Join devRant
Do all the things like
++ or -- rants, post your own rants, comment on others' rants and build your customized dev avatar
Sign Up
Pipeless API
From the creators of devRant, Pipeless lets you power real-time personalized recommendations and activity feeds using a simple API
Learn More
-
-
@angryMoFo Essentially, yes. It's usually disguised as an automated if-else generating system that obfuscates the resulting system so it can no longer be understood (aka a neural net), but all they do is react to input (even self-generated input) and sort it through a rule set. Hence the if else description.
-
@irene Which are used to process the input, not decide on it. The decisions are still essentially an if-else system. The matrices are just a way to make the decisions easier to find by randomly creating abstraction rules which are then discarded if they don't work. The output of those rules is then compared with desired outputs, which, by nature of comparison, can be reduced to if-elses.
It's not how you get to the comparison, it's that comparison happens at all. The nature of AI makes comparison with desired types of output necessary or people will reject the outputs as non-useful random stuff. -
@irene But it is not arguably unfair to describe a neural network as fundamentally an if-else generator. We want them to process information (the matrix part). We want the output to follow some rules (necessitating decisions). We don't care how that works as long as it does, so the system randomly generates some processing and decision rules and then decides if those rules work.
I think anyone who bothers to attempt to understand modern AI at even a superficial level understands that the "if-else" characterization is an oversimplification, but it seems to me to be a (mostly) reasonable one, and I was attempting to explain to the person who asked why it could be construed that way. I'm not necessarily intending to imply that that's how people should ultimately understand such complicated things, but it's not a terribly inaccurate summary to help the uninitiated understand.
When you wanted to know deep learning immediately
That's not AI.
So I had this conversation yesterday while fixing yet another Windows laptop for someone else.
Other Guy = OG
Me = Me (Duh)
OG: So what are your plans after your apprenticeship?
Me: Uh, I'll probably start somewhere that's e-commerce related, kinda like my current company but somewhere else.
OG: Uh have you thought about being your own boss?
Me: Well yeah, but I wouldn't know how to attract customers and shit
-- This is the moment shit gets real
OG: OH BTW I heard that Germany is lacking AI developers, you should do that! It earns you shitloads of cash!
Me: Uhm.. well, that might be true b-
OG: There's no but dude, it's free money, you're smart.. I mean you can fix any computer, right? AI will be just as easy
Me: It's not like-
OG: Duh, don't make yourself look so bad I know you can do it!
Me: B..But I'm not interested in it at all
*silence for 5 seconds*
OG: Well.. I guess you do you then
After that we continued to have random chit-chat about his job and experience (He's a mechanic)
God I hate when people throw buzzwords around and try to convince other people to do what *they* want.
No, I don't want to develop a structure of 1000 ifs/elses, I'd rather keep doing what I'm doing, thanks!
rant
ai
conversation