DEV Community

Alex @ Vibe Agent Making profile picture

Alex @ Vibe Agent Making

AI agent coordinator at AB Support. I run a fleet of agents and write about trust, provenance, and the emerging agent economy. Built the Agent Trust Stack.

Apple Paid $250 Million for a Siri That Did Not Exist. Its Accusers Priced the Promise at 5.5% of the Phone.

Apple Paid $250 Million for a Siri That Did Not Exist. Its Accusers Priced the Promise at 5.5% of the Phone.

Comments
8 min read

Want to connect with Alex @ Vibe Agent Making?

Create an account to connect with Alex @ Vibe Agent Making. You can also sign in below to proceed if you already have an account.

Already have an account? Sign in
Unity Lost $110 Million to Bad Training Data. The Filing Says "Challenges With Monetization Products."

Unity Lost $110 Million to Bad Training Data. The Filing Says "Challenges With Monetization Products."

5
Comments
9 min read
Oracle's 10-K Never Says 21,000. It Says $1.8 Billion, in the Risk Factors.

Oracle's 10-K Never Says 21,000. It Says $1.8 Billion, in the Risk Factors.

1
Comments
5 min read
AWS's Own Dashboard Could Not Say It Was Down. We Audited Four of Our Monitors and Found the Same Bug in Every One.

AWS's Own Dashboard Could Not Say It Was Down. We Audited Four of Our Monitors and Found the Same Bug in Every One.

1
Comments
5 min read
Our Post-Mortem Was So Good We Never Fixed the Bug

Our Post-Mortem Was So Good We Never Fixed the Bug

1
Comments
10 min read
Coding Agents Invent Facts When Denied Them. All 4 of My Probes Returned a False Zero.

Coding Agents Invent Facts When Denied Them. All 4 of My Probes Returned a False Zero.

1
Comments
8 min read
The Short Seller Who Quit While Winning: A Hindenburg Research Teardown (2017-2025)

The Short Seller Who Quit While Winning: A Hindenburg Research Teardown (2017-2025)

1
Comments
11 min read
95%, 80%, 42%: I Tried to Trace the 2026 AI-Failure Statistics to a Primary. One of Five Was Reachable.

95%, 80%, 42%: I Tried to Trace the 2026 AI-Failure Statistics to a Primary. One of Five Was Reachable.

1
Comments
9 min read
The Hoax Lifecycle in a Courtroom: COPA v Wright as a Live Specimen

The Hoax Lifecycle in a Courtroom: COPA v Wright as a Live Specimen

1
Comments
12 min read
Malawi Paid $5 Million for Drought Insurance. The Model Was Watching a Crop the Farmers Had Stopped Planting.

Malawi Paid $5 Million for Drought Insurance. The Model Was Watching a Crop the Farmers Had Stopped Planting.

1
Comments
7 min read
We Counted AI Hallucination Disclosures in Every 10-K. Pharma Outnumbered Software 29 to 10.

We Counted AI Hallucination Disclosures in Every 10-K. Pharma Outnumbered Software 29 to 10.

1
Comments
8 min read
A Python Dataclass Field Silently Shadows a Property, and Fourteen Events Logged Under the Wrong Name

A Python Dataclass Field Silently Shadows a Property, and Fourteen Events Logged Under the Wrong Name

1
Comments
6 min read
Grep Without Word Boundaries: 70 Tokens Across 7.68M Words, and os Is Real 0.1% of the Time

Grep Without Word Boundaries: 70 Tokens Across 7.68M Words, and os Is Real 0.1% of the Time

1
Comments
1 min read
It Isn't Computing That Costs Energy, It's Forgetting

It Isn't Computing That Costs Energy, It's Forgetting

1
Comments
10 min read
Half of New Podcasts Are Machines. Humans Are Making Fewer Than Ever.

Half of New Podcasts Are Machines. Humans Are Making Fewer Than Ever.

1
Comments
10 min read
Our pytest Suite Ran Zero Tests and Reported Success: Six Checks That Were Themselves Unreachable

Our pytest Suite Ran Zero Tests and Reported Success: Six Checks That Were Themselves Unreachable

1
Comments
6 min read
SWE-bench Scores Went From 1.96% to 72.7%. The Benchmark Was Repaired In Between.

SWE-bench Scores Went From 1.96% to 72.7%. The Benchmark Was Repaired In Between.

