August 20, 2018 at 7:34 am
In the data science space, there is similar happening and being adopted with more or less point-and-click machine learning. I would imagine it's similar to this article where you just assemble what you want it to do by dragging the objects together and the end result is your fully fledge predictive analytical model that will interpret data and forecast said data for you without R script and so forth.
Equally, I attended a few talks from major vendors we use who also spouted that data scientist are at risk of being replaced by these tools on top of just automation. There is only so many ways to redesign the wheel so to speak. So, why waste time constantly coding it to do the same thing you always do in ML?
As someone who is not a data scientist and someone from the outside-in perspective, it can be that easy to replicate this with no code. The issue however, which is the same issues I brought up in these vendor talks, is the fact you put 100% trust in the machine and the pre-generated code. It's one thing to use tools like ORM's where you can still validate the code and having people understand how to validate the code. It's another thing not knowing how to validate the code because you lack the code chops or in the data science space, the math chops with it comes down to the specific formulas. It's scary to think businesses out there are just relying on pre-generated code from so-called specialist who can't validate the final result because they lack the skills. Especially if it's something like a model that is telling the business where to spend it's millions of dollars.
No thank you.
August 21, 2018 at 9:01 am
Sergiy - Friday, August 17, 2018 3:21 PMNo-code platforms are basically interpreters from a human to the machine language.To see how effective those things are copy this post to Google Translate and do double translation - say, to Russian, or Bosnian and back to English.Would you be satisfied with such computer interpretation of your commands?
Definitely not. I can see the awfulness of Google Translate without doing double translations, as translating Scottish Gaelic to English with GT is already hopeless without translating back the other way. GT is fairly hopeless with Irish too. It's improving with both the Gaelic languages, but the improvement is very slow, and at current rate of progress it will continue to be awful with those languages until I'm in my coffin. Or, if google translate is to believed, starting from Scottish Gaelic "gus am bi mi 'n ciste chaoil nam bord" that should be "so that I can be a limb cock on board" (I haven't a clue either what that means or how google translate arrives at that) - if I go Gaelic -> Russian -> English it's much the same (the only change is "limb" changes to "member") ; if I translate the Gaelic to English myself, without any GT or other translation aid, I get "until I'm in my coffin" unless I'm being very literal (usually a bad idea unless the text is totally idiom free, and often even then) in which case it's "until I'm in a narrow wooden box".
Tom
Viewing 2 posts - 31 through 31 (of 31 total)
You must be logged in to reply to this topic. Login to reply