Hacker News new | past | comments | ask | show | jobs | submit login
Official enrollment for Stanford's online AI class has begun (ai-class.com)
229 points by epenn on Aug 30, 2011 | hide | past | web | favorite | 63 comments

If you've signed up for these classes, how about letting others know here? Perhaps we might be able to form ad-hoc groups to help each other when we're stuck?

If I've missed this suggestion in another thread, please let me know.

In the meantime, I've made a google spreadsheet so please add your details if you want to find other HN readers taking part. http://bit.ly/pLCRzg

Edit: Also found this on reddit - http://www.reddit.com/r/aiclass

There is also this doc from a previous talk https://docs.google.com/spreadsheet/ccc?key=0Aq06GgDrF8g3dFR... we should combine, not fragment. The sheet from your post is to be preferred since it's more structured.

Should it be more feasible to make a google group ?

I completely agree that we should not fragment. Happy to go with whatever folks feel is best.

NB: I cross posted links at the top of both sheets so they're at least 'connected' in some way. Hope that's cool.

Great idea, I'll be joining you. We can keep a chat room open to help each other and just talk AI.

An irc channel on freenode would be ideal. How about irc://irc.freenode.net/#ai-class

Also found this via the reddit, a study group for Cambridge (UK): https://groups.google.com/forum/#!forum/stanford-online-clas...

Good idea, I put my name on the more detailed spreadsheet. I wouldn't mind being emailed at the address I entered there, if there's any need.

Not sure yet if I'll have time for the DB and ML classes, but I'll try.

I don't know whether to take this course or the machine learning one. The both seem very interesting, but I only have time for one. I don't care much about robots, and was partly sold by Ng's separating music from background. OTOH I want to learn Bayes networks and natural language processing. I'd appreciate any advice.

My impressions: CS 221 (AI) is taken mostly by sophomores and juniors who are interested in AI, but come from all areas within CS. CS 229 (ML) is taken mostly by juniors and seniors who are on the AI track. CS 229 has a reputation for being the harder class. I get the idea that people come out of 221 with a knowledge of what techniques are out there so that they can accomplish simple tasks and know where to look when they need to learn more, and that people come out of 229 with a relatively firm foundation to solve ML problems (at least, as firm a foundation as you can get in 3 months in a large and difficult field).

for what it's worth, if you take the ML class, you will learn most of the things you would learn in the AI class and more - although it does get a bit rigorous, and will take more time than the AI class would. from personal experience, I feel like Ng's class gave me a more thorough foundation in the math behind the concepts, and was more challenging to boot - so I'd recommend it if you're feeling up for it.

I know that I hardly have time for one of them, but greedily want to try all three. I wonder if there is an expectation for the course to be available again, or at least provide the full materials afterwards. And if there is a penalty for failing/dropping out of them due to time constraints.

You can find some of the materials for last year's ML class at http://www.stanford.edu/class/cs229/. CS 221 had a similar one, although it got taken down recently. I imagine you'd at the very least be able to access the materials while the class is ongoing or shortly thereafter for archival purposes.

the course materials for the ML class are available at http://see.stanford.edu -- it's CS 229

Interesting, did anyone who subscribed receive any news via email about the registration being opened ? I didn't.

I got no mail, and was signed up for ai, ml, db. It seems only ai is open to registration so far, perhaps to get an idea of demand and discover bugs...

Not I. And I only spotted it because I brought the course up in a meeting at the office this morning.

No, although I subscribed the first time. I read it here ;-)

I love how stanford is doing this for the public -- especially empowering those who really can't afford a degree but really want to enrich their education desires. +1 to Stanford for pushing this and the instructors and TAs who will be dedicating their time to make this happen.

On a side note, I'm deciding to take this class or the ML one. In my line of work, I do believe that the ML class will be more beneficial but the AI one seems way more interesting.

Does anyone have a link to the other classes? I see from the spreadsheet there are DB (database?) and ML (machine language?) classes as well ?

DB Class: http://db-class.org/ Machine Learning Class: http://www.ml-class.org/

I thought folks that had already put an email the first announcement would be emailed by now to let us now sign up is open, but I got none.

I've signed up for the AI class. It should be interesting to revisit the topic after taking it several years ago with Bob Levinson @ UCSC. There, a good deal of the focus was on Lisp and playing Chess.

I'm likely to sign up for the ML one as well.

I like that they separated it into basic/advanced, originally I was thinking of signing up and skipping the homework when I didn't have free time for it, but now I can just do the basic and not feel bad :)

Are we gonna do this? I took AI in college but it was so-so, and I was thinking about doing this again. Anyone in a similar situation?

