AGI and the Hairs on the Back of My Neck

That Feeling You Can’t Quite Name

Let me describe something and you tell me if it sounds familiar.

You’re reading the news. Another AI announcement. Another company laying off 15% of its workforce and calling it “restructuring.” Another startup demo that makes something you spent 20 years learning look trivially easy. And somewhere in the back of your skull — not a thought, exactly, more like a physical sensation — the hairs on the back of your neck stand up.

Not because of that one headline. Because of all of them together. Because you can feel the weight of something enormous gathering speed, and even though nothing has technically “happened” yet — no mass unemployment event, no robot uprising, no single moment you could point to — something in your nervous system is telling you: it’s close.

If that’s you, I need you to hear something.

You’re not anxious. You’re perceptive.

That tingling sensation is your pattern recognition doing exactly what it evolved to do. You’re detecting a pressure change in the atmosphere before the storm arrives. And the data — 480 studies, 150,000 survey respondents, behavioral data from 180 million developers, task-level analysis of 81,000 workers across 159 countries — confirms what your body already knows.

The storm is real. It’s closer than most people think. And the window to position yourself is right now.

Let me show you what the air pressure actually looks like.

The Loaded Spring: What AI Can Do vs. What We’re Using It For

Here’s the number that should make your neck tingle more, not less.

Anthropic — one of the three companies building frontier AI models — published hard data on what their system can actually do when given real work tasks. Not demos. Not cherry-picked examples. Actual task completion across the entire economy.

The results:

• 68% of tasks: fully handleable by current AI.
  Not future AI. Not AGI. The models running right now, today, could complete 68% of the cognitive tasks humans currently do.
• 97% of tasks: AI can at least contribute meaningfully.
  Only 3% of knowledge work is genuinely impossible for current AI. Three percent.
• 33% of that capacity is actually being used.
  The observed utilization rate across enterprises. Two-thirds of what AI could do is sitting idle. Not because it doesn’t work. Because organizations haven’t figured out how to plug it in yet.

Read those numbers again. The gap between 68% handleable and 33% utilized — that’s not a technology problem. That’s the loaded spring. That’s the thing your nervous system is detecting.

No breakthrough is required. No new model needs to be invented. No science fiction needs to become science fact. The capability is already here, right now, waiting for organizations to simply use it.

When people ask “how close are we to AGI?” they’re asking the wrong question. The better question is: how close are we to organizations actually using what they already have?

And the answer to that question is: closer than you think. Because of three forces that are about to collide.

Why It’s Going to Feel Like It Happens All at Once

The tingling you feel isn’t just about AI getting better. It’s about three separate forces converging at the same time. Each one would be significant alone. Together, they create that feeling of inevitability — the sense that something is about to snap.

Force 1: The Corporate Herd

Here’s something they don’t teach in business school, but every manager knows intuitively:

You can never be punished for doing what everyone else did.

If every Fortune 500 company buys AI agents and it fails, no individual VP gets fired. They all shrug and say “nobody could have predicted it.” But if your competitor buys AI agents and it works, and you didn’t? You’re gone by Q3.

This asymmetry — where the career risk of not adopting is catastrophic and the career risk of adopting is zero — guarantees a purchase wave. It doesn’t matter if the ROI is uncertain. It doesn’t matter if 85% of current AI usage creates no business value. The herd is moving, and staying behind the herd is the only unforgivable sin.

Accenture booked $1.2 billion in GenAI consulting revenue. Agent adoption across enterprise jumped 282% in a single year. The herd has started moving.

Force 2: The Pipe Is Getting Bigger

$700 billion. That’s how much the cloud hyperscalers — Amazon, Microsoft, Google — are spending on AI infrastructure. Not over a decade. Right now.

Think of it this way. The AI capability already exists (Force 1 makes everyone buy it). But the pipes it flows through — the data centers, the chips, the network bandwidth — have been the bottleneck keeping everything from deploying at scale.

