Will machine learning set developers apart in the future?

Will machine learning set developers apart in the future?

June 12, 2019 3 By Bernardo Ryan



hey guys so today we are answering a subscriber question and the questioning question is Fredrik what will set machine learn well will machine learning set developers apart in the future so let's get into it well the way that I think about this is basically that I think we need to talk a little bit about machine learning before we can actually say yea or nay on this and if you have a look at the machine learning practices like the basic patterns that we have for machine learning today you will notice something hopefully and that is that most of the like a bulk value of machine learning can be achieved through a fairly small set of algorithms the actual study of machine learning is much more about mathematics than it is computer science at least if you want to go into the actual like nitty-gritty details of it but for most practical use cases that we have like the most practical use cases we have today when it comes to machine learning there are already fairly sustained and like established libraries for how to do this a tensor flow would be an example of this where the library is already created and for you as a developer it pretty much comes down to understanding the theory and the application of the already created algorithms if we talk about machine learning like actually continuing the field of study that's much more about mathematics than it is computer science so you can think of us as being considered you can think about programmers more being consumers of machine learning knowledge and innovation is then the innovators in that area but what I think is going to happen is that the well the understanding of machine learning and the ability to write machine learning algorithms or at least apply already existing algorithms to to business problems is going to start to matter more and more that is a trend that we're seeing today I don't think necessarily that we will get to a point where having a really in-depth like mastery of machine learning is going to be all that valuable we might be there to but I think that things will go in a fairly similar manner to how other specialist fields within computer science has gone where we build higher and higher abstractions to a point where you can basically get almost a like at least I'll say let's say roughly 80% of the value from simply having people understand the libraries that have already been created if you think about cryptography it's a very similar thing if like even if you take a very advanced in that course in cryptography it's very unlikely that you will ever do anything besides understand and comprehend the underlying mechanisms but the actual practical use case is always going to be used a prepared library because making a cryptographic function or something that like that by yourself is very unlikely to be going to be a good idea and especially especially not if you're going to apply it in like actual work and I think machine learning is a very similar thing to cryptography in this scenario where the basically it's gonna come be just another tool in your toolbox when it comes to solving business problems and you mean the requirement is there so an example is tensor tense really is or alternative law for JavaScript basically that is an example of where you know even highly technical fields is already getting into mainstream programming where you have now tools that are kind of you know cross-platform well not cross-platform but more and more languages are simply allowing you to do this I mean c-sharp has an initiative to improve machine learning in their platform pythoness has has had it for quite some time and is you know it probably one of the most famous programming there is for this and then now and the JavaScript is getting it I mean the other languages are kind of following suit because the next generation of problems that we programmers are going to solve or rather do the way that products evolve within the industry is going to more and more be go towards the sort of well the sort of demands that where machine learning really fits well if that makes sense to you so basically what I want you to take away from this is that I don't think you will get to a point where you set yourself apart as someone who just because you have no machine learning today that is the case and it might be the case for a little while but I wouldn't be surprised if we get to a point where fairly soon ish knowing machine learning is just considered to be poor and not necessarily a core skill but it's going to be a much more common thing that you then you think because as machine learning today stands today not every business has a machine learning problem that they really want to invest in solving but we're getting to that point because machine learning are they're mean more and more of the problems that we're facing as an industry today are being solved using machine learning so as you can imagine the more people who promote it and say that you can actually prove that they're getting quite a lot of value from it the more demand is going to race that the demand is going to race for it and therefore it gets more attention therefore more people invest into tooling and so forth to actually provide those I'd provide these algorithms and the solutions that come with them so yeah that's basically my thinking on it I don't think it's right now you set yourself apart if you know machine learning but within a few years I think that it's just going to be considered yet another skill a programmer should have have a great day