About the Course/Instructor

CSC 261 - Artificial Intelligence - Weinman

Your answers to questions B-12 and C-12 of Lab 0 will be given here.

1  About the Course

Q
For writing journal entries, is 3-5 sentences an absolute limit of how much we can write or can we write a little bit more to elaborate?
A
It's not an absolute limit, but if your answer gets too long-winded, I'm nearly certain to stop reading all of it. That's fine, because the primary goal is for YOU to gain understanding. Since writing is a form of thinking, I'm happy to have you do more (not to mention more thorough) thinking about the material.
Q
When submitting assignments, should we include a header with the references.txt file?
A
Ah, that's a good one, as the documentation was silent (at best) or inconsistent (at worst) on that. It would be a good idea, so that when we are grading we are confident you uploaded the correct file. (Since you will have many references.txt files!)
Q
Any extra-credits opportunities?
A
The occasional assignment will have a few extra points available, but not regularly.
Q
What kind of questions/scenarios do you think are more appropriate to receive help from on Piazza and which are more appropriate to receive help from you during office hours?
A
Well, if you want to talk about whether your approach to solving a particular assignment is correct, we should probably talk in office hours. On the other end of the spectrum, if they are pretty direct clarification questions about the assignment or reading, then every benefits from an answer posted to Piazza. Plus, everyone gets the opportunity to answer that question on Piazza, which does count positively toward your participation evaluation in the course.
In general, if you have a question that you can clearly write out (per our journals, writing is a form of thinking), then I'm happy to take them via Piazza. However, if you're completely unsure even how to formulate a question in writing, or just plain think that negotatiating a conversation in which comments happen by turns (rather than a single longform question question and then a longform answer), then by all means come to office hours.
The only time you ask me a question in person (particularly right after or right before class) I'll say "ask on Piazza" is when it's the general, clarification sort that benefits both from me having time to reflect upon and give a clear, written answer but also benefits everyone by being public.
Good question!
Q
Are lab assignments (described in 7.3) the same thing as programming assignments (appear in Grading)?
A
Yes. Sorry about the discrepancy in terminology.
Q
What is Miscellany?
A
I'll post links to various things that might come up in class there.
Q
What is the reason for the second question mark in Why are journal posts due the night before???
A
Bad proofreading?
Q
Why is the need to allocate other classwork to a later timeframe in order to accommodate for the 10:30 PM deadline not mentioned?
A
It's a need. Of course, I can't force you to go to bed early, but I want to encourage it and facilitate it with my own homework. My conscience can do no other.
Q
Where should I mention typos?
A
If they're in assignments, example code, etc, go ahead and post it to Piazza. If it's in other less critical course-related documents you can just send me an email. If they're in the book, they're probably already known.
Q
Do Kernighan and Pike address the existence of algorithms which are clever instead of clear for more efficiency in their work?
A
I'm not sure, but probably not. There is a copy of the book in the CS Learning Center and I highly recommend it.
Q
Were there negative comments about the use of journals? What were they? Why are they not mentioned?
A
The most consistent negative comment seems to be growing weary of prioritization, or the amount of time it takes. The vast majority of comments regarding journals are indeed positive.
Q
How will negative responses on lab partner reports be handled? Will the affected student be informed?
A
I will read them and take them into account when assessing student grades. If there is a clear and well-established pattern, I will raise the issue with the student, but I will not share feedback directly, as I want to protect the individuals who may have made the comments.
Q
Why does About the Course/Instructor not contain the information about the course from the syllabus?
A
Because I'm answering questions here, now, that are not answered by the syllabus.
Q
Will future questions about the course be included in the About the Course/Instructor page as well?
A
No. Just ask on Piazza.
Q
Are we still required to do 6P documentations, or just some phrases that make sense are fine?
A
See "Grading", Section 1 (Comments).

