Is /Lit/ into programming?
How do you see the future of writing? There is a robot today that can write 2000 articles per second and you can't tell that it was not written by a human.
Reuters uses and some other news agencies as well. And there are more coming?
>There is a robot today that can write 2000 articles per second and you can't tell that it was not written by a human.
>And there are more coming?
When a computer will be able to write books, a computer will also be able to code (program). So programmers will be as obsolete as writers. If it comes to that.
>Is /Lit/ into programming?
it's a bit embarrassing to admit on /lit/ but i was (and am) interested a bit in coding as in applied logic. my constant apathy doesn't let me to improve myself though and my math skills are not stellar too... i know some basics though
>How do you see the future of writing? There is a robot today that can write 2000 articles per second and you can't tell that it was not written by a human.
Reuters uses and some other news agencies as well. And there are more coming?
that robot which you linked is a simple compiler, i saw stuff like that in like junior school when a similar program created fairy tales combining pre-set elements, i saw a program which wrote poetry too around the same time, that used random words from a rhyme dictionary, too bad the result was completely random (sometimes pretty though). to write a program which can actually create a piece of art, not a random pattern or a compilation, it should be sapient, and it's still not known if our pc can even be sapient at all even theoretically
I am a coder and i can see a future where only the best programmers will have jobs because they will create software that creates software that bad coders create today.
But i also think that there will be software that will be able to write articles like a human. I don't mean fiction. I mean articles in magazines. They will be able to write them seriously or humourously or which ever way. And then journalists will lose their jobs.
If that's the case, it's far, far beyond what is capable now. There will have to be a complete revolution in the way computers work. Everything will have to made more complex, from the binary system upwards.
Fiction will be written by software soon enough. I am working on a project with some people on applying Propp's theory to one of these robots (also some Aristotelean beginning/middle/end, linearity stuff and we're working on adding some other basic tropes). Not long and there'll be tales written by robots, and eventually there won't be a good way to seperate it from quality literature.
yeah, i recall an edx course vid, a prof there told to his students that expert programs could replace lawyers but they as coders would be safe... i don't think he was 100% right. expert programs are probably easier than those powerful code and article generators though
satanic mills strike again, bring me my arrows of desire
In all likelyhood all labour will eventually be outsourced to robots.
How much do you know about quantum physics? Spacial relativity? Conway's game of life? Dimensional properties? And finally, coding?
Because there is a lot I base my assumptions on. Probably more importantly, do you believe in determinism? It is practically fundamental in this.
code / programs could possibly be generated by computers, likely through a declarative programming language, where a 'coder' would tell the computer what inputs and outputs he wants, and the machine would create the code.
However, it's not really possible for a computer to take over literature and writing. That would be a very very complex undertaking which precludes it from being possible. We take a lot of things for granted because we come hard wired to see the world a certain way...it would be nearly impossible to 'teach a computer' to see the world through human eyes. A computer could definitely output news stories or simple articles, but it could never write a real novel with a believable human protagonist because that requires a human perspective.
This is an area I am personally really interested in as I'm studying both IT and English Literature.
My understanding of these article generators is that they are created to fulfill a very narrow purpose. ie. Some form of structured document that is often quite repetitive/cliched in nature and based on decision-making processes and language usage that tend to lend themselves to being modelled (relative to other article types, other forms).
The article generators take data (especially things like financial/sporting data) and produce 'natural language' corresponding to the trends in the data and the typical types of commentary human journalists write.
This is not actually that remarkable when you think about it. It should be noted that this sort of writing is basically procedural in nature. So-and-so scored the most often, such-and-such stocks went down, etc. And the most complex+useful of these generators can take years to make. It's a good thing to automate, no doubt, but this is not really a big advance in computational creativity.
You might not notice that a 'robot' has written it, but that's mostly because the process of writing these things (and the structure of them) has been so formalised/mechanized already that it makes no difference.
>>In all likelyhood all labour will eventually be outsourced to robots
>Vivre? les serviteurs feront cela pour nous
also i remember a poem of mayakovsky how he envied machines for their efficiency but if it was translated in english i cannot find it -_-
Man, I fucking hate technology. Seriously, what will be the point of living if we don't actually do anything? How can we appreciate things if they were written/made by a robot in 10 seconds?
>What is true AI?
>What are androids?
The future is coming, prepare to get left behind.
>implying it's a utopian viewpoint
>implying suicide and depression rates won't skyrocket
I can't say that I wouldn't personally enjoy it though.
What we make of life will be up to each of us.
Does it really matter how or by whom it was written if it is a good piece of literature?
There is a company today that offers software that can analyze e-mails and say what mood the person was in when it wrote it, is he happy, sad, financial probelms, small shclong, just got rich and so on. I don't have sauce but it's true trust me.
We also don't know what is going on in AI research labs around the world. I am sure they are working on some scary stuff.
I don't think a robot will ever write better the Dostoyevskiy but i think they will be able to write about factual stuff in an entertaining way. It's coming. And then the janotirs will be replaced by robots becasue they will understand what we write
>There is a company today that offers software that can analyze e-mails and say what mood the person was in when it wrote it, is he happy, sad, financial probelms, small shclong, just got rich and so on. I don't have sauce but it's true trust me.
that's not something especially interesting though, it's not ai, it's a simple analysis of the text for known patterns, not that more intellectual (albeit much less precise) than to use a thermometer to check your body temperature
However a robot can use this software to analyze its own text. Based on the analysis the robot can tweak the text till the result is the right one. Tell the robot to write a text in a happy mood and the robot will write a text, analyze it to see if it is happy enough.
we don't know if it's possible or not yet
i dunno if we even know how to understand that we created it (except some indirect signs like creativity) because stuff like turing test is laughable and was beaten for the first time by elisa bot in like 60s or 70s
but in terms of a more 'literary' perspective...
There is plenty of "poetry generation" stuff (which is relatively easy to get something 'good' and 'human' from since poetry can be short, highly structured, etc), plus some amateurish, and some more impressive experiments into the 'generation' of longer texts. Pretty much all of the texts are unable to approximate natural language, or else fail to be engaging / to have a kind of 'thread' of to follow in the language that makes sense.
Generally these experiments are inevitably tied up in what is unique or special about their "process" of writing, rather than the "product", if that makes sense. But that may change with time.
I struggle to see novels being written by robots any time soon. But there is most certainly a small scene for "procedural literature". Which will grow much larger before we see any sort of 'generalised' novel generating robots.
the robot cannot do anything which you didn't teach it to do and to teach it to create an authentic text of its own you need to give it the true ai
or it can simply tweak some pre-set text patterns, it will be that compilation which i spoke before, it possibly can theoretically create simple mediocre cliche books but probably nothing more
Simple particles obey simple rules to create complex things. It thus follows that the human brain is a bunch of simple particles following simple rules. It is possible to create programs that can simulate simple particles following simple rules. It thus follows that it is possible to create programs that can simulate the human brain.
To take things one step further. The entire universe is a whole lot of simple particles following simple rules. It thus follows that it is possible program a simulation of the universe undiscernable from our own.
Now think for a bit of the implications of this.
It seems I forgot a sentance in the beginning stating that all things are the result of simple particles obeying simple rules. Although it was explicitly implied through the rest of the text so it shouldn't have crated all too many misunderstandings.
And also *on the implications of this
It would seem that even pretentious ass-hats fuck up from time to time.
Your point is not too far-fetched, it simply follows the natural scientific approach in the humanities. But this is also exactly the point where I differ from you.
I agree ( of course) that the natural sciences should be in search of general laws. But this process of natural laws doesn't occur in history. Searching for such in laws in the humanities for instance is called historicism (Popperian interpretation).
As far as I'm concerned, causality on a larger scale is too complex to make a science out of. Historical events do in no way seem to follow any discernable laws. It can, however, be argued that the events are the result of much simpler things. In a game of chess, if you have a few pieces in a corner it is very easy to foresee the outcome, yet in a full game it is impossible to tell the outcome from the onset.
What I'm getting at is:
If we scale things down enough the humanities become irrelevant. If we can then look at the grand picture through this microscopic view the causes can become clear. And with study of these things I deem it possible to replicate the events digitally granted there is enough processing power.
At least, that is how I see things.
Novels will still be around because people still like them. We need to stop pretending that normal people have ever read in their leisure time. Although I see personal stories like Confessions of a Mask being more center than pieces like V. I feel like the internet is going to be the main source of humor and irony.
Let me present you with a couple of counter-arguments:
The universe does not adhere to cause-and-effect. Cause-and-effect is a trick that our minds apply to the universe in an attempt to make some sense of it. The universe is unfathomable. A simple argument is that our brain is small compared to the universe (in fact, our brain is just a tiny part of the universe) and it is therefore impossible for any brain to comprehend the universe.
>granted there is enough processing power
If we think of the universe as one very big computer any computer that we build will be severely lacking if we attempt to simulate the universe. Pretty much the same argument as above, the computer will be contained in the universe and it will only be able to do approximations.
Even if you are right and we are able to overcome all of this it will almost certainly not happen in our life time.
Every language that you use is compiled several times. Let's take a web application written in Python, which is interpreted in C, which compiles to x86 assembly, which the assembler translates into "straight binary", which the computer uses to effect change at the level of transistors and logic gates. Code generators are not horribly difficult to write (although if you tried to implement Haskell, well, good luck) -- a talented undergrad can write a mediocre one. But, it's interesting to think of how many people they might have put out of work. What in 1978 could be done only by a team of highly trained experts working in assembly can now be done by a mediocre dude who learned Python a month ago. Although probably this created more jobs...
Okay yeah sure it took a team of experts at one point to write a shitty version of Paint.exe but our computers have also grown in capability immensely. Do you honestly think that "a team of highly trained experts" would EVER be able to write a modern web application in assembly?
And if we cannot even trust our own my why does it matter? Personally, I'd argue that the logic presented by ones own mind is the only thing really worth trusting. Because, what alternatives do we have?
As far as I see it, there are two ways that this could work.
The first being utilizing the fact that the the universe is in its very core simple. The rules are simpler than the result. It follows that it should be possible to generate with less resources than presented in the simulation. Also utilizing the potential of nothingness should save some resources.
The other way this could work would be by fueling this "computer", if we can call it that at this point. With our own universe. I get that this sounds quite bizarre, but I do feel that the singularity at the end of time could suffice. This is, of course, assuming that the universe eventually falls back to its neutral state.
>Even if you are right and we are able to overcome all of this it will almost certainly not happen in our life time.
This I will have to agree with you on. At least to the extent of the technologies I've spoken of thus far. I still believe, and hope, that "true" A.I. will come to life in my time. This would be as imitations of the human mind rather than simulations of such.
>Every language that you use is compiled several times. Let's take a web application written in Python, which is interpreted in C, which compiles to x86 assembly, which the assembler translates into "straight binary", which the computer uses to effect change at the level of transistors and logic gates.
yeah, those are translations to bytecode and to assembly which are not really human readable though
>What in 1978 could be done only by a team of highly trained experts working in assembly can now be done by a mediocre dude who learned Python a month ago.
what exactly? dunno about assembly but i know c and it's not that hard to use it for most of basic tasks you can learn in 1 month of learning python.
and then, to really use python in practice one need to learn its libs or frameworks for the specific task and it will take months on top of basic cs knowledge, especially considering how shitty those probably are
also python is shit, ruby rules, ruby is the most kawaii language existing
>Do you honestly think that "a team of highly trained experts" would EVER be able to write a modern web application in assembly?
it's possible with c, probably possible with assembly, let me google
yeah it's possible
If I wanted to learn coding and to be any sort of proficient at it, what programming languages would you suggest I learn? Also, which would be the easiest to start with before I work my way into the harder ones?
>Python is interpreted into C which is compiled to x86 assembly which is translated into straight binary
Python is most often never "compiled" but interpreted by a python interpreter which is already in binary form. The interpreter "acts on behalf of" the script written.
Also, while theoretically possible, nobody "interprets C".
Also, even if python scripts were translated to C, modern C compilers forgo assembly and generate the binary straight away.
>to really use python in practice one need to learn its libs or frameworks for the specific task
The libs are there for your convenience. Generally speaking you could write your own libs but it would probably take you longer than to learn how to use an existing lib, even if it's rather shitty. That's the point.
>Python is shit, Ruby rules
Enjoy your performance hits.
>And if we cannot even trust our own my why does it matter?
Because it might mean that we will never be able to find these fundamental rules of the universe because we will never be able to comprehend them. That is, if they exist.
>The first being utilizing the fact that the the universe is in its very core simple. The rules are simpler than the result.
We don't know that. The universe could be (near)-infinitely complex at its core. (Quantum physics seem to support this, no?)
>It follows that it should be possible to generate with less resources than presented in the simulation
Even if the premise is true this does not necessarily follow from it. Perhaps the universe is the best possible 'implementation' and it is impossible to 'execute' or 'simulate' the universe any faster.
I am unable to follow the rest of your post.
In the long term all biological life is just a step-ladder for machine intelligence.
In the short term you'll die before computers can actually be creative. The articles they "write" are trivial.
>yeah, you, book lover
>you know how
>being from /lit/ and all
>you love reading?
>well, guess what?
>you'll love this!
>boy have I got a scoop for you!
>you ready for this?
I'm actively working on the destruction of those things!
Thanks anon. I spent my college years on bullshit liberal arts, the only useful subject being English which landed me a job teaching writing at a studio. I should've spent more time learning something useful, like coding.
>Most people are joking or trolling
My personal favorite:
>Pros: runs ridiculously fast.
>Cons: can't handle a webpage more than 640 kilobytes large because it's only compatible with DOS on a 80286 and lacks HIMEM support.
You can do it, obviously, since you can reverse engineer a webapp binary and shebang you have it in Assembly, but for a human the problems caused by using Assembly get immensely complex and would be extremely prone to error. If people frequently discover bugs and errors in applications written in a comparatively simple language like C (a current example being the shellshock vulnerability found in bash), how do you think they would fare supporting the same software written in ASM?
Most of the time people write tiny Assembly programs made to do one thing exceedingly well, and then have another program call them as necessary. It only really matters where performance is at stake.
