Expert System Disadvantages

okay let me have it what did you come up with expert systems / experts what do you lose what kind of problems do you end up having yeah emotionless okay give me an example why that's important because I think we said before that the emotionless this led to consistency right it's always the same thing which could be a plus so why could it be a negative can't give me an example okay so when when these decisions affect people okay yeah they can be problems that come in there if something's like a system is not sensitive to those kinds of issues can you give me an actual concrete example okay yeah all right so there's a there's a sort of awareness that humans have of human beings how they feel and how they've been affected by situation which not just affects each individual decision but also how they relate to each other all right that's fine okay because I will give you a totally separate issue problem weakness disadvantage hmm okay so I'm trying to think about that's true so we're talking about you know the hardware limitations on the system I'm trying to I'm trying to steer away from the general ones right because we can always say things like you need you need electricity and you need all these kinds of things which are gonna apply to every piece of technology and we can list them off that's fine but I want to can we be more specific about how to face decisions okay did anyone have one it's a little more related up the back anyone any takers okay alright so the job theory the employment situation I'm not going to put that under here okay for two reasons even though it's actually valid it's a very important issue together okay employment as you saw you may remember really did multimedia right he's always been a two-edged sword with technology because you destroy jobs and simultaneously you create create them right so depending on what kind of technology sometimes you end up gaining more jobs than you lose or vice versa okay however the real reason I'm not including under here is that I'm trying to get at the the way the decision support system functions and does its job like how does it not not the social impacts outside I'm like walking in this situation what is better and what is worse does that make sense okay now just keep my eye on the time Rocco's gonna come soon so I'm just gonna zip through that's okay I've got I've got four reasons okay so number one it can be done here's what I mean by that okay excuse me common sense and intuition right are characteristically human sort of qualities right you're like look this is obviously confiture okay they they clearly convey that this wouldn't make any sense now for a computer that kind of thing is actually very difficult to program common sense right so you say Dom right computers make completely ridiculous decisions just think of the last time GPS told you okay if you want to get from A to B the best way to do it this go some crazy other way okay and then you'll land there half an hour later okay it's just following the rules that it knows to provide you with assistance in making your decision and doesn't see in a common sense of well just because there's no you know part of the map that shows you go from here to here directly it doesn't mean you can't right so common sense is something that it's very difficult to program into a system secondly this is the sort of flip side of the consistency thing right expert it's one of the great things about that is that often they are marked by creativity right they can come up with a new kind of solution for an old problem or being a new problem that they never met before and through all of their accumulated wisdom and experience they'd be like okay I can work and what to do here now an expert system by contrast usually is my fun fact that it'll keep doing the same thing every time very predictable right this is one of the challenges in producing a system that could beat a chess master right because if you play game after game after game it'll keep on doing very similar things if you provide it with the same inputs okay that's great for consistency it's bad if what you need is actually something novel something new something creative okay all right thirdly um data integrity now what's that integrity do you know it has to do with reliability right it includes two things number one accuracy okay the data that you have has to be right and number two remember what was the difference what's the distinguishing between accuracy and integrity that's still that's still part of accuracy yeah yeah you mentioned that reliability I think I heard it yeah you've got to have everything okay so the diversity integrity and accuracy is not just is it right what you've got but you've got to have the whole picture okay so you know the array will you remember it is uh in in court people say I swear to tell the truth the whole truth and nothing but the truth right that's not integrity now why is that a promise for expert systems okay well the the pool of knowledge and expert system is is is getting out right it's going to be fixed okay it needs to be updated manually so a crucial feature about an expert system is that of itself it doesn't learn okay so dial integrity is a big problem for that because the world that we live in is constantly changing right so it's gonna be updated manually by subset of programmers and some other experts in the domain who actually know this stuff and they see how the situation is changing their data integrity is gonna go out the window very quickly okay not only will things change in the situation okay but also there will be new things discovered which the echo system doesn't know looking whereas humans learning is part of what makes us feel it right so an extra system very important this is a distinction between this and the next kind of level artificial neural networks we classify both of these as artificial intelligence but expert systems they don't learn that a change unless you do it yourself whereas an artificial neural network the cool thing about it is it learns on its own we'll get to that later ok ah one last reason one last disengage it has to do again with this arm not learning ok we have a rule in to your expert system ok and it learns to make these decisions but what if the decisions turned out really bad ok you do you make this kind of choice I get it again and again and it's it's obvious that you know you get very very poor outcomes ok if it's gonna be consistent like an expert system is it's gonna keep on doing it again and again again right it won't respond to its stimuli ok when it when it makes a decision and it turns out bad it's not gonna change like a human would okay how many times do you have to press something you get zapped Eli okay I'm gonna stop doing that now right where is the expert systems just gonna keep on going after a particular line of action a particular series of logic even if it turns out badly okay I suppose you could call it unresponsive