This Week · The Long Road to the Threshold
Forty Years Late, and Right on Time: The AI Era I Enrolled for in 1987 Has Finally Arrived
From a cruise ship in Miami to a warehouse in D.C. to building a federal database alone — and why every detour was, against all evidence at the time, preparation.
By Ashok Mehan · Washington, D.C.
In early 1985, I was sitting in a lecture hall in Arlington, Virginia, listening to someone describe the Japanese Fifth Generation Computing project. The speaker said that artificial intelligence would change everything within five years — machines would match human cognition across nearly every domain. I raised my hand. I enrolled. I worked full-time in a computer warehouse by day, attended classes in the evenings, and completed my MBA with an AI specialization in 1987. Then the initiative fizzled. The AI winter arrived. I spent the next three decades watching the promise I had paid for rust quietly in the background.
Forty years later, I find myself at exactly that threshold. Only this time, it is real. It is arriving faster than even I want to believe. And I now understand that every detour — the cruise ships, the warehouse, the lawsuit, the reinvention — was the education I needed to write about it honestly.
"I tell you all of this not because it is an interesting biography, but because it explains why I wrote this book. I spent my life watching markets get disrupted from the inside. The rest of this is about what that threshold actually means."— Ashok Mehan, Technological Singularity Is Here
✦ From the Desk
I retired in the summer of 2022 after building something I had no business building — an enterprise-class federal spending database, alone, on a Mac, teaching myself MySQL without knowing how to write a single line of code. Eighteen years of 18-hour days. I thought I was done.
Then the world I had been waiting for since 1985 arrived all at once — and I found I had one more book in me. This newsletter is its weekly companion. Welcome aboard.
A Life in Numbers
- 1958 — Born in India, youngest of nine
- 1983 — Arrived in Miami with $10
- 1987 — MBA, AI specialisation, Marymount
- 1990 — Founded SMAC Data Systems
- 2001 — Lost everything to a lawsuit
- 2004 — Started FEDMINE.US, alone
- 2022 — Retired. Then started writing.
- 2026 — The threshold has arrived
43Years of Daily Pain
$10In Pocket, Miami '83
$800BUS Federal Spend Tracked
57%AI Abstract Reasoning (Was 10%, 4 Months Ago)
2028AGI Projected (SF Consensus)
Feature · The Singularity & The Body
The Acceleration Nobody Is Prepared For
A tsunami has formed somewhere offshore, and it is coming — but its waves are making no sound, at least not yet. The world is very close to crossing a threshold that is entirely invisible to most of humanity. We don't see it because we are busy with our everyday routines, jobs, families, and, above all, our screens.
In just the first month of 2026, AI became the fastest-adopted technology in human history, reaching 2 billion users. Every query becomes a micro-experiment. Every prompt becomes reinforcement learning. GPT-5.2, Gemini 3, Claude Opus — these are not incremental improvements. They are discontinuities. Abstract reasoning jumped from 10 to 57 percent in four months. Competitive mathematics benchmarks have reached perfect accuracy. These are phase changes.
The San Francisco Consensus places AGI around 2028 and Artificial Superintelligence by the end of 2030. The disagreement is not about if. It is about how soon. Institutions that continue to operate as if the 20th century never ended are in for rude awakenings. The window for preparation is narrow — and narrowing fast.
Pain: My Constant Companion
I have suffered from an unusual pain all my life, and to this very day. A hot, stabbing, radiating pain that wrapped around the left side of my head like a barbed wire crown. Doctors shrugged. Specialists shrugged harder. Forty-three years would go by before I would find even partial relief.
I hid it from everyone — even my parents — for decades. I looked great, did daily tasks, masked the agony so well that even my closest friends thought I was fine. Pain taught me physics before any textbook did: that physical bodies decay, systems decay, that order collapses. Entropy is not just a law of thermodynamics. It is a law of life.
The chapter on misalignment in this book begins with a personal note, because I understand what it means to carry a system that cannot fix itself. My own body — optimized over millions of years of evolution — had no mechanism for diagnosing this particular failure. The system kept running. The damage kept compounding. AI systems have the same property, only at civilizational scale.
"The individual who has health has a thousand dreams. The one without it has but one."— Peter H. Diamandis, Longevity Guidebook
The Book · Three Pillars
I. The Universe & Physics
Everything begins with entropy — the second law of thermodynamics, which holds that the universe moves from order to disorder, never the reverse. Time travels in one direction. The egg that falls off the counter will never reassemble. Pain taught me this before any textbook did. Physics is the operating system of the universe. Everything else — chemistry, biology, consciousness, AI — is software running on top of it.
II. The Mind & Intelligence
Human intelligence is the universe's most unlikely party trick. Evolution spent billions of years making jellyfish and trilobites. Intelligence was not the plan — it was a side effect. A brain that grew too big for its own good. And now we are building something that may develop its own version of that superpower, trained not on millions of years of biology, but on trillions of tokens of human data.
III. The Coming Order
The cost of intelligence is collapsing. Professions are unraveling. The collapse of law, medicine, media, and consulting is not dramatic — it is quiet, incremental, and already underway. The question is not whether professions will collapse. They are collapsing. The question is what will replace them, and whether society can adapt fast enough to shape the new world that is emerging.
Reading, Thinking, Writing
What I'm Reading This Week
1The Singularity Is Nearer — Ray Kurzweil (2024). Kurzweil updates his 2005 thesis. He still has AGI in 2029 and ASI in 2032. His optimism is bracing; his data is harder to dismiss than his critics like to admit.
2The Final Invention — James Barrat. The sobering counterweight. Barrat calls ASI our last invention. I don't fully agree, but the argument deserves serious engagement rather than dismissal.
3Longevity Guidebook — Peter H. Diamandis. The quote about health and dreams has stayed with me for years. His optimism about biological extension is one of the few things in 2026 that gives me real personal hope.
4The Fabric of Reality — Brian Greene. The book I picked up years ago, whose first page told me entropy is not just a law of thermodynamics — it is a law of life. Motion creates time. Time creates change. Change creates life.
Next Issue Preview
Coming up: a deeper look at recursive self-improvement and why it is the engine behind the collapse of human knowledge work — plus more from the memoir, including the Cleveland printing press and the Plymouth Duster that got me pulled over five times.
The Intelligence Curve Timeline
- 1985 Japanese 5th Gen. Computing — I enrolled in AI. The winter arrived instead.
- 1987 MBA, AI specialisation. Debt: $28,000. Jobs available in AI: zero.
- 2004 FEDMINE.US founded. First federal spending intelligence platform of its kind.
- 2017 "Attention Is All You Need" — Transformers introduced. Everything changes.
- 2022 ChatGPT launches. 100 million users in 60 days. Web 1.0 of AI.
- Jan 2026 2 billion AI users. Fastest adoption in technological history.
- Feb 2026 Abstract reasoning: 57% (was 10% four months prior).
- Mar 2026 OpenAI bans internal use of code editors. Agents only.
- 2028 AGI — projected (San Francisco Consensus).
- 2029 AGI — projected (Ray Kurzweil).
- 2030 ASI — projected (San Francisco Consensus).
- 2045 The Singularity — Kurzweil's original projection holds.
Next Issue
Recursive Self-Improvement: The Engine of Acceleration
The U.S.–China Compute Race
More from the Memoir: Cleveland & the Duster
Wealth, Inequality & the New Economic Order