Mark Cuban says AI will reduce the demand for computer science degrees.

[Nobu]

Skilled software engineers will be in demand for the forseeable future. If all you can do is setup a basic Wix or Wordpress website and can't solve novel problems, you should be worried. But AI is not going to be creating and architecting and maintaining and troubleshooting complex software systems on it's own anytime soon.

Software development is not just setting up a pretty looking website.

 

[rapbeats]

Skilled software engineers will be in demand for the forseeable future. If all you can do is setup a basic Wix or Wordpress website and can't solve novel problems, you should be worried. But AI is not going to be creating and architecting and maintaining and troubleshooting complex software systems on it's own anytime soon.

Software development is not just setting up a pretty looking website.

but you still aint getting it. a basic wix website and all that basic in reality. Just because its now the norm doesnt make it basic. I remember when them super basic websites were being programmed for back in the 90's. some of your well known companies had trash websites in my opinion. but there were real website programmers coding for them. Then everyone started stepping their games up. Then they had old school versions of wix that made you a website but it looked like trash and it was hard to navigate. and finally here comes wix. Wix has put the regular joe blow HTML guy at home out of business. The best he can do is get good at Wix customization and hope the people that pays him to do their sites really hate thinking about putting a website together even if its as easy as wix. Most people/business dont need a lot of extras outside of what wix can offer online. Every generation that grows up with a WIX type of option means those people will grow up doing it themselves not asking for someone to help them with their website. you see my point? old heads might still ask. but not anyone under 35. they are at least going to give wix a go first before paying you. and thats what will happen with this AI.

 

[Nobu]

but you still aint getting it. a basic wix website and all that basic in reality. Just because its now the norm doesnt make it basic. I remember when them super basic websites were being programmed for back in the 90's. some of your well known companies had trash websites in my opinion. but there were real website programmers coding for them. Then everyone started stepping their games up. Then they had old school versions of wix that made you a website but it looked like trash and it was hard to navigate. and finally here comes wix. Wix has put the regular joe blow HTML guy at home out of business. The best he can do is get good at Wix customization and hope the people that pays him to do their sites really hate thinking about putting a website together even if its as easy as wix. Most people/business dont need a lot of extras outside of what wix can offer online. Every generation that grows up with a WIX type of option means those people will grow up doing it themselves not asking for someone to help them with their website. you see my point? old heads might still ask. but not anyone under 35. they are at least going to give wix a go first before paying you. and thats what will happen with this AI.

It's a logical fallacy to assume that because developing a basic website used to be perceived as hard and has since been somewhat automated, it must mean that all programming related tasks are just as easy to automate.

Again, developing static webpages is not at all representative of what software engineering as a field is.

Think about it this way; even humans are so bad at software engineering that good, experienced, but not genius level software engineers can command $300k-$400k+ in total compensation. Top interns like myself get $50+/hr, and yet most software even in the great firms is shyt. It breaks all the time for reasons unknown and basically we are just patching it up with more shyt to keep services running. If highly compensated humans can't write good software or good general Artificial Intelligence yet, I won't hold my breathe on AI developing complex software systems anytime soon. It will happen when general Artificial Intelligence happens, which appears to be very far off.

 

[Dr. Acula]

There is a real life example called deep coder.

https://openreview.net/pdf?id=ByldLrqlx

^^the white paper.
AI learns to write its own code by stealing from other programs
or


this is from that article


Thats exactly what i just typed in a previous reply to this thread. people will be able to create things without having to be actual coders and without having to have the knowledge of syntax for each coding language. they will just say what they want and the AI will go from there. This means the better the person is at leading the AI to create things of great use, the more money said person will make. Being creative is one thing, but being creative where it gives a lot of people either some new thing to buy for fun or to wear, or to drive, or to listen to, watch, etc. or create some thing(s) to help people in some way.

From the findings

There remain some limitations, however. First, the programs we can synthesize are only the simplest problems on programming competition websites and are simpler than most competition problems. Many problems require more complex algorithmic solutions like dynamic programming and search, which are currently beyond our reach. Our chosen DSL currently cannot express solutions to many problems. To do so, it would need to be extended by adding more primitives and allow for more flexibility in program constructs (such as allowing loops). Second, we currently use five input-output examples with relatively large integer values (up to 256 in magnitude), which are probably more informative than typical (smaller) examples. While we remain optimistic about LIPS’s applicability as the DSL becomes more complex and the input-output examples become less informative, it remains to be seen what the magnitude of these effects are as we move towards solving large subsets of programming competition problems.

There is a popular website most likely used and not named in this study called Topcoder where the competitive challenges were pulled from.
Community - Competitive Programming

As someone who has actually used the website and done the challenges, when they say the simplest problems it might mean something like.

