#controversial
#ai #cnn #nn #ML #DL #github #project #coding #intelligence #experiment #thoughtexperiment
⚠️ Warning. this is a bit Controversial! ⚠️
background: who am I?
I am Will, 24 years old at the moment, I am taking a break from work and studying and stuff like that, I have been doing reverse engineering and analysis on malware (reversing and analyzing them, to share the results to other researchers),
this I am now also taking a break from,
and I started doing AI!
so, I am not yet in university/college, I am just doing this because it is fun.
#####
FOR YOU who don't know about AI and stuff:
(like myself)
binary (yes/no, 0/1, -,+) classifiers is just models that predict if something is 1 or 0. For example..
- Is this (...) <- sentence positive(1)?
- if it's not, predict it to be negative(0)
(this is basically boolean, the 0,1, on/off, +,-, it always has this 2 choices: one or the other)
why I make it so obvious is to explain to others who don't know :) that's basically boolean or Binary classifiers, A or B basically. More than that is called something else, I think, as far as I know.
Anyway,
MAIN PROJECTS
#############
I am doing a major project in AI,
think of it like this:
This MAJOR project can be broken down into 2sub-projects:
1) Passive AI (denoted α, or alpha )
2) Passive AI (denoted β, or beta )
To explain what both of them does, lets explain it with a thought experiment (for readability reasons I will kind of not think about the ethics because it's just a thought experiment) (Einstein's thought experiment about falling in a elevator also, skipped the ethical concerns because, falling with an elevator? that's.. Yea.)
Same here, it's only theoretically speaking.
Now, the thought experiment:
Imagine having alpha and beta, 2 AIs respectively,
The user who made these AI's:
powering on α -
WHICH..Passively observes both the user and the system. The keyboard(what he types), the current applications
running(edge,chrome,virtualbox,minecraft,Microsoft store, python, etc.) The user gives Alpha also access to his diary, so it learns about human life (what makes him sad, or happy, what he ate, what he brew for kind of coffee, and so on.)
//After say 1 month, the AI is exported to a model_file "model_alpha".
And is considered done (learned from the user, and is now knowing about the user's habit and the user's system, etc)
powering on β
WHICH..is given access to //model_alpha
(so this will try to autocomplete things the user writes, like if the user coded in python3, a server in TCP, and begin writing "imports" to import network and socket libraries, the ai will ask (gray text,"tab to autocomplete a TCP Socket, because you wrote a TCP server yesterday")
if the user...
0 does not autocomplete, AI_alpha has to be trained more
1 does indeed hit TAB and autocompletes, AI_alpha has done it's job. And so have AI_Beta as well.
NOTE
I will soon request help (that is, it will not be related to the thought experiment, that's just a thought experiment, really) but I Will need help like Data generation, and stuff
(everyone who will help me will get included in a "acknowledgement/contributor" section in my repository on GitHub as thanks)
BUT I am right now only planning all of this, it's still in the 'thinking' phase.
If questions; or if anyone want to early-contribute; feel free! DM/Ping or Mention/ is always welcome, Discord I prefer because that's where I keep all of my major communications.
- is my username here on infosec.exchange, on GitHub, but also online. 
The Kid Should See This 🌈🪐✨
Molly Cantrell-Kraig 🍷