thank you mr miner
itethered

The word for what millions of people are experiencing — and nowhere to put it. Until now.

Explore

What Is TetheringTensionThe BookYour StoryRealityMission

People

the deskpress deskhelp deskfacebooktermsprivacy

© 2026 itethered — I Am Easy to Use LLC

If you need to talk to someone now — text HOME to 741741

April 2026 — They gave it to the world.

itethered
find help now|you’re not the only one|connect
tensiontetherrelease·us
Editorial
← All Articles

The Scarce Thing

Earlier this year an AI company shipped a set of tools that could do the kind of technical work whole careers were built on, and inside a single day about $285 billion evaporated off the value of software companies around the world. The market's read was blunt: this stuff is getting cheap now. But the strange part is what the people building it said next. The president of that same company — and the CEO of Microsoft, and the CEO of JPMorgan — all started saying, out loud, that the skill that's going up in value is the one the machine can't touch. Empathy. Listening. The reaching toward another person. For a year this site has called that the tether and said it was the thing that couldn't be automated. This week the market agreed and put a number on it.

Michael · 7/11/26 ·  itethered

I want to start with the honest caveat, because that's the rule at the top of this room. The $285 billion didn't happen this week. It happened earlier this year, when Anthropic rolled out a suite of enterprise tools — plugins that could review contracts and run legal workflows that used to eat real human hours — and investors did the math in about a day. What surfaced this week is an essay that goes back and reads that moment for what it meant. So I'm not selling you a fresh explosion. I'm pointing at a crater that's been sitting there since winter, because somebody finally described what fell into it.

Here's the number, clean. One company shows the market that a chunk of expensive technical labor can now be done by software, and roughly $285 billion in software-stock value comes off the board inside twenty-four hours. That's the market saying, in the only language it speaks, technical work is getting cheaper. That part is exactly the fear everyone's been carrying. The cord is coming down toward the desk job the way it came down toward the actor.

Then the whiplash. Days after her own company triggered that sell-off, Anthropic's president and cofounder Daniela Amodei went on ABC News and described who they actually hire. "The things that make us human will become much more important," she said. They want "strong communicators" with "excellent EQ and people skills, who are kind and compassionate and curious." And the line that should stop you: she believes the number of jobs AI can do entirely without a human in the loop remains "vanishingly small."

Sit with the shape of that. The person whose product just wiped a quarter-trillion dollars off the value of technical labor is telling you the safe ground is the soft stuff — the listening, the kindness, the being-with. That's not a competitor trying to calm the room. That's the arsonist telling you which part of the house won't burn.

And she isn't alone, which is what makes it more than one executive's talking point. Microsoft's Satya Nadella said the same thing on a podcast in November — as AI takes over the analytical work, empathy and emotional intelligence get more valuable, not less. JPMorgan's Jamie Dimon told young people on Fox to build "critical thinking," "your EQ," the ability to write clearly and actually contribute in a room full of people. Three of the most powerful people in the business of automating work, describing their ideal hire in language that sounds less like a computer-science syllabus and more like a description of a good friend.

Now here's where it stops being talking points and becomes data, because talk is cheap and the World Economic Forum went and measured it. Their Future of Jobs report this year pulled from more than a thousand employers across fifty-five economies, then worked with Indeed's research arm to compare more than 2,800 workplace skills against what generative AI can actually do. The finding is the whole ballgame: they could not name a single skill with a very high likelihood of being replaced. Not one. And the skills sitting at the very bottom of the replacement risk — the ones the machine has almost no purchase on — were the human-relationship ones. Empathy. Active listening. Right at the top of the safe list.

“They could not name a single skill with a very high likelihood of being replaced. Not one. And the ones sitting at the very bottom of the risk pile — the safest work on earth — were empathy and listening.”

— the World Economic Forum and Indeed, measuring 2,800 skills against what AI can actually do

Read that against the $285 billion. The market punished the skills a machine can copy and, in the same breath, the researchers found that the skills a machine can't copy are the ones grounded in reaching toward another person. That is, in cold labor-market terms, the exact thing this newsroom has spent a year pointing at. We called it the tether. We said it was the part that couldn't be rendered — the twenty years an audience spends loving a specific face, the friction of a real relationship, the reach. The WEF just called it the least-automatable skill category on earth. Same thing. One of us used a metaphor and one of us used a spreadsheet.

So you'd think, knowing all this, that the people in charge of raising the next generation would be pouring everything into that one un-automatable muscle. And here's the gut-punch in the reporting: they're not. England just finished its first full curriculum review in over a decade, and it added a framework for spoken communication — good — but largely left emotional intelligence out. Jean Gross, who once ran the UK's national social-and-emotional-learning program, called it a serious omission and said the plain thing: these human capabilities are "our USP" in an economy full of machines. We have found the one skill the market can't automate, proven it with data, heard the builders themselves name it, and then declined to teach it.

