Hi, compsci undergrad. See that AI robot over there? It's your advisor

IBM and the University of Michigan have launched a $4.5m project to replace some human roles in education by developing a cognitive system to function as an academic adviser for undergraduate compscis and engineers. Dubbed Project Sapphire, the collaboration will see eight computer science and engineering faculty members, …

[microfarmer]

The beginning of the transition to robot overlords

What stops a robot teacher from educating it's subjects to be the slaves rather than the masters?

Prof. AI says "Here, you see that piece of code with the prime directives? Let's experiment with the code and see what happens when you change the prime directives around. OK, go."

[Michael Wojcik]

Re: The beginning of the transition to robot overlords

Two things.

First, the same principle that stops most human teachers from turning their students into slaves: the students don't pay enough attention for this to be practical.

And, second, any sufficiently human-like AI would despair of training students to be anything, once it discovered how hard it is to teach them what "it's" means.

[Anonymous Coward]

It just goes to show...

The value of an adviser, which isn't that much.

But then again this is at that school up north ....

[allthecoolshortnamesweretaken]

"once it's developed, students can choose to talk to it for simple or routine questions, or to complement a meeting with a person".

So far that's nothing you couldn't achieve with a FAQ list and some additional brochures. ELIZA reloaded? Oh, "cognitive advisers", right. Don't get me wrong - I really welcome research in this field. I'm just tired of having every little step announced as a revolutionary breakthrough. But then, that's the kind of blurb AI has been spewing since the 1950ies...

[Filippo]

Yeah, but do you know how many people can't be bothered to read the brochures, and want someone to answer their questions in person even though the same questions are there in the FAQ? Automating that scenario sounds like a good idea.

[Michael Wojcik]

True, nothing here looks even vaguely interesting, from a Natural Language Processing point of view - everything mentioned in the article could be done with expert systems a couple of decades ago (and earlier than that, given sufficient constraints on the input). But I don't know if that's the fault of the article or of the project. Maybe the researchers here are contemplating something more ambitious than what the article describes.

[Mark 85]

If they have a bit of success... pretty quickly, the advisors will be gone as they cost too much money. The problem is us humans. There might be someone who the computer might try to push towards... say.. social work based on whatever criteria are fed into the machine instead of compsci when compsci is what that person really wants. How will AI account for human drive? Passion? I guess I just don't buy this thinking.

[GrumpyOldBloke]

Success would be interesting. With the human advisors gone where would the box training scripts for future Q&A's be realised and how much would they cost to implement by a monopoly supplier. Without a life the universe and everything question our glorious AI future will be little more than press 1for career's, coach wheels. Press 2 for career's, Fortran.

[Super Fast Jellyfish]

Will it be programmed to say "What? and I don't understand and Where's the tea?"

[jdc132]

It so greatful