Comments by "iorekby" (@iorekby) on "The Top 5 Programming Languages in 2021 to get a job" video.

Amjadsalem Gran He gives an "honorable mention" to Ruby at 08:27 but says that a) it has a very low market share and b) it's falling quite a bit each year. TL;DR Right now, the future looks bleak for Ruby.
40
@reasonerenlightened2456 That level of "self awareness" exists mainly in Science fiction. We're decades and decades away, anything even close to actual fully automated programming. And they could end up being a black box. Do you fancy debugging a million lines of C++ some crazy coding AI came up with, with zero documentation or explanations for the "thought process"? Neither do I.
26
Exactly. C# is basically Microsoft's version of Java with some improvements.
13
@reasonerenlightened2456 I appreciate what you're trying to say, but you're being really reductive here and I suggest you do some more research.
10
@mugenn56 How so? They made a serious comment, responded to my response seriously. If it were just a joke ofc I'd leave it but they were obviously asking a serious question. Whooosh yourself lol.
10
@reasonerenlightened2456 Again you are being completely reductive. It's like saying I want to build something as tall as the Empire State building. All I need to is get people to build it. "GET ON WITH IT" It's low resolution thinking. It doesn't get into the minutiae of the actual complexities of the task, the cost or the minutiae of the actual task. In the AI issue, it doesn't address the limitations of AI, the extraordinary computational cost of such tasks or the engineering difficulties. Tell you what, if it's so simple, why don't you do it? Seriously. If you think your overly simplistic analysis is correct, and simple to do, "get on with it" and make it yourself. Otherwise, I'll repeat myself: You need to educate yourself on this and think about it a bit more.
9
@reasonerenlightened2456 That's just irrelevant gibberish. You are being reductive; you are over-simplifying a complex challenging issue and now you have made a rod for your own back by saying it could be done simply, and yet you are avoiding the challenge of doing this "simple" task yourself. At the moment, it's coming across as you taking lack of domain knowledge and a request for you to increase your domain knowledge as a personal attack, and you are resorting to posturing online to save face. So, I'll say it again: If you truly believe this to be simple, build the AI yourself and share your Git. That's the only way to show you have indeed thought this through and have a sufficient understanding of the domain.
6
@tasheemhargrove9650 I agree, C++ has a steeper learning curve for sure. JS, more or less anyone can pick it up and do it. The other issue is the domain spaces those languages operate in have way differing levels of complexity. JS is used for a few things but primarily web/mobile dev, which is generally not that complicated. You don't need a deep foundational knowledge of much to build a website beyond some JS/CSS/HTML and some good UI principles. C/C++ on the other hand deals with pretty complex domains, like operating systems, networks, cryptography etc... you'd need a pretty strong background in say electrical engineering or mathematics before tackling that stuff.
5
@reasonerenlightened2456 More hogwash. Google will give literally 10s of thousands for free to any such project through GCP credits. Provide the cost calculations for your cost for your "simple" solution. Resorting to ad hominem snides instead of backing up what you are saying is simply discrediting you.
5
@reasonerenlightened2456 I'll summarise: You claim to have a "simple" solution to a complex problem and have not demonstrated any discernible knowledge on the domain space. When asked to contstruct this "simple" solution, you refuse to answer initially, then claim it is too expensive. When asked to provide costs for this "simple" soluition, you repeatedly ignore the asks. Your retorts to these reasonable requests have been petulant ad hominem attacks questioning someone's intelligence simply for scrutinizing your facile "solution". Do you really think any of that is making me look bad? Any reasonable person can see you're made some silly remarks, and rather than providing a working proof of concept or actual tangible detail you've debased yourself with frankly juvenile responses wrapped in quasi-intellectual nonsense. By all means continue if you wish. It's not making me look bad, it's only making you look bad.
