Anyone know Java?

[Black Rose]

I want to turn my forum posts into a chatbot. How would you do that?

 

[scorpiomover]

I want to turn my forum posts into a chatbot. How would you do that?

Make a chatbot.
Do a search of all your posts.
Copy all your posts' contents into a database table.
Feed the chatbot with the contents of your database table.
Voila! You now have a chatbot that says things you would say.

 

[dr froyd]

natural-language processing is not my strong suit but I suppose ideally it would be based on recurrent nets. One interesting algo I've looked at in the past is called "seq 2 seq", typically used for machine translation.

there's a bunch of machine-learning libraries you can try, like pytorch, keras etc.

I do believe implementing something like an RNN from scratch in Java would be pure waste of time as its extremely complicated and already available from aforementioned sources. But then again I don't know what the goal with the bot is, or whether you want to use ML at all.

 

[EndogenousRebel]

You should try out Processing. It's a coding language built over Java that is meant for intuitive artsy people, and has a decent community with plenty of videos. It's biggest selling point is the library of GUI and GUI oriented functions.

From there you can delve deeper and learn what Comp Sci is really about without the bells and whistles.

 

[Black Rose]

It seems I am not smart enough to take a one-man stand approach to this. And because I don't know anyone in real life and school is not an option I will have to take a different approach. I may need to find an online group in virtual reality that know how to program. It will be interactive and in person. They will have tools I can test out.

Now it will be that in VR I will try and make an intelligent avatar. VR is the perfect place because of physics simulations. The avatar will be embodied in this physics. The world it can interact with will be richer than the real world. Testing can be done quickly.

This is how I believe human-level a.i. will come into existence.

 

[Black Rose]

 

[EndogenousRebel]

I would say it's a lack of discipline or motivation if I wanted to be really harsh, but that's just not necessary. If these are the aspect that you find interesting, then own up to that. If you would be truly interested in reading and writing code, then do keep trying to do it. You only live once.

Plenty of people prefer this aspect of comp sci without degrading themselves with invalidating qualifications like "not smart enough". It's simply MORE FUN to MORE PEOPLE. Play testing video games sounds infinitely more fun as a career than telling a computer which ones and which zeros and in what order they have to initiate. This niches have levels of sophistication that are present in just about every craft. It's a matter of being driven and truly interested.

The computer science worker market is super saturated right now with people who got into the industry, NOT because they're particularly smart, but because they believe it will give them long-term job stability. That is not a factor for you it seems, our society/your environment seems to have given you some security that alters your motivations from the norm. Maybe you do want to be more "normal" in which case, I would say, that's not terribly motivating.

 

[Black Rose]

a.i. requires knowing about computers.

I try to think - but overwhelmed, can't think what is first.

am stuck, been stuck 15 years

want to make something, don't know but maybe something...

 

[EndogenousRebel]

a.i. requires knowing about computers.

I try to think - but overwhelmed, can't think what is first.

am stuck, been stuck 15 years

want to make something, don't know but maybe something...

Are you sure this isn't a problem of conscientiousness? Rather, you aren't optimizing for certain deficiencies you might have?


I've started using Project Management Theory whenever I can, it's really enlightening and it changes your whole perspective of what you can accomplish.

 

[dr froyd]

a.i. requires knowing about computers.

I try to think - but overwhelmed, can't think what is first.

am stuck, been stuck 15 years

want to make something, don't know but maybe something...

tbh your problem is that you usually speak in terms of artificial general intelligence (AGI), e.g. you want to have a super-intelligent AI you can interact with socially. If that's your criterion, you will never, ever implement anything because the theory for AGI doesn't even exist. I'm not saying such an AI is an impossibility, or that it's impossible to implement for 1 person, I'm saying that it's a very bad starting point if you are starting from scratch.

what people refer to as "AI" nowadays is a completely different category - it's either statistical learning algorithms or reinforcement-type algorithms that are aimed at solving specific problems, very much like a bread toaster toasts bread and doesn't write poetry.

if you want to get a starting point, lower your criteria drastically and just look up "natural-language processing tutorials" on google

 

[dr froyd]

I've started using Project Management Theory whenever I can, it's really enlightening and it changes your whole perspective of what you can accomplish.

project management is just a way for managers and lazy people to feel like they are doing something useful while accomplishing nothing. Gratuitous gimmicks.

 

[EndogenousRebel]

I've started using Project Management Theory whenever I can, it's really enlightening and it changes your whole perspective of what you can accomplish.

project management is just a way for managers and lazy people to feel like they are doing something useful while accomplishing nothing. Gratuitous gimmicks.

If you're trying to execute a goal, it'd be wise to organize the logistical and metric components.

I don't see how you can deny that. Maybe the graphics and certain methodologies are redundant, but the vocabulary used is essential, otherwise it wouldn't be universal. Stakeholders, scope, deliverables, ect

Maybe you feel like you've accomplished nothing, but assuming you're the one that assessed, planed, and executed, that's a you problem.

 

[dr froyd]

why would you need a vocabulary to do something as basic as dividing a project into components and execute the plan? I dunno it just sounds like a field whose sole purpose is to fill consultant positions and have long, pointless meetings.