That bottleneck is being obliterated with $700 billion in concrete and silicon. When those data centers come online — and they’re coming online now, this year, next year — the infrastructure constraint vanishes. The loaded spring loses its last friction point.

Meanwhile, inference costs have dropped 99% in a single year. What cost $100 to compute in early 2025 costs $1 today. The thing stopping most companies from running AI on every task, every workflow, every customer interaction was price. That excuse just evaporated.

Force 3: The Tools Are Getting Stupidly Easy

This one’s the sleeper. It’s not dramatic. It doesn’t make headlines. But it’s the force that actually triggers the storm.

Right now, using AI effectively in a business workflow requires technical people to set it up, configure it, integrate it, maintain it. That’s why 50% of enterprise AI agents are siloed — some IT team built them for one department and nobody else can touch them.

But the tools are simplifying. Fast. Every AI company on earth is racing to make their product something a non-technical manager can deploy in an afternoon. Not because they’re altruistic — because the company that cracks “easy” captures the entire market.

When the tools become self-serve — when deploying an AI agent is as easy as setting up a Zoom meeting — that 33% utilization rate doesn’t slowly climb to 40% and then 50%. It lurches. Because suddenly every department head, every regional manager, every small business owner can do what only the IT department could do before.

That lurch is the storm. That’s what your neck is tingling about.

Why Most People Will Be Surprised (And You Won’t)

There’s a reason most people don’t feel the tingling. And understanding it is your informational edge.

When you read that “88% of companies are using AI,” your brain files that under “it’s already happened.” The headline creates a feeling of completion. Like we’re already in the AI era and it’s no big deal.

But dig one layer deeper and you find the real number: only 24% of those companies have AI integrated into actual workflows. The other 64%? Someone in marketing has a ChatGPT subscription. A developer used Copilot once. The CEO mentioned AI in an earnings call.

This is what we call the Adoption Mirage. The headline number — 88%! — makes the world look like it’s already transformed. The reality is that we’re barely past the experimentation phase. And that gap between perception and reality is exactly why the storm will feel sudden when it arrives.

Most people look at the 88% and think: “OK, AI happened. We absorbed it. Things are fine.”

You look at the 24% and think: “We haven’t even started.”

You’re right.

The people who will be caught off guard are the ones reading headlines. The people with the tingling are the ones reading the second number in every statistic.

And here’s the thing that makes your instinct especially valuable: the mirage is about to crack.

There is a predictable phase in every technology cycle where public ridicule peaks just before the technology becomes undeniable. We saw it with the internet in 1998 (“nobody will buy things online”). We saw it with smartphones in 2007 (“nobody needs email on their phone”). We are in this phase right now with AI.

The ridicule is getting louder. The adoption is accelerating underneath it. The gap between public perception and private deployment is widening. And when it snaps shut — when the mirage cracks — it will feel like it happened overnight.

You can feel it coming. That’s not anxiety. That’s a six-month head start.

The Storm Map: What Gets Hit, and When

OK. So the storm is real, it’s coming, and your instincts are correct. Now the question that actually matters for your portfolio: which industries are in the eye wall, which are in the outer bands, and which are watching from the coast?

We track 28 industries across 171 cross-industry effects and 5 time horizons. Here’s what the data shows about the sequence.

The Eye Wall (Already in the storm)

The Outer Bands (Storm arriving 2-4 years)

Watching From the Coast (5-10 years out, but don’t sleep)

Three Signals That Tell You the Storm Has Arrived

You don’t need to guess when the capability overhang collapses. There are concrete, measurable signals. When you see these, the storm has made landfall.

