"That's administratively ugly."
"Good."
"You think ugly is a design requirement."
"Sometimes friction is the safety feature."
The paragraph changed.
The university invitation remained in Isaac's inbox for two weeks.
Finally, he accepted.
He drove to Austin instead of flying.
The lecture hall held engineering students, faculty, and a scattering of industry professionals who had heard enough rumors to be curious. The event title embarrassed him: WHEN THE DATA DISAGREES: TECHNICAL DISSENT AND ENGINEERING COURAGE.
Isaac removed the word courage from his opening slide.
"I don't know much about courage," he told the room. "I know about incentives, documentation, and being difficult at the correct time."
The students laughed.
Then he told them about system boundaries.
Not OBSIDIAN-7 by name. Not classified details. He used a generic propulsion example and focused on the intellectual mistake that mattered: evaluating components separately when the failure lived in their interaction.
A student asked, "How do you know when to risk your career over a technical disagreement?"
Isaac considered the question carefully.
"You don't start by risking your career. You start by making the disagreement testable. Write down what you think will happen," Isaac said.
"Write down why. Give people a way to prove you wrong. If the evidence still says the system is unsafe and the organization wants your signature anyway, then the question isn't whether you should risk your career. The question is what your signature means if you don't."
Another student asked whether he regretted refusing to sign.
"I regretted the consequences. That's different."
That answer quieted the room.
Afterward, a young woman waited until most people had left.
She introduced herself as Marisol Vega, a second-year graduate student in mechanical engineering.
"My father fixes farm equipment," she said. "People in my family think I'm wasting time doing research because they understand machines you can touch and I work on simulations. Sometimes I think they're right."
Isaac smiled slightly.
"Simulations are tools. Wrenches are tools. Both become dangerous when the person using them forgets what problem they're for."
"Did working in the garage make you a better engineer?"
He looked at her.
"Yes."
"Because it kept you practical?"
"Because broken things don't care about prestige."
She wrote that down.
Isaac noticed and sighed.
"Don't quote me."
"Too late."
On the drive home, he realized he had enjoyed the lecture.
That annoyed him enough to tell no one.
Eleanor found out anyway.
"Heard you were inspirational."
"Who told you?"
"Small profession."
"Worse than Harlow."
"Much."
The Houston position remained open.
Eleanor did not pressure him every week.
She pressured him every third week, which she considered restraint.
"You could shape the team," she said during one call.
"I can shape things here."
"You could do both."
"Houston is six hours away."
"There are roads."
"I hate Houston traffic."
"Everybody hates Houston traffic. That's not a principle."
He laughed.