 

[EndogenousRebel]

why would you need a vocabulary to do something as basic as dividing a project into components and execute the plan? I dunno it just sounds like a field whose sole purpose is to fill consultant positions and have long, pointless meetings.

Complexity, vastness, collaboration, diversity. A project can easily become overwhelming depending on a lot of factors

The cancer of consulting is prone in just about every field. Operations and project management however are just closer to money, thus of course you're going to have more contention.

Ideally, having a level of project management cuts down the amount of human and capital resources you spend. Cutting down 10% in either of things allows for allocation of these resources elsewhere. Your loss if you don't believe in the craft itself. You'll lose to competition that that do.

 

[ZenRaiden]

Complexity, vastness, collaboration, diversity. A project can easily become overwhelming depending on a lot of factors

That is the job of management. Workflow and eliminating unnecessary constrains.
Lots of bad management is putting unnecessary constraints on things, and not delineating work in ways to prevents people from working efficiently.

Motivating people is matter of policy really, because yes, sometimes people need bit of extra motivation, but having motivated people is matter of hiring and pay and reward system.

Really the goal is to make people not compete between each other, but work as one group for objective. Since this is very hard to do for most companies, that is why smaller groups of people can outperform bigger companies hiring lots of people.

You are better off with 3 people doing all the work with proper workflow than with 30 people or working like little imps on their little tasks micro managing every step they do.
Micro management is just by product of myopic tracking of workflow and management worried with making mistakes.
Thus people who are more focused on not making mistakes are lot less efficient than people focusing on achieving larger objective and goal.
That is why anal retentive analysis of work is really not good long term.
Its not even good at eliminating errors since those happen no matter how much you try to make the process seamless.

People only stop making errors with training and experience.
All this work tracking and telling people what to do and how by management is really just resulting in people having more constraints than they really need.

So if a person has to worry about constraints on the job, be it one or two or three or more, its always +1 thing to worry about at the job, thus making workflow lot less.

 

[Black Rose]

You are better off with 3 people doing all the work with proper workflow than with 30 people or working like little imps on their little tasks micro managing every step they do.

Like I said, I can't be a one-man team. I can't do it by myself. If I had been able to do it by myself I would not be stuck for 15 years. I am a theory guy, I find it hard to implement. And I don't have the resources I personally need to learn given my depression and other things. I am trying really hard.

Practically speaking I may have something "Like" ADD, I try and think about what I need to do but can't figure it out, get really depressed and stop. An example is files. Java has a way of opening editing and saving files. I never figure it out. Because I cannot figure out simple things I cannot make progress. And I have no emotional support.

I never took an official class so I don't know what code is or the history of programming or the big picture. I just wanted to make my program. Now 15 years is a long time but figuring is out by yourself has pitfalls. I looked at books but nothing helped. Programming is not modular where you can plug and play something that works. You need pages of pages of just the basics. It's not a recipe like they say it is a language and fluency requires a good teacher you cannot learn language from a book and understand complex movies/books. Give a person from a foreign country a dictionary they won't learn shit about English.

You cannot learn a language by yourself. With a dictionary, and no native speakers. I know shit about programming. Management schemes won't help. I need school.

 

[ZenRaiden]

I need school.

School will do the same what you already have.
They will tell you to do stuff, which you already know or have to figure out.
You will just sit in class and read stuff, that you already have on the net or available for purchase. You also have online class without the commute.

 

[EndogenousRebel]

Complexity, vastness, collaboration, diversity. A project can easily become overwhelming depending on a lot of factors

That is the job of management. Workflow and eliminating unnecessary constrains.
Lots of bad management is putting unnecessary constraints on things, and not delineating work in ways to prevents people from working efficiently.

Motivating people is matter of policy really, because yes, sometimes people need bit of extra motivation, but having motivated people is matter of hiring and pay and reward system.

Really the goal is to make people not compete between each other, but work as one group for objective. Since this is very hard to do for most companies, that is why smaller groups of people can outperform bigger companies hiring lots of people.

You are better off with 3 people doing all the work with proper workflow than with 30 people or working like little imps on their little tasks micro managing every step they do.
Micro management is just by product of myopic tracking of workflow and management worried with making mistakes.
Thus people who are more focused on not making mistakes are lot less efficient than people focusing on achieving larger objective and goal.
That is why anal retentive analysis of work is really not good long term.
Its not even good at eliminating errors since those happen no matter how much you try to make the process seamless.

People only stop making errors with training and experience.
All this work tracking and telling people what to do and how by management is really just resulting in people having more constraints than they really need.

So if a person has to worry about constraints on the job, be it one or two or three or more, its always +1 thing to worry about at the job, thus making workflow lot less.

Competence in general is hard to find, so this might be moot. But if someone is orchastrating the whole organizational operation, it might do well to limit/restrict what people down the line do. The key is having no contradictions nor unnecessary limitations in the design, but that's not always easy to see.

I don't practice micromanagement. Micro-measuring sure, with compassion. Micro-objectives as well, while avoiding dead-ends please. Errors are inevitable, are only a problem if they keep happening.

Outside of project management there are plenty of collaborative frameworks, of which are all equally contentious with new age productivity gurus. They all have their pros and cons, so if you know them well enough, you should be able to apply a mixture of them depending on the project.