The message said "I'm a doctor!!!" It was from another successful candidate whose dissertation was accepted and who passed her oral defence. I love getting those messages, but this one was particularly rewarding.
Although nobody becomes a doctor immediately after a defence (there's paperwork and a formal graduation approval and ceremony to go through first) this candidates was delighted to be at the end of the journey and I'm happy to celebrate with her.
We had worked closely together for quite a while to address the fact that TurnItIn had rated her thesis at over 80%, which is truly excessive. The problem was that her area of study is a branch of engineering, which is a very complex, highly technical field with many specific terms, calculations and formulae to report. TurnItIn went crazy and gave a false negative.
Our only choice, given the short deadline, was to go through the thesis sentence by sentence, to try and identify what words had triggered TurnItIn. Some of the sections flagged could not possibly have been written by AI, because they related directly to her study.
But when we re-submitted sections, the TurnItIn report changed so that flagged sections were cleared and new, previously un-flagged sections were newly highlighted. Worse, some of the sections that my client manually rewrote and I proofed were still marked as AI generated.
It's upsetting to think that a robotic program would pass judgement on scholarly work. It will take persistent to prove TurnItIn wrong until the algorithm is fixed. But persistence does pay off, and this client proved it. After working diligently on her manuscript, she managed to beat it at its own game, and her thesis was accepted by her university.
Nothing beats human determination and ingenuity.