/sci/ - Science & Math

>java code

>>11131316
i bet that has something to do with
(((Microsoft))) acquiring Github

This would be AI complete if true

>>11131316
>be incel code monkey
>aggressively yell at everyone about how their jobs are retarded and how they will easily be automated in 20-30 years by robots
>get 90% of your jobs in the industry automated during the next 10 years when Google and Microsoft launch an AI assistant for software design

>>11131338
meanwhile, we mathematicians are comfy knowing that our job will never be taken by robots

>>11131484
someone tell him

>>11131499

Mathematicians actually are robots and /sci/ is just where incels come after they've masturbate before shortly returning to /r9k/?

>>11131316
lol, who do you think made this? Distinction between software developers and computer scientists, typical /sci/ poster
you fail to realize most of CS academia is made up of people who double majored in math and CS, especially those in theory subfields, right?

>>11131484
https://en.wikipedia.org/wiki/Automated_theorem_proving

one day

>>11131338
>the safe jobs are ones that require empathy or manual dexterity
>give handjobs
>laugh in my job security

>>11131316
I thought this was going to be real, but it isn't sadly.

>You might notice that once you start typing in “call:readLine”, a drop-down box with suggestions will appear below the cursor. You can simply select an element from this list to auto-complete it instead of typing it in full. In fact, it is required that all queries be elements that occur in this list – otherwise an error would be shown when issuing the query to Bayou.

You still have to type every single thing that you want it to do, just with shorthand, all this is, is a large parser/generator for a specific set of shorthand. In other words, this saves me about as much time programming as tab-completion in Eclipse or Intellij already does.

>>11131484
Once it becomes common for mathematicians to write their proofs in a formal language that can be checked by a proof-checker, an AI will be unleashed on all of that data.

>>11131577
You are clueless about your own field.
The first computers needed a half a dozen of technicians to simply do basic operations, today you barely need one.

With the advance of languages, libraries and compilers programming is now accessible even to brainlets and that's exactly why it will be automated much sooner than expected.

Most of the code in an applications today is absolutely trivial, you even see the average code monkey brag about what they really do half of the day is copy pasting stack overflow.
Google hires any retard who did a bootcamp specifically because they currently need little hands to maintain their codebase and do tasks assigned by the software architect.
The moment they automate this they will save tens of billions a year, you will only need one or two guys to handle a project requiring 15 people today. App builders like bubble.is (which doesn't even use any machine learning) are a good example of how it begins.

CStards think they are sheltered from the trend because they believe inputing basic logic with a keyboard is hard. But the truth is that they are on the frontline and unlike other jobs there are no material, economical or social obstacles to their automation.

If your job mostly consists in inputting shits in a computer then you are the most accessible thing to replace in the near future.

>>11131316
>companies that develop software's biggest cost is developing software
>>n-n-no they won't replace me the good little code monkey they need me
>mfw sitting here studying a field that is about creativity, art, and human thought.
>>11131484
based cedric villani poster
>>11132656
>thinking math is only about being formal and rigorous
HAHHAHHHAHAHHHAHAHHAHA
hurr duur calculus is the epitome of math

>>11132735
Hell yeah, tell em mathbro
Sorry computer inputists, soon it will be over

>>11133061
the confident man shot himself in admiration of his own firearm.
>me talking to CS majors in a few years.

Millions of Kode artisans and experts will rise up as one and destroy any company that attempts to implement this.

>>11131338
As a person who knows how to write AI, and also work as a software engineer, I really don't think that an AI could write the things I do. It's not as much code as it is system design and architecture. Writing the code itself is the easy part, designing the solution and configuring the different parts is the problem.

>>11131316

Ted predicted this

>>11133169
DMS SRT Semantic Designs.
Amend a blackbox. No AI.
It wont be the military doing what is quoted.

>>11133213
>Future very scary, better give up and blow up some buildings, that will certainly fix the problem
Even ironically idolizing this madman is a slippery slope, you'll find yourself in a rubber room next to him. Not to mention the fact that his beliefs are completely anti-human

>>11131484
I agree, but only because doing math is a cultural endeavour.
It's a realistic-drawing vs. taking-a-photo kind of relation. Clever exhaustive automated search in and beyond mathematical topics that people are still good in could very well take the swing out of any math topic, just like Dürer pictures don't have the same impact as for people who don't know that you could take photos of things that then look like the original.

>we mathematicians are comfy knowing that our job
My point is that it shouldn't be seen as "job" at all, not even now.

>>11133232
>cultural endeavour
And I mean "cultural endeavour" as in coffee culture. We try to make sense on Grothendiecks ideas and build on them like we curate themes, techniques and topics in cinematography - we do math not actually to solve the problems and (as pretty evident in the algebraic geometry case) not to help pump some efficiency of a process like an engineer is supposed to. (Algebraic topology may find application in computational problem, but that's more to the detriment of math than to its advantage.)

>>11132728
You legitimately ignored everything in my post. I am aware of the number of people it took to keep machines working decades ago, but my comments were not about babysitting computers.
I made a statement about the difference between CS and software engineering as career pursuits and from their internal culture.
You went on to spurt out more about “waaah codemonkeys!” I realize this was inevitable because I’m almost certain you have no idea what academic CS is, and I’d probably surmise that if I didn’t get a answer about codemonkeying, you’d spurt out another meme like ML.

Anyway, your post isn’t exactly relevant to the nontrivial work out there. Automation of the menial is inevitable but even specifying simple tasks takes a heap of work. I really don’t see ML or even general forms of AI taking over cryptosystem design - I see large scale AI assistants in that field being used by actual qualified software engineers to help focus on core design and implementation

>>11131328
>(((Thanks for all those projects you've been spending thousands of hours on for free.... )))