• Signal 1 — Watch now
  Consulting revenue peaks, then drops quarter-over-quarter
  Right now, consulting revenue is surging — Accenture’s $1.2B GenAI bookings, Deloitte and McKinsey scaling AI practices. This is the “getting rich helping people adopt it” phase. When this revenue peaks and starts declining QoQ, it means the tools have become self-serve enough that companies don’t need help anymore. That’s the inflection point. Watch Accenture and Infosys quarterly earnings like a hawk.
• Signal 2 — Watch 2027
  Entry-level white-collar hiring falls below replacement rate
  It’s already down 14% for 22-25 year-olds in AI-exposed occupations. Down 20-30% in professional services specifically. When this crosses the replacement rate threshold — when companies are hiring fewer juniors than they’re losing to attrition — it means AI has moved from “augmenting” to “replacing” in practice. Not in press releases. In headcount.
• Signal 3 — Watch 2028
  A Fortune 500 CEO says “we don’t need that department anymore” on an earnings call
  Not “we’re using AI to make that department more efficient.” Not “we’re restructuring.” The actual words: we eliminated this function because AI does it now. When a CEO says that to Wall Street analysts and the stock goes up, the storm has arrived. Everything after that is just the rain.

What to Do Before the Rain

Here’s where I have to be honest with you, because honesty is the only thing worth selling.

Nobody can predict stock prices. Anyone who says otherwise is selling you something. And I’m specifically not doing that — nothing on this page is investment advice, and you should make every financial decision with a professional who knows your situation.

But here’s the thing people get wrong about prediction.

You’d be insane not to bet your house at 2:1 odds that the Buffalo Bills will score at least one touchdown in the 2026-2027 NFL season. You’d be equally insane if you did bet your house that you could predict the temperature in Timbuktu to two decimal places three and a half years from now.

Most people hear “prediction” and think Timbuktu. They picture someone claiming to know exactly which stock hits $400 on March 15th, 2029. That’s not what we do. That’s not what anyone can do.

But the Bills touchdown? That’s knowable. And there is an enormous, valuable middle ground between those two extremes — more than most people realize. You can know that cloud infrastructure demand will grow. You can know that per-seat SaaS pricing is structurally breaking. You can know that consulting revenue will peak and then decline. You can know that 95% cognitive-labor industries transform before 30% physical-labor industries. You can know the sequence.

That middle ground — the space between the obvious and the unknowable — is where we operate. And it’s much wider than the “nobody can predict anything” crowd wants you to believe.

What our data shows about that sequence: not which stocks go up, but which industries transform first, second, third — and what that means for where value migrates over time.

The most counterintuitive thing in this analysis: the industries that scare you most right now are the ones closest to the other side of the storm. Cloud platforms, chip designers, enterprise software — they’re in the eye wall. It’s violent. It’s volatile. And they’re the ones that emerge transformed and dominant.

The industries that feel safest — the ones with big regulatory moats, physical labor floors, and “AI can’t do what we do” confidence — they’re just further from the eye wall. The storm reaches them too. They just don’t know it yet.

Your Edge Is That You Can Feel It

I want to end where we started. With that tingling.

Most investors are reading the headlines. “88% of companies use AI!” Great, sounds like it’s priced in. Move on.

Most analysts are building models. Spreadsheets with linear growth curves that assume AI adoption continues at its current pace. Neat, orderly, wrong.

Most people have turned the tingling into noise. Background anxiety they try to ignore. Another thing to worry about in a world full of things to worry about.

But you read the 88% and felt something was off. You noticed the junior hiring numbers. You caught the gap between the headlines and the reality. And instead of dismissing the tingling, you followed it. To this page. To the data.

That’s not common. That’s the thing worth knowing. Most humans suppress pattern recognition when it produces uncomfortable conclusions. You didn’t. You leaned in.

The storm is coming. You can feel it. The data confirms it. And the question isn’t whether you’re right — you are — it’s whether you use that head start.

We track 28 industries, 171 cross-industry effects, and 5 time horizons — updated continuously as new data arrives. The model doesn’t predict stock prices. It maps the storm. Which industries are in the eye wall. Which are in the outer bands. Which think they’re watching from the coast but aren’t.

You’ve been feeling the pressure drop for months. Here’s the barometer.

Related reports: The Human Bottleneck  |  Hype vs. Reality  |  AI Hatred and Corporate Adoption  |  How We Score 28 Industries