Deep learning AI software, (Neural networking)

[seen_this 0]

Is neural networking and deep learning AI software anywhere in our horizons? 

Its not rocket science, its just a constant algorithm checking likely hoods based on prior data collections. I know a whole community would be okay with sending their bot data off to improve scripting capabilities. 

It was accomplished in 2011 by rsbuddy and rsbts.nt and he made millions when he was only 17. The snellman brothers made even more with their nexus AI software.

[yeeter 520]

lol its not just as simple as sprinkle a pinch of AI into the code and bam done.  Also take a deeper look into the Snellman brothers lmao.  making millions isn't useful when big daddy Jagex rips your asshole apart and takes you for everything you have and more.

[Nuclear Nezz 2051]

Let's put it this way

OpenAI Five(the dota machine learning) was created by a full research lab, which was founded by ELON MUSK.

Originally, the full research team (not just the game portion) had a funding of one billion dollars.

If anyone could use machine learning to play RS, I don't think they'd be here on the forums talking about it.

 

Now if you're talking about maybe just using machine learning on very small aspects (mouse movement, for example) that's something entirely different than playing runescape.

Also, machine learning is probably pretty close to rocket science. To actually know what you're doing on a real scale, you need to know your shit.

[beezdul 56]

On 6/20/2019 at 10:56 AM, Nuclear Nezz said:

Let's put it this way

OpenAI Five(the dota machine learning) was created by a full research lab, which was founded by ELON MUSK.

Originally, the full research team (not just the game portion) had a funding of one billion dollars.

If anyone could use machine learning to play RS, I don't think they'd be here on the forums talking about it.

 

Now if you're talking about maybe just using machine learning on very small aspects (mouse movement, for example) that's something entirely different than playing runescape.

Also, machine learning is probably pretty close to rocket science. To actually know what you're doing on a real scale, you need to know your shit.

I made a bot to do agility by feeding the output of a custom mobilenet model and a low resolution image of the map into an LSTM. It kind of worked, but was pretty bad.

[seen_this 0]

On 6/21/2019 at 5:58 PM, beezdul said:

I made a bot to do agility by feeding the output of a custom mobilenet model and a low resolution image of the map into an LSTM. It kind of worked, but was pretty bad.

This is genius thinking! I encourage you to refine and work out the kinks and you might have a full time paying job on your hands as a revolutionary new type of bot. Just try not to get copyright infringed somehow in court like the Snellman gods did.

They had a full blown machine learning program to make all those unbannable bots in only sometimes a week from start to beta then premium! 20-50$ a pop all day long