5
Python is great but boy is it overhyped in terms of "taking over". It really isn't. It's great at some things but terrible at others. It's not as good a language for OOP support compared to say, Java, and like OOP or not it's still the dominant paradigm in the marketplace today (for now). I think Python gets hyped so much simply because it's easier to teach people with no programming experience, ergo a lot of YT channels which are sponsored by bootcamps etc... push Python. Rather than seriously looking at Python in the marketplace and it's trends. In Western Europe where I live, the big 3 languages today on the job market for devs are Java, JS and C#. It's been that way for at least the last decade. Python hasn't made this huge dent some people on other YT channels claim it has.
4
@reasonerenlightened2456 More drivel. Google provide software solutions, they would want a more efficient and less error prone solution to employing thousands of software developers. Google would make something proprietary in such cases which you would know if you'd a) worked on an AI solution before b) actually applied for GCP credits. None of that by the way stops you from providing detail costing of this "simple" solution you keep talking about. So where are they? Provide them.
4
@reasonerenlightened2456 I was about to respond to your comments when I noticed one of them disappeared, no doubt from YT who took offense to your racial slur. Should that not give you pause for thought? And what you said makes no sense: If it's such a monumental paradigm shift, how do you know what the costs will or won't be and ipso facto how can you tell how expensive it will be? You obviously cannot. If you cannot refrain from being uncivil/offensive, don't expect YT to keep your replies on here.
3
@reasonerenlightened2456 You're continuing deflection. You've painted yourself in to a corner by making grandiose claims with no tangible way to prove or implement them. You then resort to ad hominem attacks which it seems YT keep deleting, so once again does that not tell you something? You then arrogantly compare yourself to bona fide scientists and engineers who achieved great things. It's just hucksterism and cheap debating tactics of blathering idiocy to hope the other side simply acquiesces to the tsunami of nonsense. Even if the costs are prohibitive (which you still have not shown) that does not stop you going in to detail about the AI/DL algorithms you'd use in your "simple" solution, for example. You have provided no details, no proof you understand any of this. You have made ad hominem attacks that have repeatedly broken YT standards. By all means continue with this line, but your phony intellectualism, arrogance and hostility will likely end up with you being banned from YT. By all means carry on.
3
Think you mean C++ bud but that's still a very good point. For sure C++ devs would be making more. There are wayyyy less of them available, and the spaces they normally work in (e.g. fintech) pay serious dough. And C++ usually requires more CS or electrical engineering knowledge that Python doesn't necessarily. You could get away with knowing little about CS and nothing about EE as a C++ dev. You couldn't really get away with that as a C++ dev. So, C++ Devs get paid more for that extra know-how.
3
@tasheemhargrove9650 Relative to say coding an OS though, that stuff isn't anywhere near as complicated and requires less domain knowledge. And that stuff you said isn't really complex to understand for most people. The challenge comes in managing a large code base. Sure, it's not as easy as say frying some eggs, but relatively speaking it's not as complicated.
2
@reasonerenlightened2456 It's not pathetic, it's simply asking you to detail your "simple" solution without any esoteric gibberish. The fact that you quickly descended in to childish insults simply shows you can't substantiate what you've said, and are trying to get away with making outlandish claims but not being held to account for them. You could provide more information if you knew what you were talking about but you obviously don't, and the infantile returning to a conversation even after YT deleted some of your previous comments is frankly pathetic.
2
@reasonerenlightened2456 I'll make it simpler for you then: Assume that your god-like intellectual abilities are above mere mortals like me (which I'm sure your arrogance can accommodate). But then also realize it's not just me reading these comments. Perhaps some other Olympian-like intellect will happen across this discourse. Explain it to them. Break it down step by step, with copious detail. The fact you continually return to ad hominem responses simply shows how utterly vacuous and pretentious you're being here.