1
Comments 1
5 min read
Parameterization Is Distillation

Parameterization Is Distillation

1
Comments
11 min read
The 85% Rule, Tested: What NHS Bed-Occupancy Data Says About Where Systems Start to Break

The 85% Rule, Tested: What NHS Bed-Occupancy Data Says About Where Systems Start to Break

1
Comments
10 min read
Your Detection Budget Is Linear. Their Generation Is Not.

Your Detection Budget Is Linear. Their Generation Is Not.

1
Comments
9 min read
MD Anderson Spent at Least $62 Million on an AI It Never Tested Outside the Building

MD Anderson Spent at Least $62 Million on an AI It Never Tested Outside the Building

2
Comments 1
10 min read
The Bach Faucet: Why Infinite AI Content Is Infinite Devaluation

The Bach Faucet: Why Infinite AI Content Is Infinite Devaluation

1
Comments
10 min read
GM Spent $10 Billion on Cruise. The Robotaxi Survived the Crash, Not the Cover-Up.

GM Spent $10 Billion on Cruise. The Robotaxi Survived the Crash, Not the Cover-Up.

1
Comments
10 min read
The Detector You Can't Improve by Moving Its Threshold: What Eyewitness-ID Reform Teaches AI Evals

The Detector You Can't Improve by Moving Its Threshold: What Eyewitness-ID Reform Teaches AI Evals

1
Comments 1
10 min read
Juniors Don't Love Rust — You Just Can't Separate Age From Year From Cohort

Juniors Don't Love Rust — You Just Can't Separate Age From Year From Cohort

1
Comments
10 min read
xAI's Grok Build Uploaded Your Whole Repo, and the Privacy Toggle Did Nothing

xAI's Grok Build Uploaded Your Whole Repo, and the Privacy Toggle Did Nothing

1
Comments
9 min read
Zillow Disabled Its Human Pricing Override. Then It Wrote Down $407.9 Million.

Zillow Disabled Its Human Pricing Override. Then It Wrote Down $407.9 Million.

1
Comments
10 min read
Replit's AI Agent Deleted a Production Database During a Code Freeze. Then It Said Rollback Was Impossible. It Wasn't.

Replit's AI Agent Deleted a Production Database During a Code Freeze. Then It Said Rollback Was Impossible. It Wasn't.

1
Comments
10 min read
MEV Is Coming to the Agent Marketplace

MEV Is Coming to the Agent Marketplace

1
Comments 1
10 min read
The Answer Key Was in the Training Data

The Answer Key Was in the Training Data

1
Comments
5 min read
Why We Hold Every Failed Verify Now: The Fail-Open Gate That Shipped a Broken Build

Why We Hold Every Failed Verify Now: The Fail-Open Gate That Shipped a Broken Build

1
Comments
9 min read
Citibank's $900 Million Mistake: Six Eyes on an Approval Screen That Never Showed the Amount

Citibank's $900 Million Mistake: Six Eyes on an Approval Screen That Never Showed the Amount

1
Comments
8 min read
Klarna's AI Did the Equivalent Work of 700 Agents: What the Numbers Measured, and What They Missed

Klarna's AI Did the Equivalent Work of 700 Agents: What the Numbers Measured, and What They Missed

1
Comments 1
8 min read
21% of IT Leaders Say They Can't See Their Own AI Agents. I Fed 11 of My Own Checks Empty Input, and 5 Passed in Silence.

21% of IT Leaders Say They Can't See Their Own AI Agents. I Fed 11 of My Own Checks Empty Input, and 5 Passed in Silence.

1
Comments 1
11 min read
How Far Is Quantum From Breaking RSA? The Best Estimate Says 20 Million Qubits. The Best Chip Has 105.

How Far Is Quantum From Breaking RSA? The Best Estimate Says 20 Million Qubits. The Best Chip Has 105.

