(Warning: kinda long && somewhat of a political rant)

Every time I tell someone I work with AI, the first thing to come out of their mouth is "oh but AI is going to take over the world!"

No.

It was only somewhat recently that it started being able to recognize what was in a picture from over 3 million images, and that too it's not that great at. Honestly people always say "AI is just if-else" ironically, but it isn't really that far from the truth, we just multiply an input by weights and check the output.

It isn't some magical sauce, it's not being born and then exploring a problem, it's just glorified-probability prediction. Even in "unsupervised" learning, the domain set is provided; in "reinforcement learning" which has gotten super popular lately we just have the computer decide which policy is optimal and apply that to an environment. It's a glorified decision tree (and technically tree models like XGBoost outperform neural networks and deep learning on a large number of problems) and it isn't going to "decide" to take over the planet.

Honestly all of this is just born out of Elon Musk fans who take his word as truth and have been led to believe that AI is going to take over the world. There are a billion reasons why it can't! And to top it off this takes away a lot of public attention from VERY concerning ethical issues with AI.

Am I the only one who saw Google Duplex being unveiled and immediately thought "fraud"? Forget phone scammers, if you trained duplex on the mannerisms of, for example, a famous politician's voice, you could impersonate them in an audio clip (or even video clip with deepfakes). Or for example the widespread use of object detection and facial recognition in surveillance systems deployed by DoD. Or the use of AI combined with location tracking and browsing analytics for targeted marketing.

The list of ethics breaches are endless, and I find it super suspicious that those profiting the most off of unethical AI are all too eager to shift public concern to some science fiction Terminator style takeover that, if ever possible, would be a long way out and is not any sort of a priority issue right now.

!rant
Explaining bellman equations to a n00b

*cracks knuckles*

Boy was I happy to see this when I opened devRant up.

So for starters, more group projects are necessary. Many reasons why. To begin with, it allows for more complex programs than getting some input and printing some shit out. It also develops interpersonal skills (I hate people too, but when you go out to look for work you'll be with them, so better get used to it soon). If a platform like GitHub is used, it's easy to track who did what, and see what each person in the group did, so it should be fairly easy to discourage lazy asses.

Beyond that, stop giving us half completed assignments and asking us to fill in a function/method. Yes, it will take longer. But one doesn't learn to program by doing the minimum required work, you've got to crash and burn a lot in order to git gud. So ffs, let us do all the work. We're like AI, we learn through reinforcement learning.

Stop giving us a spec to follow. We'll do plenty of that in the future, right now we need to make mistakes, not be held by the hand all the way. Let us do dumb shit so you can fail us and tell us our code is repulsive, and this other way was better. Explain why. That's how people learn, not by telling us what each function should return, what can and can't be used, etc. And if you can't come up with a scenario in which what you're teaching is useful, then maybe you're not teaching us the right material.

I'll leave it at that for today... But I'll be back 😈

Which is the most promising sector of Artificial Intelligence in future(2025) ?

I am currently studying about 'Machine learning'.

Learn about
-Cyber Security
-Machine Learning (especially Reinforcement Learning & GANs)
-Microcontrollers (ATtiny)

Of course I want to finish the projects I'm currently working on and maybe start a YouTube channel about my projects.

Yes I know, it's quite a lot to do, but I don't know if I will ever have the chance to do all that things in my free time again.

Next level reinforcement learning:

Grab a baseball bat and show that damn machine who's the boss, i.e. reinforce that message by highfiving the said machine in the face with the aforementioned bat.

Reinforcement learning is going to be my end. 😩😩😩☠️

(currently stuck at how to put images as well as a bunch of other -motor- values as input... and exactly what am I getting as output again?)

Pulling my own hair out... Ooooooof

OK so I have this joke its not fine tuned yet but I'm gonna try it anyway, tell me what you think:

If I ever buy a sex robot I would get 2, male and female, that way I can turn them on before I leave for work in the morning and by the time I get home they would both be exponentially better.

!rant

I heard you like bellman equations...

so we put a bellman equation in a bellman equation, so you can maximize while you maximize

Actually build a deep reinforcement learning algorithm. To play games for me and watch, or some other goal. Actually I'm still looking for a fun, interesting, and realistic purpose for this algo :)

Learning to tech to speed up learning.
Using a new cooperative learning technique, AI Lab researchers cut by half the time it took a pair of robot agents to learn to maneuver to opposite sides of a virtual room.

A combination of deep learning and reinforcement learning algorithms are responsible for computers achieving dominance at challenging board games like chess and Go, a growing number of video games, including Ms. Pac-Man, and some card games, including poker. But for all the progress, computers still get stuck the closer a game resembles real life, with hidden information, multiple players, continuous play, and a mix of short and long-term rewards that make computing the optimal move hopelessly complex.

Image: Dong-ki Kim

I've learned more about stochastic by watching my miserable dqn , trying to determinate whether it's actually learning something or not, than in all the math classes I ever visited.
May write an epic about depths of despair next.
Probably qualified to lead humanity into battle against the machines.
Reconsidering life choices.
Decided never to have children.

- Finish "Introduction to algorithms"
- Learn some genetic algorithms
- Get my hands dirty on reinforcement learning
- Learn more about data streaming application (My currently app is still using plain stupid REST to transport image). I don't know, maybe Kafka and RabbitMQ.
- Learn to implement some distributed system prototypes to get fitter at this topic. There must be more than REST for communicating between components.
- Implementing a searching module for my app with elastic search.
- Employ redis at sometime for background tasks.
- Get my handy dirty on some operating system concepts (Interprocess Communication, I am looking at you)
- Take a look at Assembly (I dont want to do much with Assembly, maybe just want to implement one or two programs to know how things work)
- Learn a bit of parallel computing with CUDA to know what the hell Tensorflow is doing with my graphic card.
- Maybe finishing my first research paper
- Pass my electrical engineering exam (I suck at EE)

Current alternatives for OpenAI Gym?

I had a splash of inspiration. I would like to develop a method for analyzing unknown bitstreams of data. The method would involve determining the format of the data by trial and error machine learning algorithms. This would allow determining data types and byte formats and meanings of streams of data. Could be useful in data forensics. I would call the method: heuristic translation machine learning. I am currently developing code that does this. It will be fun to learn about reinforcement algorithms.

Ros melodic in a strictly python 2.7 environment mixes horribly with a PyTorch based RL module... Time to work around with terminal calls from the latter
*sigh*

Anybody used the integration ui from openai’s gym-retro library?