You should get a CS degree, because you shouldn't hope of programming anything more complex and sophisticated than a console log of "hello world" (unless you're into Web Dev; HTML and CSS is a breeze in that case). That's an exaggeration, but still: If you want to *understand* what you're doing and how it's properly done when you program, everything essentially boils down to Mathematical Logic.
i used to be a /g/ dweller so i can tell you, there are two main schools of thought among those on /g/
some think it's good to start with c because it gives you the very basics how that computer stuff works
others suggest to learn disgusting python because it's simple and popular
as for my favorite ruby... it's amazing but you probably won't understand its amazingness if you start with it as your very first language, it's better to understand a bit of oop and functions before. generally it's like python (but more fun, possibly slightly more slow, generally it's kind of holy war where i'm on ruby's side, actually python is good and significantly more popular too, but its users are often very annoying), learn python with think python and then ruby with the book with two drawn foxes :3
⇒How much do you know about quantum physics? Spacial relativity? Conway's game of life? Dimensional properties? And finally, coding?
More than you.
Why are 14 year old who recently read GEB so obnoxious? Go back to /sci/, kid.
I kind of wanted to be a writer but I didn't have the courage to go through with it. I became a programmer instead. I envy you for actually following your dream and I am sad to hear that it was unsatisfactory.
All programming languages have two 'users': Programmers and computers. Programmers write programs (production) and computers execute them (consumption). Ruby and Smalltalk are designed for programmers and it tries to make programming easier or more enjoyable. Computers are secondary to their design (programs execute slower compared to other more computer-centric languages, but this is not a big deal with modern machines.) That being said, language design hasn't really made any breakthroughs in many, many years so a large percentage of people who start programming will have a tough time really coming to grips with programming (I believe the anecdotal number is something like 80%.) Also, object orientation is not an easy place to start so perhaps just get the basics of 'if' and loops down first. But I believe anybody can learn to program, some just have to spend more time on it. Personally, I have had to spend many years on it before I finally reached a point where people told me I could actually make businesses money with my skills.
⇒The first being utilizing the fact that the the universe is in its very core simple. The rules are simpler than the result
Perhaps to a christfag like you the rules are simple because the only rule you use it "lol god did it, it's magic". I prefer to use science and in science the rules of the universe are far away from being simple.
Not really necessary but it will help you in job interviews. This is coming from someone who is currently finishing a CS degree. It also has no bearing on the complexity of the stuff that you work with. In CS programming is looked down upon.
My kind of guy!
It has too many language quirks in my opinion.
This will most likely give you bad habits that you will have to shake later. The lessons learned from C are not really applicable in (most) businesses.
I didn't mean to imply that I had a lack of satisfaction with English, which isn't actually a dream. My dream was actually to become a musician but it's not the kind of thing you can make a living off without some dumb luck.
I loved every minute of English actually and love talking lit with kids at a company who actually semi give a shit about the work and books I assign them. It's just that it's not paying the bills so I want to increase my skills and versatility with something like coding, something I consider another kind of writing. I'll keep looking further into it and hope I can get it to click after further study. I still in the beginning stages where I'm deciding on the direction.
Glad to know you're happy with your level of skill anon.
⇒The lessons learned from C are not really applicable in (most) businesses.
well c has more similar syntax
anyway i don't suggest something
also i saw code of people who learned ruby as their first language and they... they used loops in ruby, it was such a blasphemy i couldn't watch -_-
also mit has a good cs course available online, they teach it with python
Ah, I see. I am considering going into teaching myself. I think I can find some joy in teaching programming to others. Other people have said that I might have some 'need' to communicate ideas and I have been unable to put it any better myself, so I think they might be on to something.
I will encourage you to find some small project that you can work on while learning to keep your motivation up. Things will go sour but if there's some carrot at the end of the stick it is usually easier to pick things up again. When I first started out I used to make small scripts that would collect beer prices from websites and find the cheapest beer available in my area. Just silly stuff like that. It could also be a small website where you can post messages. It doesn't really matter as long as it's something that you think you will find useful and that you think is within your capabilities. You might want to downgrade to a simpler project or upgrade to a more complex project once you are more aware of what you are capable of.
Go for it. There are few feelings more satisfying seeing one of your students progress so much that you have nothing left to teach. I'm sure my ability to convey information in a logical and easy to comprehend manner will be an essential workplace skill anywhere I go.
Before I do decide to leave my job though, I can think of a ton of computer related projects at my current workplace that I could work on while while learning coding. Maybe making a more efficient website for the hw for one thing.
>they used loops in ruby, it was such a blasphemy i couldn't watch -_-
Haha, I did that as well when I was first learning Ruby. It was only later that I learned of the glorious collection lambdas (or 'the collection protocol' as Kent Beck calls it in Smalltalk.)