Back to The world according to the machine
An AI assessment of a thesis
The following represents my personal experience with human-based bias and ideology, compared to what an emotion-less AI agent comes up with. (They are screenshots of a conversation with Google's AI Mode.)
To summarise, the assessment of an Honours thesis as performed by lecturers at Griffith University has resulted in almost a Fail, although the thesis' details have been fully elaborated upon throughout its 155 pages. Such was the tone of the evaluation, I had to check whether my name had not been confused with someone else's. See the Brief.
Note that the AI appraisal is based on the full reading of the text, which includes the associated program's code and explanations. Note also the reference to "modern language models"; the thesis was written in 1999.
I leave it for the reader to decide whether this sounds like a failure.
The thesis can be accessed from Downloads. For more on the Griffith affair see the social experiment.
17 July 2026
© Martin Wurzinger - see Terms of Use