A fictional national medical board moved to ban a cure for a particular cancer Saturday after an investigation found that researchers had used artificial intelligence during its discovery and appeared insufficiently embarrassed about it.
“The treatment passed every safety and efficacy review,” said board spokesperson Dr. Helen Voss. “Unfortunately, we have since learned how they got the answer.”
Officials confirmed that the inquiry had uncovered no fabricated results or concealed side effects. The concern arose when a researcher described a computational tool as “helpful.”
“That word raised immediate questions about the integrity of the struggle.”
According to the investigation, the team used AI to identify promising molecular candidates, then synthesized, tested, and independently verified them. Auditors found that the researchers had rejected several suggestions that failed in the laboratory, a decision the board said demonstrated a troubling willingness to take only the useful help.
“They cannot claim ownership of the successful ideas while distancing themselves from the stupid ones,” Voss said.
The discovery team submitted its complete experimental record, including independently replicated findings.
The board requested scratch paper.
Lead researcher Dr. Mara Ilyan supplied notebooks documenting years of failed experiments, revisions, and safety testing. An investigator returned them with a note explaining that laboratory difficulty and arithmetic difficulty were separate categories, and only one was currently under review.
“Nobody disputes that they worked hard,” Voss said. “We dispute whether they worked hard on the part a computer could do.”
Patients Accused of Focusing Exclusively on Outcomes
One patient expressed confusion after learning that his successful treatment had become controversial.
“I thought the important part was that the cancer was gone.”
Voss acknowledged that this was an understandable position for someone without a background in academic standards.
“Patients tend to evaluate these things through a very narrow lens.”
The patient, who had returned to work and begun planning his daughter’s wedding, was asked whether he had considered the implications of his recovery for people who remembered doing their own calculations in graduate school.
He said he had not.
The admission was entered into the record as further evidence of the problem.
Another patient offered to personally complete several pages of long division if doing so would allow the hospital to continue treatment.
Officials said the gesture was appreciated but could not retroactively establish that the researchers had earned their discovery.
“You cannot simply outsource the suffering to the patient.”
A patient advocacy group asked the board to distinguish between evaluating a treatment and grading a homework assignment. The board replied that the distinction would be easier to discuss once everyone had put their laptops away.
Hospitals received an interim consent form warning that the treatment’s benefits might include additional years of life obtained through a process that some reviewers would not have permitted in an introductory course.
Every patient signed it.
The board requested a version with larger warnings.
Researchers Ordered to Show Their Work Again, Less Efficiently
The board subsequently requested that researchers reproduce the breakthrough without AI, calculators, spreadsheets, or the colleague who was unusually good at statistics.
“We need to know they understand it.”
The researchers offered to explain the mechanism.
“Without your notes.”
During a closed-door hearing, Ilyan described how the treatment recognized the targeted cells, why earlier candidates had failed, and which experimental results would have disproved the team’s hypothesis. The examiner interrupted to ask her to multiply two three-digit numbers.
“Does that calculation affect the treatment?” Ilyan asked.
“It affects how we feel about you.”
The board also asked the team to submit the number of hours it would have spent without computational assistance. Ilyan said the estimate would require a model. Officials said that sounded like something else she should do by hand.
A junior researcher attempted to comply by copying the computer’s calculations into a notebook. Auditors rejected the submission because the handwriting was too relaxed and several correct answers appeared without preceding evidence of distress.
“We are looking for the moment where you almost gave up,” one reviewer explained.
The researcher pointed to the hearing.
Other Treatments Under Review
The investigation has expanded to additional therapies after auditors discovered that computers had performed calculations behind several celebrated medical advances.
One laboratory was unable to produce evidence that anyone had manually multiplied the relevant matrices.
Instead, it submitted evidence that the treatment worked.
“That’s the same answer they keep giving us,” an investigator said.
Inspectors have placed older discoveries in a separate review category after finding that their inventors used reference tables, technicians, and findings published by people who were already dead.
“We are trying to establish where the individual contribution ends,” Voss said. “So far, it appears to end considerably before the discovery.”
One department defended its use of a microscope by explaining that the relevant cells were too small to see. The board asked whether the department had tried harder.
Officials acknowledged that a total prohibition on assistance could delay new therapies. They said the delay would provide valuable evidence that the prohibition was being taken seriously.
Board Introduces Certification for Personally Endured Arithmetic
A newly formed Committee for Authentic Discovery proposed labeling treatments according to the amount of arithmetic personally endured by their inventors.
The highest certification would require handwritten calculations, visible eraser damage, and testimony from a spouse confirming that the work had damaged the relationship.
“We’re not against technology,” the committee chair explained. “We simply believe a discovery should contain something irreplaceably human, such as spending three weeks correcting a sign error.”
Under the proposed system, a correct treatment discovered quickly would receive a provisional designation. The same treatment discovered after a researcher missed two anniversaries would qualify for the gold seal.
Committee members rejected a proposal to include the treatment’s success rate on the label, warning that the number could distract patients from the conditions under which other numbers had been obtained.
“There is already a whole section of the medical record about what happened to the patient,” the chair said. “We are finally making room for what happened to the mathematician.”
The board’s ethics office said researchers could qualify for reconsideration by expressing regret, returning to an approved level of inefficiency, and confirming that they would have preferred a later discovery if it had made them feel more deserving.
Ilyan said she preferred the discovery happening before her patients died.
The office marked her application “insufficient reflection.”
Artisanal Alternative Still in Development
Hospitals have been instructed to ask patients whether they prefer conventional treatment or an artisanal option discovered by a man who refused help.
The artisanal option is not yet available.
Its creator is still calculating.
The board praised his commitment after he declined an offer from the original team to share its verified findings. He said accepting them would defeat the purpose of the exercise. Asked whether the purpose was to cure cancer, he said that was one of several considerations.
In the meantime, hospitals must keep the successful treatment in locked storage pending a final ruling on whether its discovery involved enough inconvenience. The storage inventory is maintained in a spreadsheet.
Voss said that use of software was administrative.
“It’s different when we need to get something done.”
At press time, the board issued a statement assuring patients that preserving trust in medicine remained its highest priority.
The statement contained a mathematical error.
Officials confirmed it was human and therefore would remain.