Yes I'm in the same boat as you. My AI wasn't taught by Norvig in college( though he was a pretty good professor) so it might be interesting to see a different opinion on what "intro to AI" should consist of. I'm also interested in what the delivery system is going to be like. This might be the way courses are offered in the future.

having taken an AI class (using AIMA) at my undergrad college, I can't say that Norvig co-teaching the class made it all that different - probably the most interesting difference was being able to hear him (and Thrun) talk about real-world applications of the concepts we were covering, be it in Google products or Stanford research.

Signed up!

Is there a way to buy a pdf version of the text? From the links at http://aima.cs.berkeley.edu/ I got:

CourseSmart: eTextbook rental. Amazon: Kindle version. Nook: eTextbook. not sure if I can download pdf cause you have to sign up to find out. Cafe Scribe: cafe scribe format. Kno: format only works on iPad.

Edited for formatting.

And does anybody know why the kindle version is not available outside the USA? I'd prefer to buy the kindle version, but since I can't I'm thinking about buying at Kno. Any experiences with it?

Edit: WTF? I just discovered that the Kno ipad App is not available outside USA as well (at least not in my location).

I can't find any list of prerequisites for this class, though their FAQ https://www.ai-class.com/registration/faq implies that there are some. Does anyone know if this class is noob friendly? (as in someone with no CS or programming background)

Doubt it's going to be too hard but maybe not noob friendly. The "basic" level of the class at least should be accessible for the most part. It will cover some probability theory in the beginning so if you have a grasp of that already but forgot it this will help. It looks like machine learning is going to be a larger focus than in the AI class I took so you will need to know linear algebra as well.

Here: http://robots.stanford.edu/cs221/ is the outline with the prereqs and the schedule of topics.

Here are some suggestions of preparatory material:


There are a lot of Stanford CS classes available online. Here's a list of the best ones according to Stanford CS majors - http://raunk.com/list/669,682,1364,2394,2395?filter=4,5

Does anyone know approximately how much dedication would this course require (in hours per week)?

It depends on your background (CS and math are very very useful) and which track you're going for (standard vs advanced). I believe Norvig & Russell's textbook is being used for the class, which is the standard for AI classes at universities and a FANTASTIC textbook. If you skim through it, you can get an idea. I took a similar class in grad school with the same text (I'm a CS major) and sometimes spent 2 hours in a week...sometimes 15. It depends on what you're good at and how much detail they go into.

This book looks interesting but there is no review on Amazon.


I was wondering if someone can comment on the suitability of this book?

Does anyone know why the class is only something like 8 weeks long? Stanford is on a semester system, and even in the quarter system classes are 10 weeks.

Stanford is on the quarter system. You can check out their academic calendar online.



I'm all for the increasing availability of free online education, but it's interesting to note I'm paying 50 grand (partially) for this...

You're really paying for a degree from Stanford. You can get a similar course for much cheaper at a state school. Or now, online for free!

I think there's no doubt that you'll have much more resources at your disposal (like proper office hours) if you are actually enrolled at Stanford.

if nothing else, part of your 50 grand is also paying for the fact that our career fairs are always super crowded :)

You better make sure to get your money's worth by picking Norvig and Thrun's brains at office hours :P. And if you get an A in the course, you can probably do undergraduate research with them through CURIS, something that online attendees won't be able to do.


If you're taking a CS degree in Stanford, earning the 50 grand back is trivial after you graduated.

The "terms and conditions" checkbox was not enabling the "Register" button for me under IE7.

It works under Chrome though.

Try using IE9 instead

I signed up. I'm guessing that we don't have to write the exams in person. Does anyone know?

"The exams will be offered within a 24 hour period around Nov 19/20 (midterm exam) and Dec 17/18 (final exam). ... The actual exams will only take 4 hours and there is flexibility when you start the exam within these dates, but once you begin you must complete the exam in 4 hours"

I don't think anyone's expecting to take these in person.

This is a great idea. I've signed up to the DB class. Looking forward to getting started!

There is one Q-A community for this class at aiqus.com

I'm in...

just signed up!

im in :)

Why does this require students to submit a birthday? This seems like an unnecessary disclosure of personal information.

I imagine they want demographic data on who's taking the course. If you were running a large-scale education experiment, wouldn't you want to be able to measure how things are going and account for variables such as gender, age, and education level? I bet they are going to look at location as well, via IP addresses.

...and as we just learned, this is enough info to uniquely identify you, even without your name :)


> I bet they are going to look at location as well, via IP addresses.

Why would they do it that way rather than self reporting?

I would say gender should be irrelevant too ...

gender could give them all sorts of interesting statistics about what happens to gender ratios in CS when you take the course outside of a traditional classroom.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact