@ki7sun3 and clients will have to want logically consistent things. We're safe.
@jeffcliff @ki7sun3 I wish all programmers where I work were replaced by machines.
That way I would understand them better and there would be no more
Dev - "The build clusters are broken, our build takes 2 times as long as last year"
Me - "The product you are building went from X lines of code to 4X lines of code since then"
Dev - "Oh.... Anyway we need builds to be faster"
@qrsbrwn @ki7sun3 did you catch the link i posted about GPT-3 as jsx programmer?

@jeffcliff @qrsbrwn GPT-3 looks nice and generates on me great expectations about programmers on the future more than in GPT-3 technology itself

@qrsbrwn @jeffcliff I'm in both sides (devops/programmer), I think if your colleagues think like that are not the best at understanding how hardware/OS/compiler work, but its my opinion based only on your side of the history =)

@ki7sun3 @jeffcliff I find that developers rarely understand computers at all. Wait, I'm gonna rephrase that.
Being very good at software development is a completely different skillset than understanding how computers work.
I'm not saying that one is better than the other or anything like that.
In my naive mind when I started working I kinda thought they were different parts of the same skill set. Now I know better. You can be a super talented and good developer and still not know anything about computer.


Neural networks... You no longer need to know what you want, all you really need is 1 000 000 examples of what you do or do not want, and the computer figures it out on it's own.

@LukeAlmighty sadly technologies like neural networks doesn't resolve the problem. We trains them to resolve particular problems, but this technology today are to far far away of "understand" what is a complex bug? or for why its is a bug on certain program context?

But remember the tweet is a joke, don't take too seriously =)

I believe that the differenciation between our "deep understanding" and their "simple training" is an artificial one, based only on the believe in human "special" status.

Of corse neural networks are too shollow and lack cluster referencing system, that would enable them to jump between individual "concepts", but I still firmly believe, that our understanding is based on a same level of repeated training.

PS: I find it funny to take jokes dead seriously. It creates a disonance, that I consider halarious for some reason.

@LukeAlmighty I don't think that humans are "special", my opinion is based on actual state of art of IA/Neural Network/Machine learning.
Repeated training is actually one of many ways that a human used to understand or learn the reality or to solve problems but we are no limited only to it

Sign in to participate in the conversation

Welcome to your niu world ! We are a cute and loving international community O(≧▽≦)O !