One year after Mike first raised concerns about the algorithm, Algorithm, Inc. was a transformed company. Congressional hearings had led to new regulations. Competitors had emerged with genuinely fair algorithms. Mike was now the head of the Ethics Division—a real division with real power.
But the transformation hadn't come easily. There had been more articles, more investigations, more resignations. The CEO had stepped down. Dr. Chen had retired. A new leadership team had taken over, genuinely committed to doing things differently.
Mike stood at the window of his office, looking out at the city below. A year ago, he had been a junior analyst, afraid to speak up. Now he was shaping the future of ethical AI—not just at Algorithm, Inc., but across the industry.
The class-action lawsuit had settled that morning. $500 million in damages to affected borrowers. A commitment to rewrite every algorithm from scratch. Independent oversight for the next decade.
It wasn't perfect. It wouldn't undo the damage that had already been done. But it was something.
Lisa knocked on his door. "Ready for the press conference?"
Mike nodded. Today, he would stand in front of the cameras and explain what had changed—and what still needed to change. He would talk about the importance of human oversight, the dangers of blind faith in algorithms, the need for constant vigilance.
But first, he had one more thing to do.
He pulled out his phone and typed a message to his younger self, the one who had been afraid to speak up:
"You did the right thing. It was hard, and it cost you more than you expected. But in the end, it mattered. The world is a little better because you refused to stay silent. And that's all any of us can hope for."
He deleted the message without sending it—there was no one to send it to, after all. But writing it helped him understand something he hadn't fully grasped before.
The algorithm had been the problem, yes. But the real issue had been human decisions—the decision to prioritize profit over fairness, the decision to ignore evidence, the decision to trust the machine instead of questioning it.
In the end, it wasn't about algorithms at all. It was about people. And people, unlike machines, could choose to do better.
Mike walked out of his office, ready to face the cameras. The work wasn't done—it would never be done. But for the first time in a long time, he felt hopeful about the future.
And that, he realized, was worth fighting for.
Three years later, Mike stood in the same conference room where he had first questioned Dr. Chen about the 0.3%. But everything else had changed.
Algorithm, Inc. was now considered a model for ethical AI development. The company had published its fairness metrics, opened its algorithms to external audit, and established a precedent that other companies were forced to follow.
Mike had become a reluctant public figure - interviewed by journalists, invited to speak at conferences, consulted by lawmakers. He never sought the attention, but he used every opportunity to push for more transparency, more accountability, more human oversight.
"Mike?" Lisa appeared in the doorway. "The new batch of loan applications is ready for review."
He nodded and followed her to the ethics division. The team had grown from just him to over fifty people - analysts, lawyers, ethicists, community advocates. They reviewed every flagged case, every exception, every person who did not fit the algorithm's neat categories.
"Anything interesting?" he asked.
"Actually, yes." Lisa pulled up a file. "This one. The algorithm approved it, but something feels off."
Mike studied the application. On paper, everything looked correct. But Lisa was right - there was something the algorithm had missed. A pattern that only a human would notice.
"Good catch," he said. "This is why we are here."
That evening, Mike walked home through the city. He thought about Maria, whose business was now a neighborhood institution. He thought about Dr. Chen, who had left to teach ethics at a university. He thought about all the people who had been denied opportunities by algorithms that did not care about fairness.
The algorithms would keep evolving. New systems would emerge with new biases, new blind spots, new failures. But as long as there were people willing to question them, to push back, to fight for the exceptions - there was hope.
The next morning, Mike arrived at the office to find an unexpected visitor waiting. A woman in her thirties, dressed in a sharp business suit, stood in the lobby. She introduced herself as Jennifer Walsh, a senior executive from a company called DataFlow.
"I have been following your work," she said. "And I think we have a problem you need to see."
She handed him a folder. Inside were documents detailing a new AI system - one that made Algorithm, Inc.'s original model look primitive. This system was being deployed across dozens of industries, making millions of decisions every day.
"The scary part?" Jennifer lowered her voice. "No one knows how it works. Not even the developers. It has started making decisions that no one programmed it to make."
Mike felt a chill. This was exactly what he had been warning about - the black box problem, scaled up to a level no one had anticipated.
"We need your help," Jennifer said. "Will you take a look?"
Mike looked at the folder, then at Jennifer. He had built something good at Algorithm, Inc. He had a team, a mission, a purpose. But this was bigger. This was the next frontier in the fight for fair AI.
"Tell me more," he said.