Computing Paradigms

I’ve been trying to pull together a list of computing paradigms. The list got long enough where it needed some structure:

Maybe you can help identify missing perspectives and/or categorizations?

1 Like

I think this is great @markNZed, thanks.

Some possible topics:

Thanks for the feedback and thanks even more for the suggestions. I added CSP. I think VSM fits better into another set of slides I’m trying to pull together: Grand Unifying Theories of Intelligence. I had forgotten about VSM so it is great to have that reminder.


Associative memory. Locality sensitive hashing. Bloom filters. Linear random projections, non-linear random projections.

Hi Sean, locality sensitive hashing, Bloom filters, Linear random projections, and non-linear random projections might be techniques or algorithms that would fit within different computing paradigms. Associative memory raises the question of whether different approaches to memory organisation are within a computing paradigm or lead to a different compute paradigm, if we start listing all the different ways data can be strutured then it will lead toward things like databases. For now I’ve added a section “Relational” in the slide 8 on “Representations” with databases and associative memory. Please let me know how you think we should integrate your other suggestions. Cheers, Mark

I don’t mind at all. I guess hash tables, Zobrist hashing and tabulation hashing can go under Relational as well. Which are totally underused things in AI.
You know in a computer game if you attach a specific fixed random number to each sprite type, then if 2 sprites collide you can xor their numbers and see if an action should take place with a simple if statement. You can avoid a lot of specific code that way, or having to design large interaction tables.
If 2 sprites hit you can just get the joint hit number h=x xor y for the two sprites and write if(h=…) then …
You could use a switch statement for even greater efficiency.
That is just one tiny example.

1 Like