data entry.
I've been doing exactly that with AI for a while now, joyfully seing how AI can:

Enhance virtual (3D office) presentations that can present real-world scenarios with "digital twin" hybrid office technology (REAL WORLD SOLUTION VISUALIZATION - For Every Idea LLC)
Recursively enhance old blog articles (Ultra Lean Technology - For Every Idea LLC) with emerging AI Narrated technology summaries
Produce an AI Accelerated MVP from a reusable prompt library (AI Prompt Engineering - For Every Idea LLC) for continuous improvement
A 2026 educational video presentation of how AI ready ideas accelerate exponential breakthroughs ( New Year - Exponential Opportunity - For Every Idea LLC)

For now, these inquiries are mostly a story about the 10,000 things that AI can't do as good as a modern business professional. A modern business professional knows that if you can't describe your idea in under 30 seconds or at least make it fit on a simple napkin sketch ... then it is doomed to fail (see Behaviors for Getting Things Done - For Every Idea LLC). Fortunately, with modern information standards we also have the option utilize machine-readable data formats to significantly enhance the messy version 01 of our transformational idea with thousands of meaningful updates that enhance the Currency, Relevance, Authority, Accuracy and Purpose with a self-improving framework of (web-based) API that can seamlessly update hundreds of people with thousands of incremental updates.

One Rule is that you only get one penny of compute and only a few kilobytes of bandwidth

That could have been an email

One thing that is not joyful is the administrative process of onboarding new resources. Even though the web initially made the job market more accessible (How can I communicate more effectively with a communication model?) the volume of phone calls, email attachments, and video interviews required to make this happen has exponentially increased from 2006 and has made it less enjoyable to be in the fractional workforce (where projects can be completely delivered in less than 90 days). Spending a few days on administrative cold-calls calls with strangers that need your resume, portfolio, and other personal information before you can commence work on a 90 day contract is a burden that could be replace with a trusted link that simply handles all the administrative tasks with a self-improving framework of (web-based) API capabilities with less than one penny of compute and a few kilobytes of bandwidth to send the link. This framework accelerates the delivery of urgent tasks with the exponential power of a (web-scalable) platform to seamlessly update hundreds of people with only a few kilobytes of bandwidth.
In a world where things can be accomplished in less than 30 seconds, navigating phone trees to get in touch with someone is too long. With modern office technology and cloud-first meeting materials, an AI assistant can simply schedule a meeting and quickly get everyone on the same page within minutes.

The public performance problem

March 1989 Tim Berner's Lee proposal for the (www) Hyperlink

For too long this disruptive (web-based) technology has been underutilized in favor of single-use (paper-based) technology. The reliability of physical office supplies and human-intensive governance has dominated the modern world for more than 30 years (Thirty Eight Years Later - For Every Idea LLC) for a variety of reasons. This monolithic inertia has suppressed many incremental improvements due to the cost prohibitive expense of consuming more (single-use) office supplies and human-intensive governance procedures. Even though a modern computer can seamlessly implement thousands of updates with less than a penny of compute or a share a dynamic web link with only a few kilobytes of bandwidth, most "version 01" presentations are continously discarded instead of enhanced with the modern "paperless" cloud-first technologies at our fingertips.
However, the industry adoption of (modern-web) AI capabilities is finally accelerating the adoption of 30 year-old (web interface) technology that can transcend the inertia of human intensive logistics and expensive monolithic office overhead with a reusable URL that can seamlessly replace the physical logistics of monolithic worksites with virtual office technology.

Disruptive opportunity that No One Talks About

Retroactive Improvement

Retroactive improvement only works when ideas are stored as meaning, not formatting. Ideas that are locked in single-use (read-only) office supplies are not eligible for AI accelerated enhancement. However, since the introduction of AI accelerated solutions, more professionals are authoring those messy versions 01 ideas with reusable (web-based) technology that can dynamically improve with bi-directional (Services - For Every Idea LLC) collaboration that can retroactively apply lessons learned and professional epiphanies with only a few kilobytes of bandwidth and less than a penny of compute.
Instead of sifting through thousands of versions 01 ideas scribbled on the back of a napkin that failed (From Scribbles to Solutions: The Future of Idea Delivery - For Every Idea LLC) because of a subsequent lesson learned that could be retroactively applied... a modern professional (or their AI accelerated technology) can resurrect that reusable (machine-readable) idea and polish with exponentially improving (modern-web) technology at the press of a button. This automation represents a transformational shift in the professional landscape.
To be clear, this disruption isn’t that AI can suddenly write a spreadsheet formula or summarize an email thread. The disruption is than any meaningful improvement can be implemented with less than a penny of compute and only a few kilobytes of bandwidth. For a modern professional with a laptop and a credit card, this improvement can be exponentially accelerated with automated workflows that can provision human and automated resources within minutes. This is in stark contrast to the classic start-up that needs to suddenly acquire new hardware and resources to meet the needs of a new project.

0 Comments