Hi,

I recently dropped out of my PhD at a UK university. My background is that I'm a middle aged technologist.working in the city who enlisted with a British university for a PhD program in explainable ai.

So I self funded a part-time PhD as I wanted a change in direction but after much thought dropped out this year and wanted to offer my thoughts to anyone who is thinking of doing a PhD based on my experience.

1. Check that the university is actively engaged in research - conference and journal papers that are recent as well as papers in the pipeline. Without this I would not recommend engaging in the program - the university need to establish their credentials. I unfortunately believed that they were going to build up the capability in explainable ai but never did.
2. Fail fast and fail early - the warning signs with my supervisor were there in six months - no knowledge of the area, no understanding of ai/ml, didn't read papers etc. Basically it was obvious that they were unable to participate in the program and so it would have been better to call it a day then.
3. Be aware of constant excuses from supervisor why things were done, unavailable for long periods of time and never reviews research and unable to offer reasonable opinions.
4. See what the university actually can offer - for me I have superior technology in the cloud with the latest libraries where the university was five years behind on computers that were lower spec than my home pc :( with the advent of arxiv and a bit of chutzpah I am able to maintain a network of academics in this area. So what do I need a university for if I just want to do research?
5. When I did decide to leave I received no support from the wider university including other staff and student organizations. For all the stated support I felt there was near zero available surprisingly when you wanted a conversation about leaving.

Overall, coming from industry, I found my unfortunate experience to be least intellectually stimulating experience in recent times. I am lucky enough to work with really smart people who have provided a much higher and deeper standard around explainable ai that available at the university I attended.

I hope this helps someone at some point.