2  About the Instructor

Q
What are your current research interests? How did you get interested in artificial intelligence or computer science in general?
Q
How does AI fascinate you? What made you choose CS as your career?
A
When I wrote my first program as a child, I was enthralled with the idea I could make the computer do what I wanted it to do. I suppose it's a little ironic that I'm now writing programs that do what they learn to do. I still love that behavioral aspect, and I'm curious with the general properties of sensible and efficient information processors (whether biological or not). I love teaching the course because I find the material fun and interesting and I appreciate getting to share it with students who also generally find it fun and interesting.
Q
What are you working on this semester and how is it going? I always found computer vision somewhat interesting when I was first exposed to it when Google released an update to the translation app that could translate text in real time by just having a user point their camera at it and I found it fascinating. What type of problems are being faced in the computer vision world right now that are keeping it from progressing?
A
I'm still working on tying together various ends on my map processing project, which many MAP students have contributed to over the years.
I suppose one set of problems I could point to is that while we have some powerful new function regressors (read: deep neural networks), we still need lots of data to train them, and folks haven't quite fully adapted away from labeled training sets to a more biological "learn from self-directed experience model". (People are doing it, and have explored it for a while, but the new wave of computer vision hasn't totally found its way there yet.) Moreover, these models do not yet seem to be very robust; they can be tricked by small changes to the input that are completely imperceptible to humans, and that's a big problem for safety-critical systems.
Q
What do you find most interesting about AI and this course?
A
Our human relationship to AI; we are social creatures and have not really evolved (fast) critical systems for interacting with AI. From the earliest days of Weizenbaum's Eliza program, people (Weizenbaum's secretary in his famous example), have treated programs they even know are simple rule-based machines as something more.
Q
how do you think AI is related to neuroscience, and do you think it is possible that we can insert some chips into human bodies that serve as sensors and stimulate some certain parts of the brain to help people with disabilities?
A
Yes. It's already happening.
Q
Is this course your favorite one to teach? If not, what is your favorite?
A
Well, courses are definitely not like children, there are favorites. But I've grown to enjoy teaching a lot of my regular courses. This one might be it, though. Each one is certainly different in what I enjoy about it.
Q
How did you get into artificial intelligence? What is your career path like?
A
I took an undergraduate AI course about the same time I took a course in the mathematics of image processing. I loved how visual the math was, and eventually got more into computer vision, but my work in that area was always supported by some sort of AI.
I have a pretty straight career path. Although I worked as a software intern in the financial industry after my sophomore year of college, after that it was REU, grad school, Grinnell!
Q
What do you think is the most important feature to success in computer science major?
A
Learn to satisfice. (And develop good debugging skills!)
Q
What is your goal for the class?
A
That you meet our learning objectives.
Q
What have you gained from teaching the class?
A
I've learned more about AI than I never knew before, including teaching subjects I never learned (AIMA is a book book) in either my undergraduate (AIMA first edition) or graduate (AIMA second edition) courses, but also details about the algorithms I never had time/inclination to consider.
Q
How do you feel about past times you have taught the class?
A
It's evolved, and the enrollments have grown. (It was fun teaching to single digits up to the low teens.) Nevertheless, I feel like I've made tweaks and incremental improvements based on notes I make after each day of each offering.
Q
How do you feel you have changed over your time as a professor?
A
I've learned a lot more about creating inclusive classrooms, about how learning actually works (cognitively), and grown to understand more of the varieties of student experiences. All of these have changed how I teach. I've also become a parent, and that definitely changes your outlook on the world (and - for better or worse - gives me many new points of reference in my lectures).
Q
What are interesting/unexpected/weird/funny results/behaviors that you have encountered from artificial intelligences?
Q
What are interesting/unexpected/weird/funny results/behaviors that you have encountered in other fields of study?
Q
Why do kids love the taste of Cinnamon Toast Crunch?
A
Why do I love the taste? Just the right combination of sugar, spice (cinnamon, etc.) and salt. I don't like how fast it gets mushy, though.
Q
Which languages do you wish you could study and why do you not study them?
A
Computer languages? I'd like to know more about all the crazy stuff you can do in C++ now that didn't exist when I first learned the language more than quarter century ago. This summer might be the time! (Answer: limited time and need.)
Q
Is it possible to add more questions at a later date?
A
Probably not. You can ask me in office hours, before class (if I'm not too busy setting up), etc.
Q
Do you ever feel like a sophisticated machine yourself? And how do you find your place between human and machines?
A
Occasionally. Fortunately, machines are not sophisticated enough yet to be confused for full AGIs, so it's more about not being annoyed or seduced by them.
Q
Why did you decide to grow out your beard?
A
I almost always stop shaving during fall semester's finals week. I usually shave before the first day of class, but it's been so cold and I've been so busy I figure having one less thing to do in the morning would be better (and keep me warmer).