View source for A quick draft for a curriculum for Computer Science

{{pubdate|13|February|2022}}

The other day, on Facebook, I was [https://www.facebook.com/vrandezo/posts/525194596099 asking the question who would be the person closest to being a popularizer for ideas in Computer Science] to the wider audience, which lead to [https://www.facebook.com/vrandezo/posts/525194596099 an interesting and insightful discussion].

Pat Hayes asked what I would consider the five (or so) core concepts of Computer Science. Ernest Davis answer with the following short list (not in any particular order):
# Virtual machine
# Caching
# Algorithm
# Data structure
# Programming language

And I followed up with this drafty, much longer answer:

# how and why computation works; that a computation is a mapping from your problem domain into some machine state, then we have some automatic movement, and the result represents an answer to your question; that it is always layers of interpretation; that it doesn't matter whether the computing machine is made of ICs or of levers, marbles, and gravity (i.e. what is a function); that computation is always real and you can't simulate computation; what can be done with computation and what cannot; computational thinking - this might map to number 1 in Ernest's list
# that everything can be represented with zeros and ones, but doesn't have to be; it could also be represented by A and B and Cs, and many other ways; that two states are simply convenient for electric devices; that all information, all data, all input to all computation, and the steps for computations themselves are represented with zeros and ones (i.e. the von Neumann architecture and binary encoding); what can be represented in this paradigm and what cannot - this might map to number 4 in Ernest's list
# how are functions encoded; how many different functions can have the same results; how wildly different in efficiency functions can be even when they have the same result; why that makes some things quick to calculate whereas others take a long time; basically smearing ideas from lambda calculus and assembler and building everything from NAND circuits; why this all maps to higher level languages such as JavaScript - this might map to ideas from 2, 3, and 5 on Ernest's list
# bringing it back to the devices; where does, physically, the computation happen, where is physically the data stored, and why it matters in terms of privacy, equity, convenience, economics, interdependence, even freedom and independence; what kind of computations and data storage we can expect to have in our mobile phones, in a data center, in an RFID card; how long the turnaround times are in each case; how cryptography works and what kind of guarantees it can provide; why centralization is so alluring and what the price of that might be; and what might be the cost of computation for the environment
# given our times, and building on the previous lessons, what is the role of machine learning; how does it actually work, why does it work as good as it does, and why does it not work when it doesn't and where can't it work; what does this have to with "intelligence", if it does; what becomes possible because of these methods, and what it costs; why these methods may reinforce inequities; but also how they might help us with significantly increasing access to better health care for many people are allow computers to have much more intuitive interfaces and thus democratize access to computing resources

I think the intuitions in 1, 2, and maybe 3 are really the core of computer science, and then 4 and 5 provide shortcuts to important questions four ourselves and society that, I think, would be worthwhile for everyone to ponder and have an informed understanding of the situation so that they can meaningfully make relevant decisions.

{{tag|Simia}}
<noinclude>{{simiapost|english}}</noinclude>