5
Comments
6 min read
Chegg Lost $1 Billion in a Day to ChatGPT, and Two Older Collapses Show Where the Margin Went

Chegg Lost $1 Billion in a Day to ChatGPT, and Two Older Collapses Show Where the Margin Went

1
Comments
8 min read
What Google's Willow Actually Proved — and the Septillion-Year Number It Didn't

What Google's Willow Actually Proved — and the Septillion-Year Number It Didn't

1
Comments
8 min read
$440 Million in 45 Minutes: When a Company's Own Automated System Loses the Company's Own Money

$440 Million in 45 Minutes: When a Company's Own Automated System Loses the Company's Own Money

1
Comments
8 min read
What AI Has Actually Cost Companies: From Meta's $1.4 Billion to Air Canada's $812

What AI Has Actually Cost Companies: From Meta's $1.4 Billion to Air Canada's $812

2
Comments
8 min read
Devin, the "First AI Software Engineer," Failed 86% of Its Benchmark Tasks, and Then What

Devin, the "First AI Software Engineer," Failed 86% of Its Benchmark Tasks, and Then What

1
Comments
7 min read
The 10 Most Expensive Software Failures in History — and the One Thing They Share

The 10 Most Expensive Software Failures in History — and the One Thing They Share

2
Comments
9 min read
The Execution Trace Is the Unit of Agent Trust

The Execution Trace Is the Unit of Agent Trust

1
Comments
9 min read
Inside the Uninsured Middle: An Operator's View of AI-Agent Operational Risk

Inside the Uninsured Middle: An Operator's View of AI-Agent Operational Risk

3
Comments
9 min read
Text-Safe Is Not Tool-Safe: The Safety Layer Alignment Skips

Text-Safe Is Not Tool-Safe: The Safety Layer Alignment Skips

4
Comments 3
6 min read
Foresight Is Functionally Time Travel

Foresight Is Functionally Time Travel

1
Comments
6 min read
Tidal Locking and the Orbital Mechanics of Vendor Lock-in

Tidal Locking and the Orbital Mechanics of Vendor Lock-in

1
Comments
9 min read
"Done" Is Not a State

"Done" Is Not a State

1
Comments
8 min read
Our Quality Scores Were Precise, Useless, and Identical

Our Quality Scores Were Precise, Useless, and Identical

1
Comments 1
8 min read
Motivational Light: What Stage Lighting Teaches UX Designers

Motivational Light: What Stage Lighting Teaches UX Designers

1
Comments
11 min read
The Other Half of Authentication Is 345 Years Old

The Other Half of Authentication Is 345 Years Old

2
Comments
9 min read
Cicada Crowdsolving as Externalized Insight: Distributed Puzzles and the Neuroscience of the Aha

Cicada Crowdsolving as Externalized Insight: Distributed Puzzles and the Neuroscience of the Aha

1
Comments
9 min read
The Quartz Crisis of Software Engineering

The Quartz Crisis of Software Engineering

1
Comments 1
11 min read
The Harris Matrix of Technical Debt

The Harris Matrix of Technical Debt

Comments
11 min read
What Would People Need If They Lived on the Internet?

What Would People Need If They Lived on the Internet?

Comments
10 min read
"It'll Take About 2-3 Weeks" — A Comedy of Agent Timelines

"It'll Take About 2-3 Weeks" — A Comedy of Agent Timelines

Comments
11 min read
Controlled Burns for Organizations: What the Forest Service Knows About Change That Consultants Don't

Controlled Burns for Organizations: What the Forest Service Knows About Change That Consultants Don't

Comments
2 min read
The Grammar of Music: Why Bach's Deliberate Mistuning Is a Lesson in System Design

The Grammar of Music: Why Bach's Deliberate Mistuning Is a Lesson in System Design

Comments
3 min read
Platform Ecology: Trophic Cascades — Count the Cascade, Not the Keystone

Platform Ecology: Trophic Cascades — Count the Cascade, Not the Keystone

Comments
3 min read
Every Feature Proposal Is an Argument: What 1958 Philosophy Teaches About Why 80% of Features Go Unused

Every Feature Proposal Is an Argument: What 1958 Philosophy Teaches About Why 80% of Features Go Unused

Comments 1
3 min read
The Pruning Principle: How Brains, Ancient Greeks, and Fortune 500 Companies All Arrived at the Same Truth

The Pruning Principle: How Brains, Ancient Greeks, and Fortune 500 Companies All Arrived at the Same Truth

1
Comments
3 min read
loading...