> ## Content Index
> Fetch the complete content index at: https://www.oliphonic.com/llms.txt
> Use this file to discover other available public pages before exploring further.

# Experiment #2: Can AI write me a PhD thesis?
- URL: https://www.oliphonic.com/experiment-2-can-ai-write-me-a-phd-thesis/
- Published: 2026-10-20T18:28:51.000Z
- Updated: 2026-10-20T18:28:51.000Z
- Author: Oliphonic

I've decided on what my next AI experiment will be - **"Can AI write me a credible & defensible PhD thesis on a topic of my choice?"**

I asked the Q on whether AI is replacing academia on an [earlier blog post](https://www.oliphonic.com/wait-is-ai-replacing-academia-too/). So let's put that (very humorously) to the test.

**I think a few rules:**

- The thesis needs to be **wholly created by AI** \- research, writing etc. I can prompt but that's about it.
- The output needs to be research of **excellent, world-leading quality. 100,000 words-ish.**
- I need to be **incredibly illiterate** in the chosen subject field
- The subject field can't require any 'physical' experimentation - I mean, last time I checked I can't ask chatGPT to do some experiments in a lab
- The topic needs to be absolutely hilarious

💡

For all my sins, **I actually did a PhD back in 2011** \- in Mechanical Engineering at the University of Bristol. It was good. It was hard. Thousands of hours of research, literature reviews, experiments, conferences (they were the really fun part), writing, publishing etc. But I got the piece of paper at the end of it, and I'm officially a 'Doctor'.

![](https://storage.ghost.io/c/9c/fa/9cfabe9c-bcc0-471f-a6fa-271bee166fa4/content/images/2026/10/The-Mysterious-Ecology-of-Traffic-Cones.png)

I'm gonna get the answers to the questions everyone has been asking.

## Now for the fun part

**I get to choose what topic to do this PhD in.** Why does this make me unreasonably excited. I haven't decided what topic yet - but I'm thinking something that is both completely new research ground and also completely absurd. We have the opportunity to change the world here. Some ideas:

- **Global Seagull Criminality** \- can I explore & identify the world's most criminal seagull population? Analyse geotagged reports, news archives, tourism data and coastal demographics to construct a statistical model of seagull-related theft.
- **Passive Aggressive Trends & Shifts** \- can I explore & ascertain whether the world has become more passive aggressive? longitudinal studies of emails, exchanges, reports, books etc.
- **The Ecology of Traffic Cones** \- can I model how long abandoned traffic cones remain in place? Analyse historical Street View imagery, council records and geotagged photos - outputs could include a traffic cone half-life index.
- **Pigeon Superiority Analysis** \- are pigeons becoming increasingly fearless around humans? Analyse historical and contemporary street imagery, wildlife observations and published behavioural studies.

They're absurd topics but if they're genuinely advancing the latest academic thinking (I mean...if any exists there...) then that's worthy of consideration for that acclaimed Doctor of Philosophy...........right??

**Oh this is going to be really fun.** Please send me any more completely absurd suggestions or ideas for what I can do this PhD on.

Once completed, I guess I'll be a double Doctor! Dr Dr!

##