Meanwhile, I've been chatting to various people about this. Thanks to all who've commented on previous entries - I'm not adding one line "thank you" comments to each one unless I have something specific to say, but they are all appreciated. I think the main issue here is whether I'm still hoping to do a PhD afterwards, and whether that's still feasible. The related issue is that if I only need 50% to pass, and I can't get a distinction, then is it worth putting in loads of effort to get 90%, or should I have more modest aims? Basically, there are two options when it comes to choosing a project:
- Do a project that's well specified, and closely related to my MSc. Something like "Animation of string algorithms" on this page. This would play to my existing strengths, so it would be fairly easy to get a passing grade. However, it would be very difficult to excel in a project like that, since I can't do anything particularly imaginative.
- Do a project that's very open-ended, and closely related to a potential PhD (i.e. something that interests me enough to spend 4 years on). The semantic web has been suggested, although I don't know whether there will be a project offered on that theme. This would allow a lot of scope, so in theory I could produce something very impressive. Equally, I could wind up falling flat on my face, by failing it again (and I wouldn't be able to retake it again next year).
So, if I'm just concerned about my future (commercial) career, then it makes sense to take the first option. That would then clear up the black hole on my CV, and with a bit of creativity I wouldn't have to mention the retake at all. ("I have an MSc, and I spent 2002-2003 studying for it. I was then working again from September 2003 onwards, with a few months off over the summer.")
If I want to do a PhD, then it makes sense to take the second option. But it is a gamble, and given that I've screwed up the last two projects I did (final year BSc and MSc), that doesn't bode well. And even if I do well, it may be too late, if I've burned my bridges getting this far.
I'm leaning towards the second option. One thing I've learned in life is that I generally regret the things I haven't done more than the things I have. So, if I give it my best shot, and I can't get a PhD place, that's something I can live with; at least I'll know. By the same token, if I screw it up again, that's probably a fairly clear indication that I'm not cut out for a career in research.
I'm not sure how much useful preparation I can do between now and Christmas, although the suggestions to read up on AI in general are useful. I think I'll try to clear as many of the non-project-related items off my "to do" list as possible, to reduce the distractions that I'll have to deal with from January onwards. So, trying to clear the eternal backlog of unread books/magazines would be a good start there.