An actual problem from the site (you need to register to see it)
Topcoder Arena

You are given two s: N and K. Lun the dog is interested in strings that satisfy the following conditions:

• The string has exactly N characters, each of which is either 'A' or 'B'.
• The string s has exactly K pairs (i, j) (0 <= i < j <= N-1) such that s[i] = 'A' and s[j] = 'B'.

If there exists a string that satisfies the conditions, find and return any such string. Otherwise, return an empty string.

They also give you the function definition and parameters too so you already are given what structure your function should be constructed

Class:

AB
Method:

createString
Parameters:
int, int

Returns:
string

Method signature:
string createString(int N, int K)
(be sure your method is public)

This is a simple "if then else" problem. Where you're doing a conditional check on if something is true, in this case the string conditions are met for the strings where its a certain length, and the characters can only be A and B and there must be K amount of pairs in the string to satisfy a true condition. If found, return the string, else return "".

Fairly simple problem and this complicated AI programming could only do problems of this difficulty. This is like 3rd week stuff you do in your intro to programming course freshman year, first semester in your 4 year CS coursework.


In short, it will be A LONG TIME before automation replaces programmers. Also to iterate Mark Cuban sounds like an idiot. Not saying it won't ever happen but all this recent fear mongering of the idea in the past year or two is weird given the technology isn't anywhere near being there. Why now? Can't help but feel its part of a propaganda campaign to get workers ready for more unemployment and become comfortable with "Basic universal income".

 

[KillbertArenas]

Everyone believes their job is automation proof until it's too late. It'll be an interesting development to follow.

I drive a bus for Metro in DC

How long you think I got??

lol

 

[Rarely-Wrong Liggins]

I drive a bus for Metro in DC

How long you think I got??

lol

Us drivers have awhile. I'd say a couple of more decades before we're phased out en masse. Society prefers behind the scenes or low level work to be automated. Most people aren't ready for daily interactions with machines doing complex jobs.

 

[dora_da_destroyer]

"Creativity, collaboration, communication skills: Those things are super important and are going to be the difference between make or break," Cuban said. "In an AI world, you have to be knowledgeable about something, right?"
.

Communications degree gang we finally winning

 

[dora_da_destroyer]

Certain jobs yeah.



:laughatpeasantry:

Mark Cuban doesn't know what he's talking about. It'll just be a different type of coding. People won't be coding from scratch but from whatever API the system has available.

He’s right on this one, he may be overstating the (timeline of the) impact, but the AI, deep learning, and neaural networks people are building now are meant to train themselves, learning from more data than a human could ever hope to process in a lifetime. The more data, the more refined their learning and better their ability to “think” is. They’ve written algorithms that allow machines to identify bad code in their own program and repair it themselves.

AI is tackling complex thinking problems today, so of course it would be able to code, solve UI/UX problems, configure network architecture, build and manage databases, build and prioritize product roadmaps, build and test software, structure and analyze data (data science), complete and publish research, and of course build models to train other machines. That’s a lot of the areas CS majors are employed in (they’re not just coders), deep learning is still in its infancy yet accomplishing mindblowing shyt. Between the rate of advancement and the fact that so many people in this field participate in open souring projects (code sharing) there will definitely be an impact in the next 20-25 years.

 

[dora_da_destroyer]

They've been saying this shyt forever and we're still not any closer.

Coding isn't just math. You will not get innovative code just using math. In fact I'd argue coding is very light on math in a x+y=z context. There is a lot more logical deduction used than math with coding.

Also all these think pieces on ai coders yet never a mention of a working example of this being done.

Not sure why cuban is quoted. He was a software distributer. Not coder.

New A.I. application can write its own code - Futurity
Google's AI can create better machine-learning code than the researchers who made it

 

[dora_da_destroyer]

It's a logical fallacy to assume that because developing a basic website used to be perceived as hard and has since been somewhat automated, it must mean that all programming related tasks are just as easy to automate.

Again, developing static webpages is not at all representative of what software engineering as a field is.

Think about it this way; even humans are so bad at software engineering that good, experienced, but not genius level software engineers can command $300k-$400k+ in total compensation. Top interns like myself get $50+/hr, and yet most software even in the great firms is shyt. It breaks all the time for reasons unknown and basically we are just patching it up with more shyt to keep services running. If highly compensated humans can't write good software or good general Artificial Intelligence yet, I won't hold my breathe on AI developing complex software systems anytime soon. It will happen when general Artificial Intelligence happens, which appears to be very far off.

Coding isn’t the only thing comp sci majors do...

