>>13169533This is the biggest larp dunning Kruger post I've read today.
You will learn what you need to learn in grad school for ML/AI. Your background isn't nearly as important as the projects / research you've been involved with.
>discrete algorithms ...as opposed to what? All algorithms we write for computer are discrete in two senses: they have a finite number of steps in their description and they work with at most countable structures. Even when you start working with things from continuous spaces, you have to be able to discretize in a way that preserves the properties you care about.
>OS and lots of other bullshitA big part of ML, especially the practical side, is knowing how to design / deal with custom architecture and how to do massive data retrieval. These problems are nontrivial because of the scale at which you need to work with. A CS major learning these topics is fine.
>applied mathematicsyeah fuck off. You can't even be fucked to mention the actually useful math for ML, which falls under functional analysis, differential geometry, etc.. which falls under what you'd learn in a pure math degree anyway.
It's super clear that you're an applied math major trying to boost up your confidence by dunking on muh CS. I swear, fucking /sci/ undergrad fags treat intro analysis and algebra like it's a lost sacred art form. It's not. People pass these classes every semester. If you can recognize basic patterns, you can do basic analysis on the real line. All the real math you need for real ML, ie to understand RKHS, are things that take time but you can pick up.
If physics PhD candidates can get into ML having no analysis background, CS majors can as well. You have an inability to frame knowledge pretty trivial in the face of intuition, mathematical maturity, and design sense. I almost believe your larp, because you sound just as stupid as your typical HR department at a tech company.
