10K views in three days. That is the power of a bu...

# tornadofx

[thomasnield]

10K views in three days. That is the power of a buzzword I guess. Just use the words "neural network" and people travel from near and far.

[ValV]

Indeed, words neural network and artificial intelligence works like a spell, perceptually it's just like some kind of magic and superpower, but if you say optimization and statistics it will take exactly opposite effect, I think 😏

[thomasnield]

@ValV Unfortunately, I really do think that's the case. I kind of regret titling my KotlinConf talk Mathematical Modeling with Kotlin when it might have better marketing with Optimization and Machine Learning with Kotlin.

[thomasnield]

I talked about the pros and cons of buzzwords in an article I wrote in January. I find myself often in a dilemma where I want to be technically accurate, but I also want a turnout. http://tomstechnicalblog.blogspot.com/2018/01/is-it-time-to-stop-using-term-data.html

[ValV]

@thomasnield that was fair enough title for KC, I payed attention to it, because it was titled Mathematical Modeling, 'cause had to deal with modeling once and it turned out to be an interesting thing

[ValV]

Here, in our country, we do not suffer much from data science hype, but words artificial intelligence, neural networks, and machine learning yet still have strong effect. Where artificial intelligence means exact artificial mind, and chess game hardly comes to mind here, neural networks it's kinda brain, and machine learning is teaching a machine

[ValV]

Anyway that hype around those terms made me look at old and boring math from a new point of view. I was not good at math, 'cause we've been taught it without reasonable explanation were and how to apply that knowledge in practice -- just tons of differential equations and integrals, obscure analysis, geometry, and linear algebra. But the word machine learning attracted attention exactly to application of linear algebra and other mathematical facilities

[ValV]

Now, I hope, I see things from the other point of view: math is strong, everything that concerns mathematical problems can give incredible results (in case of successful application), but ai, nn, and ml -- are just for hype (and, highly likely, misconception). There's no teaching of machines, but hardcore math and computation

[thomasnield]

That's pretty much my perspective now. I am taking the approach of learning and teaching with math modeling and implementation, first rather than using black box libraries. It really makes the subject more productive and fun