Furthermore, a big reason things break is because some small syntax error is missed or when discrete pieces of code are finally linked in the software, there’s a bug/bad connection. AI alogirthms would automatically produce perfect code, could code faster (24-7), do bigger projects themselves, and also run tests/models on its own OS to look for bugs that wouldn’t have to be discovered in a whole separate QA process, or even worse, by customers.

 

[Secure Da Bag]

He’s right on this one, he may be overstating the (timeline of the) impact

That's exactly what he's doing. 20 years? Nah. It'll be much further along than that. Not to mention Quantum Computing/Programming.

 

[dora_da_destroyer]

That's exactly what he's doing. 20 years? Nah. It'll be much further along than that. Not to mention Quantum Computing/Programming.

I don’t think he is, but I’m also not taking it as every single coding role can/will be automated. But a lot of rank and file coding, the vast majority of employment in that field, could either be filled by AI or so supplemented by it that it devalues the role/degree...which is exactly what he said, the degree (which isn’t the best way to talk about coding as that’s just one segment to which a comp sci degree is usually applied) will diminish over time - not have no value or utility. 20 years is very much a realistic timeframe for AI to be extremely commonplace in generic coding roles


And quantum computing is even more nascent than AI, a long ways off, and as AI continues to be built out, it should be in a place to assist in quantum computing greatly once quantum computing is in the phase of the life cycle that AI is currently in. But bringing up a hyperspecialization isnt much of an argument against his point, the vast majority of comp sci people are not in quantum computing nor doing extremely high level research/dev...that’s like using private equity as a representation of the finance industry...it’s way too small a segment to represent where the majority of people employed in finance fall.

 

[rapbeats]

Wix makes the process simpler, but doesn't eliminate the need for people to develop websites. If your end goal is to just build a website, it's great. But if we are building something that scales well, It's going to be coded by developers. You can't drop and paste backend components. Even with wordpress, you still need web knowledge and companies pay up to the wazoo for it.

i agree. and tell me, how many random businesses need all of that? They dont. They just need a web presence. thats it.

one long scrolling page with what they are about, a few pics, some write ups, and a couple of links. thats it. thats all most businesses want and need at this point in time.

i'm talking about companies that dont do business online. meaning they dont and wont sell products nor services online via e-commerce. They just want a presence on the web like its the old school yellow pages. thats it.
If i'm selling my consultant services. i have a contact us page, I mAY have a few packages showing you what you could get for what price, but ultimately i want you to email us and call us on the phone. I'm not doing actual transactions over the internet. thats a nice simple wix page.

for all the wanna be artists out there. Wix all day long. This was not the case before wix like programs came out. they've had them before, but a lot of those programs were so super stale and had very little options. that every 3 websites would be identical . so people would still end up hiring a web person to do their website to be different. Wix kills that level of web designer. because it looks professional, and there are enough options and enough light weight customization where you can have your own look going within an hour or so without programming knowledge. the better designer you are the better your page will look. you dont need a HTML person at all. You dont need to know a lick of CSS. and the fact that it works on multiple browsers as well as mobile (apples and android) shoot, you cant lose with wix. that alone was the thing that separated a web developer from one of those old school wix like software. they looked good on one or two browsers and were trash on everything else. wix is functional on everything.

because again, if you're not conducting heavy business online, where you're using your website as a server to hold media, or a bunch of different transactions. then wix is the way to go. the better you can design your stuff the better your site will look and thats all you need at that point. people that are good at coding html or not always that great at the design aspect.

 

[Secure Da Bag]

I don’t think he is, but I’m also not taking it as every single coding role can/will be automated. But a lot of rank and file coding, the vast majority of employment in that field, could either be filled by AI or so supplemented by it that it devalues the role/degree...which is exactly what he said, the degree (which isn’t the best way to talk about coding as that’s just one segment to which a comp sci degree is usually applied) will diminish over time - not have no value or utility. 20 years is very much a realistic timeframe for AI to be extremely commonplace in generic coding roles

The things that WiX and the like can do will be automated for the most part. In 20 years, I can see that. But the rest is wishful thinking. The AI scare is to get more people to ask for UBI so companies can lower their costs. IMO.

 

[dora_da_destroyer]

The things that WiX and the like can do will be automated for the most part. In 20 years, I can see that. But the rest is wishful thinking. The AI scare is to get more people to ask for UBI so companies can lower their costs. IMO.

i don't know what you mean by "the rest", but more coding than website building could be automated. whether it's fully replacing entry level programmers or automating so much of their job that the price they command drops, either outcome diminishes the value of their degree.

and the only AI "scare" is all the sci fi driven "what ifs" that the media pushes for clicks. if you're referring to the automation scare (AI is much bigger than automation), well that does pose a big threat to various types of low level labor and is a threat. don't know about it being connected to UBI because i never see that passing in the US