Takeshi sat at his desk, staring at the document on his screen. His fingers hovered over the trackpad, the cursor blinking over the words "Demographic bias acknowledged; model adjustments deemed unnecessary for initial deployment."
He had suspected the truth, but seeing it in official records made it real. Someone—the hospital, the research leadership, whoever was in charge of making final decisions—had known about the flaws in the AI and chose to ignore them to push the project forward.
He exhaled sharply and leaned back in his chair. His chest felt tight. What the hell am I supposed to do with this?
Dr. Saito's face was unreadable as she read the document. When she finally spoke, her voice was steady but quieter than usual.
"I knew there were gaps in the data," she admitted. "I didn't know they deliberately ignored them."
Takeshi searched her expression for anything—denial, anger, regret—but all he saw was resignation.
"I don't know what to do," he said. "If we call this out now, the whole project could collapse. But if we let it go, we're letting bad science dictate patient outcomes."
Dr. Saito sighed and set the tablet down on her desk. "If you go public with this before the conference, do you know what will happen?"
Takeshi swallowed. He knew the answer. He would immediately be shut out—from the project, from the hospital, maybe even from future research opportunities. No one would want to work with a postdoc who exposed internal problems, even if he was right.
Dr. Saito met his gaze. "I won't tell you what to do. But I will tell you this—if this project gets shut down, there might not be another opportunity to fix it. Right now, the algorithm is still in our hands. If you want to make a difference, you have to be smart about it."
For the first time, she wasn't telling him to back off. She was asking him what he wanted to do.
Takeshi exhaled and nodded. "Then we need another way."
That night, Takeshi and Hiro sat in a corner of the hospital's research wing, going over their options.
"If we just drop this bomb at the conference, we're done," Hiro said. "They'll shut it down and spin the story however they want."
"I know," Takeshi muttered. "We need to make sure they have no choice but to acknowledge the issue."
They stared at the presentation slides, the carefully curated summary of the algorithm's success. The problem wasn't just what was in the slides—it was what was missing.
Takeshi tapped his fingers against the desk. "What if we change the presentation?"
Hiro's head snapped up. "You mean sabotage it?"
"Not sabotage," Takeshi said quickly. "We just... adjust it. We include the demographic bias in a way that forces them to acknowledge it."
Hiro frowned. "They'll never approve that."
"They won't notice," Takeshi said. "Not until it's too late."
The plan was risky. If they could subtly include one key slide with real numbers—just enough to raise a question publicly—the hospital would have to address it on the spot.
Hiro sighed. "You really think this will work?"
Takeshi rubbed his temples. "It's the best shot we have."
As he was finishing his work that night, Takeshi's inbox pinged with a new email.
Unknown Sender
Subject: I know what you found.
Takeshi's pulse quickened. He opened the email.
"If you don't act now, someone else will—and you won't like how they do it."
His hands tightened into fists.
He wasn't the only one digging into this.
Someone else was already making a move.
YOU ARE READING
The Algorithm
Short StoryIn a cutting-edge hospital known for its advancements in medical research, Takeshi, a postdoc in data science, is given the responsibility of validating a groundbreaking AI algorithm designed to detect early-stage pancreatic cancer. The algorithm is...
