Introduction to Quantum Information Processing
QIC 710, CS 768, CO 681, PHYS 767, AMATH 871, PMATH 871 (Fall 2023)
[main]
Comments regarding project presentations
Here are some additional comments beyond the basic explanation of the projects [here]. I apologize that some of these comments may be obvious to you, and may sound patronizing. But from my experience in previous years, some students will benefit from this.
Timing
You have 25 minutes allocated, after which there are 5 minutes for questions and for the next speaker to set up. These slots are tight. What should you do if you see your time is running out before you get to your last slide? What you should not do is start talking fast to try to race through many slides quickly. Instead you should skip over what you can, perhaps with very brief comments, and then go to the last slide. It's OK if you didn't present everything that you prepared, as long as you get your key points across. During your preparations, think about whether your presentation can be done in 25 minutes, and also think about a backup plan of what to skip over if your time runs out sooner than you expected it to (as well as additional material in case you end up finishing much sooner than 25 minutes).
Think about your target audience
The target audience is the other students in the class (who have varied backgrounds and cannot all be assumed to know much physics, specialized mathematics, etc). Your talk should be comprehensible to the entire class. For example, if you are in physics, you should remember that many students in the class are not in physics and certain terminology and equations will be unfamiliar to them. You should explain in a language comprehensible to the non-physicists in the class. And it's exactly the same situation if you are in other fields, such as mathematics or computer science: you should not expect people not in your area to have the specialized knowledge that you have. (If you are thinking that most people in the class will not understand your fancy equations or fancy terminology, but gosh they're going to be impressed by it all then you're probably thinking about your presentation in the wrong way.) In your career, you will sometimes give seminars to specialists and sometimes be speaking to broader audiences.
Scope of your talk
You need to make careful choices about what subset of material from your paper(s) to present, and present that material in a manner that the audience can follow. Some papers are short and self-contained, but many are not. You can select a subset of material to focus on in technical detail. For example, if your topic is fault-tolerant quantum computing, you can briefly give a broad overview of the field, state a few technical results, and explain only some of them. Or, for a quantum algorithm, you can give a brief overview of how it works and say these are the key steps and explain only some of them. But here should be some technical content explained in your talk.
Slides vs whiteboard
You are allowed to use slides or the white board. Although it's possible to give a good presentation using the white board (and some people have done so in previous years), in most cases, preparing slides is the safest approach—especially because of the tight timing. For slides, you should bring a laptop or other device that can connect to the data projectors in the room. I prefer that you use your own device. Having said that, in special cases, I will endeavour to make a Mac laptop available, if you send me the slides the day before or bring a USB drive with your slides; however, I cannot be responsible for system incompatibilities (such as Mac vs Windows), so probably a PDF file, at least as a backup, is good if you end up using my laptop.
Please number your slides
This makes it easier for me to provide feedback comments to you. I can say "in slide 6, ..."
What works in talk slides is different from what works in papers
People often mistakenly think that what works well in a written paper also works well in a talk slide. But when someone is reading your paper, they control the time. They can spend a lot of time studying parts that are challenging. And they can refer to previous pages and future pages as they see fit. However, in a live talk, the audience doesn't get to control the time. They see a slide for maybe a couple of minutes and then suddenly you advance to the next slide. Moreover, they are listening to you speak while reading the slide. This all means that the audience cannot be expected to be doing a lot of reading of slides. So you shouldn't overload slides with a lot of information. Try not to make your slides too cluttered.
It helps to try to imagine how the audience (especially the people in the audience for who the material is new) will experience each of your slides. Will they be able to digest this list of equations? Often a theorem can be stated in a manner that conveys its meaning without all the details of the statement being explicitly stated.
Consider the following road sign (that exists in the Waterloo area):
It is meant to be read at a glance, while a driver is thinking about other matters, such as the positions of other vehicles—and how to avoid colliding! Now imagine if the road sign instead looked like this:
The second sign contains the same information, but it's hard to imagine how a driver could absorb the information at a glance. You should try to avoid making your slides look like the second sign.
Here are a few more points (some of them are admittedly picky):
Do not put in references (like "[5]") that are only resolved in a bibliography at your last slide. Of course, in a paper, the reader has the choice of whether to flip to the end to check out a reference or just keep reading. What's the audience supposed to do during your talk? Memorize your third slide and then, at the very end of your talk, realize that reference [5] was this specific paper? If you put a reference in, it makes more sense to put in the slide where it's referred to (and the reference can be in abbreviated form).
If you cut and paste a diagram from somewhere, you should cite it (which could be a footnote on the slide).
Think about how to write mathematical expressions so that they are easily readable. You probably already know that, for complicated expressions in an exponential, it's better to write as exp(expression) than as a superscript of e. More generally, if a subscript is complicated, it's harder to digest, especially if you have subscripts of subscripts. And lower case letters with upper case subscripts often look like products. In a talk, the audience doesn't have a lot of time to parse these expressions.