Which is the most promising sector of Artificial Intelligence in future(2025) ?

I am currently studying about 'Machine learning'.

Please stop refering to ML as AI. It is not AI, not by a long shot.
And yes, ML is an interesting subject, as long as you are willing to give up on undrstanding the reasons for the results it gives, and accept GIGO as the underlaying priciple of the whole thing.

@magicMirror Isn't ml is an approach to ai ?

@Xenon an approach doesn't make it AI. If machine learning is artificial intelligence, algorithms are machine learning and that is just not so

I like where this is going... *grabs popcorn*

@FintinaM *sits next to you and opens beer*

@Wack cheers.. *continues watching*

ML Msc. here. ML is a subsection of AI. ML is, in the easiest terms, an approach to automating some problems AI tries to solve by letting machine basically tune itself. This tuning requires decision making, the resulting tuned system also makes decisions.

@ctmalloy Isn't what you're saying basically that if milk is a beverage, water is milk? How is machine learning not a form of AI?

@FintinaM @Wack *cheers 🍻 *

@kkalem Does ML still have a promising future ?

I believe it does. Not in the sense that it takes over everything but rather it becomes a useful tool, something to take some of the burden off our shoulders and offload that to machines. Just need the hype to die off. My personal opinion is that ML is to AI what higher order functions are to programming.

@Xenon ML is not AI. not by a longshot. In order to explain why, you need to understand the whole AI topic. in short, AI means a general problem solver. Able to define new problems, design solvers, and be able analyse itself in order to improve how it works and create new AI as required.
ML is a way to solve a well defined problems, based on analysis of past observation and their solution. as new observations are consumed, new predications are generated.
AI can use ML as one of the ways to solve problems, but not the other way around.

@theScientist ML will give good results as long as the classifier gives good results. to make a good classifier you need "data". take that and reverse it. it is GIGO.

and all the popcorn eaters: this shit affects you too. True AI will be able to do some incredible things, or destroy humanity. Just see some scifi movies.

bottom line: stop calling ML as AI. This is also true for the academia who decided to call it AI bc it is more sexy.

@kkalem acedemics also decided to call Neural networks with a Hidden layer (or layers), "deep learning". Causing some very idiotic previous Managers to really annouy me.
Just try to explain why deep learning is nothing new to some idiot, buzz word spewing, single nueron monkey manager and/or why it should not be applied to everything.

@magicMirror I'm absolutelly aware, that it will affect us pretty soon. There with popcorn and beer for the fight :P
On a more seriouse note, you may find this interesting: https://ethz.ch/en/news-and-events/...

@Wack they are not the first one to try this. IMO it is nearly impossible, as the definition needs to be understandable by the algorithem. You can consider what that guy is doing as a complex macro abstraction lang.

also take a look at this for the intersting side of ML:
https://theguardian.com/technology/...

Also - this is the main problem AI faces.
sarcasm.