Related Posts
What is salary package for AR VR Engineer?
Fishes, OK so this is true. Wipro is delaying the joining by half a month due to some restrictions or organisational changes & not ready to compensate for that half a month loss of salary. Very rude & absurd behaviour from them. Hence rejecting the offer as such company if u aren't sure when to ask for joining u may revoke the offer letter later on. Hence multiple options in hand is necessary nowadays. Infosys Tata Consultancy HCL Technologies
Any hot leads on Workers Compensation roles?
In-hand for below offer?

More Posts
Anyone ready for NYC gathering?
Additional Posts in Data & Analytics Consultants
Has anyone else begun to resent data science?
New to Fishbowl?
unlock all discussions on Fishbowl.




I’ve spent 8+ years in data science, and recently joined Microsoft. It’s crazy how easy it is to build models on Azure (I’m sure it’s the same for other cloud providers). Sure knowing the basics help, but these no code low code platforms are definitely a game changer.
And since COBOL is so easy for anyone to understand, there’s really no need for expert-level data engineers anymore since the 60s
Just look at how ETL tools went from manually writing code to low/no-code and for some odd reason everybody went back to writing manual code in python…
For directional accuracy sure, but for things that require precision it’s not good.
Like EY1 mentioned, ETL went back to coding because the complexity increased quite a bit. It’s not just move data from place a to b any more. The workflows are very complex and code management on git helps a ton!!
New software will remove a barrier to entry, but there will still be a huge premium for people who truly understand what they're doing.
Seems to me like this has already happened
Making a model is easy, making a model that scales to production on the cloud and is cost effective to run... less so.
For some applications you will absolutely need the data scientists that know their stuff, ninja level, for work that is more consequential, I.e.: development of new drugs/ treatments, trials data, for other applications, you can get close enough with low code tools, drag and drop and be okay, because the tolerance for error is acceptable. For data scientists I think the bar will be raised in terms of knowledge required to get in the field as there are a bunch of people out there running data science projects with prepackaged tooling.
Mentor
Given the prevalence of innumeracy I don’t feel threatened, I do worry about bad inferences en masse tho