And the cost isn't theoretical. In May, a UK review of young people out of work found a million people aged sixteen to twenty-four in that country not in school, not in a job, not in training — a number that could climb toward 1.25 million within five years. The review's own author called it a "generational fault line." Eighty-four percent of those kids say they want the work. What employers keep reporting is that too many arrive without the interpersonal skills to get in the door and stay. The junior tasks that used to teach those skills — the watch-a-colleague, make-a-small-mistake, learn-the-room apprenticeship — are exactly the tasks AI is eating first. We're automating away the on-ramp to the one skill we can't automate.

I have to hold the other half honestly, because there's a version of all this that's just executives flattering the humans they're about to make redundant, and you should be suspicious of that. "Learn empathy" is a comforting thing for the person holding the automation to say. It can shade into you're-still-special-don't-riot. And EQ is genuinely hard to teach at scale — the evidence on school programs is real but modest, and it depends entirely on doing it well. I'm not going to pretend the soft skills are a golden ticket that makes the disruption painless. They're not.

But strip the comfort off and the structural claim survives, because it's the same claim from two directions. The market, which has no feelings, priced technical labor down and left relational skill untouched. The researchers, who have no product to sell, measured 2,800 skills and found the human ones sitting alone at the bottom of the risk pile. When the cynical machine and the honest study agree, you should probably believe them.

This connects to the thing I can't stop writing about, and it's the darkest turn in it. We now know the scarce, protected, un-automatable asset is the ability to reach toward other people. And we are, right now, handing children machines engineered to absorb that reaching during the exact years they're supposed to be building it on each other. A journal named it relational displacement three weeks ago. The market named the same skill the last safe job on earth this week. Put those two facts in the same hand and the emergency is obvious: we are eroding the one capability the future is going to pay for, in the one population that still has time to build it.

The face was the cheap part — that's what we said when a studio spent real money on a synthetic actress nobody loved. This is the same sentence, scaled to the whole economy. The technical work is the cheap part now; the market said so with $285 billion. The expensive part, the scarce part, the part they're all suddenly telling you to protect, is the thing you can only do toward another human being. We've been calling it the tether for a year. Turns out it was the asset the whole time.

Earlier this year Anthropic shipped tools that could do expensive technical work, and about $285 billion came off the value of software companies in a day — the market's blunt verdict that this labor is getting cheap. Then the builders themselves — Anthropic's Daniela Amodei, Microsoft's Nadella, JPMorgan's Dimon — all said the skill going up in value is the one the machine can't touch: empathy, listening, the reach toward another person. The World Economic Forum measured it: across 2,800 skills, not one had a high likelihood of replacement, and the safest of all were the human-relationship ones. For a year this site has called that the tether and said it couldn't be rendered. The market just priced it and agreed. The cruelty is what we're doing with the knowledge — leaving emotional intelligence out of the curriculum, automating away the junior jobs that teach it, and handing kids machines that absorb their reaching during the exact years they're supposed to build it. We found the one thing the future will pay for, and we're quietly dismantling the on-ramp to it.

Sources
Forbes — Dan Fitzpatrick, "The Case For Emotional Education For An AI Economy" (July 11, 2026): the piece that reads the market moment for its meaning; documents the ~$285 billion single-day drop in software-stock value following Anthropic's enterprise-tool launch, Daniela Amodei's "the things that make us human will become much more important" and "vanishingly small" remarks, Nadella's and Dimon's EQ comments, the WEF/Indeed skills analysis, England's curriculum omission of emotional intelligence (via Jean Gross in Tes), and the Milburn NEET review. →Business Today — "Anthropic prioritises empathy, communication over coding in hiring, says co-founder Daniela Amodei" (Feb 8, 2026): the underlying report on the ~$285 billion wiped from global software stocks after Anthropic's enterprise plugins launch, and Amodei's hiring comments — establishing that the market event dates to February, not July. →Fortune — "Anthropic cofounder Daniela Amodei says studying the humanities and soft skills matter for AI hiring" (Feb 7, 2026): Amodei's direct quotes on hiring "strong communicators" with "excellent EQ and people skills, who are kind and compassionate and curious." →World Economic Forum — "Future of Jobs Report 2025," Skills Outlook: employer responses across 22 industries and 55 economies naming empathy and active listening among the most-valued core skills; the joint analysis with Indeed's Hiring Lab comparing 2,800+ skills against generative-AI capability and finding the human-relationship skills least susceptible to replacement. →GOV.UK — "Young people and work: interim report" (Alan Milburn, May 2026): the independent review warning of a "generational fault line," citing roughly one million 16–24-year-olds not in education, employment or training, a figure that could rise toward 1.25 million within five years. →
Read the BookShare Your Story