2
@reasonerenlightened2456 You have simply given a very high level overview of this paradigm of yours but haven't given any minutiae on the detail. You mention LISP previously. When McCarthy created LISP he didn't just say "Here's a notation system for computers" He broke down the theory and steps of what he'd worked on and produced papers explaining his work: http://www-formal.stanford.edu/jmc/recursive.pdf Compare that brief paper to what you've provided. You've basically said "Get AIs to observe and replicate programmers". That's nothing really. It's not even a something that could be considered a model. It's just a very reductive analysis on what you perceive the problem to be, and you your paradigm to provide a solution is incredibly simplistic and bereft of any scientific or engineering detail. When asked for detail, you've resorted to childish name calling and questioning the intelligence of anyone who is skeptical of your overly simplistic interpretation of the problem and the supposed paradigm to offer a solution to the problem. The fact that you have responded with personal barbs rather than functional detail is not a good look. McCarthy didn't start calling people names because they asked for a white paper on LISP. If you want to aspire to be on the level of McCarthy, do the work McCarthy did. Write a white paper for peer review to show in detail the proposed paradigm. Stop hiding behind cheap debate tricks a high school student would be embarrassed to use and show you understand the problem and the domain space. You really haven't done that yet.
2
@reasonerenlightened2456 I'm not confused about anything because you haven't actually said anything. You've essentially said we need a new paradigm, let's get the AI to mimic programmers and propagate this to build up an AI model of sufficient robustness. That's nothing. It's like me saying I want to build the world's fastest car that has an engine that makes it go faster than any other car. It's a facile statement, not a "paradigm" or a solution. I don't want you to use Goggle and Control + C and Control + V a specific response to a question, any semi-literate person could do that. I want you to do something you have repeatedly failed to do: Spell out, in copious detail, how your proposed paradigm will work, including all mathematical, scientific and engineering information related to it. I have asked for this about half a dozen times now and you are still avoiding providing this information despite you claiming it's so very "simple".
2
@reasonerenlightened2456 I don't have an ego about this, I can freely admit it's a possibility that the issue is my inability to grasp something here. It's a possibility. Now, can you admit, freely that it's possible that: a) you don't understand this domain as well as you think b) you cannot articulate whatever it is you're trying to say and the failure lies with you.
2
I was browsing the other day and found a forum where the author claimed with strong authority Java was dead, and would be obsolete in 2 years. Their post was made in 2003 lol!! People saying "Java is dead" are really saying "I don't like Java/find it too confusing and I wish it would go away", rather than having any tangible reason for thinking Java is "dead".
2
@dcngn_ Same in Ireland and the UK, Java is the lingua franca for teaching CS and SE students in those countries at university. I had a friend from France who said he learnt Java in his CS degree in France too. It seems pretty common in Europe.
2
@danielrazulay He mentions Go/Rust but said the market share isn't there yet to make it on to his top 5, but it is growing.
1
@reasonerenlightened2456 Ha ha that's a new low. I don't know whether this is just ego or delusion on your part. You've made repeated grandiose claims that are absolutely vacuous and meaningless, made pretty crass yet rudimentary logical fallacies from ad hominem responses to argumentum ab auctoritate, and blended them with a debate tactic a 12 year old wouldn't use: The Gish Gallop. What you have not done, at any point, is demonstrated any actual knowledge about AI or more specifically the problem around coding automation. You've simply named dropped other people you seem to believe are your peers of yours while behaving in an incredible rude manner simply because someone asked you to demonstrate you have knowledge on the domain in which you felt inclined to preach stentorian from the mount about. You have not only repeatedly failed to do that, you responded in a petulant, uneducated manner, to the point where Youtube actually remove some of your more offensive tantrums. That says it all really.
1
@joecalderon3158 Perl 7 was released last year too. Do you think Perl is making a big comeback? Any language with a modest community will still continue to develop new features and iterations. Doesn't mean they are going to regain their marketshare anytime soon though.
1
It definitely is back en vogue again. With PHP 8 not long here, and the speed/JIT enhancements it's a very attractive option for web/mobile dev. Plus people are doing more with it beyond web dev e.g. ML, so it's definitely getting some heat again.
1
@lexsoft3969 Of course. There are ML libraries and frameworks for most major languages. Some of them are more awkward to use, and Python is still hard to beat in terms of the amount of libraries and community support for implementing ML. Basically all machine learning is is Computational Statistics with the emphasis on Statistics . You are just implementing mathematical/statistical algorithms against data. You want to do it in a way that uses computational resources in the most efficient way though. Low level Libraries are usually written in C and C++, and basically you can use Python, Java or even Kotlin to make calls to those libraries at a high level to implement your algorithms.
1