/sci/ - Science & Math » Thread #13281164

17KiB, 275x183, download.jpg

View Same Google iqdb SauceNAO

Anonymous Tue 15 Jun 17:00:21 2021 No.13281164 View Reply Original Report

Quoted By: >>13281340 >>13281365 >>13281611 >>13282290 >>13282368 >>13282438 >>13282502

TL;DR Should I specialize in Quantum Computing or Artificial Intelligence?

I have a CS degree and 8 years SWE, and I'm interested in studying pure mathematics.
I want to combine my CS skills with math skills
but I'm torn between specializing in Quantum Computing vs Artificial Intelligence.

Both QC and AI require a strong understanding of Linear Algebra, so my curriculum is set for a while on brushing up my LA and learning Group Theory.
I scraped past LA in school by just learning how to do the computations, but I don't like how little I understand rigorously, so I'm focusing on understanding the proofs this time around.

The big pro/con I see right now is that in AI I can actually own the AI and play around with it, but in QC I'll be very limited in what I can compute without buying time on some company's computer.

What should I do /sci/?
I need your opinions
And does anyone have good books/resources on either QC or AI, Linear Algebra, Group Theory?

Anonymous

Anonymous Tue 15 Jun 2021 18:15:28 No.13281340 Report

Quoted By: >>13281478 >>13281486 >>13281769

>>13281164
working in either of these fields requires graduate study. not to learn it per say (though it's the most efficient way to learn it and get to the forefront) but for the basic credentials to get into the field
>just brushing up on LA
>learning group theory
Anon I'm telling you now that you need to work on your math fundamentals.
Hoffman and Kunze is good for LA
The MIT OCW courses on intro to AI are fine
de Wolf's notes and Mike'n'Ike are good for QC

a lot of the big work in these fields is locked behind harder math though.

Anonymous

Anonymous Tue 15 Jun 2021 18:28:06 No.13281365 Report

Quoted By: >>13281398

>>13281164
>AI AI AI AI
Machine learning is not AI.

Anonymous

Anonymous Tue 15 Jun 2021 18:38:23 No.13281398 Report

Quoted By: >>13282434

>>13281365
It's literally a subset of AI

Anonymous

Anonymous Tue 15 Jun 2021 19:02:38 No.13281478 Report

Quoted By:

>>13281340
Thanks for the suggestions!
What else should I work on for my fundamentals besides LA and group theory?

Anonymous

Anonymous Tue 15 Jun 2021 19:04:47 No.13281486 Report

Quoted By:

>>13281340
You saying I couldn't get into a job without a masters?

Anonymous

Anonymous Tue 15 Jun 2021 19:48:15 No.13281611 Report

Quoted By: >>13281769

>>13281164
I'm in QC and QT more generally. If you're coming from CS and you're not already strong in mathematics AND physics, don't come to QC.

It's already filling up with CS grads who are just trying to make the nearest facsimile of classical machines, and who are pushing out largely redundant research. There is a minority who understand enough maths to actually grasp the field, and therein to understand more fundamentally QAlgo development and the like, and an even smaller subset that also understand physics and therefore have any grasp of the hardware layer (which is necessary if you want to do anything substantial).

In terms of research, the plethora of CS memers who thing quantum is cool is becoming a joke. I don't advise it as a career unless your maths and physics abilities are as good as, well, mathematical quantum physics phds, because that's what you need.

If you are willing to get that good, go for it. There's a shortage of real talent, as with everywhere, but don't come in assuming that you'll be in a sub-discipline of CS.

Anonymous

Anonymous Tue 15 Jun 2021 20:45:56 No.13281769 Report

Quoted By: >>13282280

>>13281340
I found a Hoffman and Kunze 2e pdf!
Got any more good book suggestions?

>>13281611
My main interest is math, and I'm midwit popsci interested in physics.
I'm going to study pure math either way, but QC and AI seemed like the most interesting applications for me.
I'm super interested in studying formal systems and languages, so I'm trying to get better at reading properties and proofs.
I don't know if I'll be able to follow through all the way to QC or QT but it seems a challenge which aligns well with my interests.
I'm more interested in understanding how the universe works than I am about building a new computer though.

Anonymous

Anonymous Tue 15 Jun 2021 23:50:33 No.13282280 Report

Quoted By:

>>13281769
Being interested in how the universe works is the right motivation. Ontology is really what separates maths and physics, and, much as a mathematically inept physicist is useless, so is one with no creativity to links maths and theory to ontology and empiricism. You need a driving curiosity that is tied to understanding natural phenomena. Popsci is trash though, physics isn't nearly so obvious, nor so unnuanced - as in maths there's so much beauty, and it's not at the trivial surface level.

Pure maths is okay to study, I'd either have a foot in information theory and computation, or one in physics. There are some masters for QC now, you can also look into those. Be prepared to do a PhD in any case.

Anonymous

Anonymous Tue 15 Jun 2021 23:54:35 No.13282290 Report

Quoted By:

>>13281164
Cryptography has lots of linear algebra and group theory.
How come you want to get out of SWE? How did you go from CS-> SWE in the first place?

Anonymous

Anonymous Wed 16 Jun 2021 00:12:51 No.13282368 Report

Quoted By: >>13282392

>>13281164
Go artificial intelligence.

Quantum Computing isn't going to be completed by a human.

If you start trying to design a Quantum ISA, you will quickly realize, it's not possible for any human to create a fully encompassing Quantum ISA during their lifetime.

If you want to make the biggest possible impact that you can, on our current technology and understanding of one of these two fields, then artificial intelligence is the way to go, because it can potentially solve both problems.

Only working on Quantum Computing, may potentially solve neither problem.

The Quantum Computing issues are too tenuous for one human to overcome through an average human lifetime.

An AI has the potential to solve those issues more efficiently than any human could.

Anonymous

Anonymous Wed 16 Jun 2021 00:26:28 No.13282392 Report

Quoted By:

>>13282368
Well the two are also linked; QML is showing situational advantage, and more interesting HTM tasks and the like can be mapped onto QUBOs, which suit Q-Annealers very well.

As usual, the cross disciplinary work is the important part.

Anonymous

Anonymous Wed 16 Jun 2021 00:51:12 No.13282434 Report

Quoted By:

>>13281398
Texas is a subset of the USA

Texts =/= USA

Anonymous

Anonymous Wed 16 Jun 2021 00:52:55 No.13282438 Report

Quoted By:

>>13281164
Quantum computing is more PHYSICS than anything else. As a CS and Math guy, you're more suited to Machine Learning, which btw is not AI

Anonymous

Anonymous Wed 16 Jun 2021 01:12:04 No.13282502 Report

Quoted By: >>13282546

>>13281164
AI, be in the forefront of SKYNET

Anonymous

Anonymous Wed 16 Jun 2021 01:25:15 No.13282546 Report

Quoted By:

>>13282502
This, do your part or Roko's basilisk will keep bringing you back here to ask this stupid question with an obvious answer until you do.

Capcode	All Only User Posts Only Moderator Posts Only Admin Posts Only Developer Posts
Show Posts	All Only With Images Only Without Images
Deleted Posts	All Only Deleted Posts Only Non-Deleted Posts
Ghost Posts	All Only Ghost Posts Only Non-Ghost Posts
Post Type	All Only Sticky Threads Only Opening Posts Only Reply Posts
Results	All Grouped By Threads
Order	Latest Posts First Oldest Posts First

Your latest searches