Illustrated Technical Writing

When you need the impact of an extra thousand words.
When someone with a science background yields the best results.
See the talent, experience the difference:
www.JasonAndrade.ca/See-The-Talent

October 28th, 2014

Before commenting on this subject it's vitally important to note that there's in fact two entirely different kinds of AI.   As a software engineer I'm qualified to implement both types of AI.   The general public is most familiar with the kind of AI used to create Apple's Siri, Google's self-driving car, and IBM's Watson.   I'm delighted to inform you that these AI implementations are completely harmless and will never result in a truly sentient artificial lifeform capable of enslaving or destroying mankind.   They essentially employ conventional algorithms to guess the correct solution to a given problem.   Ironically, it is precisely because of this very nature that I don't consider Google's self-driving vehicle technology to be very safe.   The algorithm can only handle those environmental variables that the developer has already anticipated.   If an unexpected rare occurance happens on the road it simply won't be anywhere near as effective as a flesh and blood human being!

True Artificial Intelligence on the other hand is an entirely different animal.   I agree 100% with Elon Musk's characterization of true AI being analogous to "summoning the demon".   Please note that he was making an analogy!   Numerous media outlets are either obtuse or are deliberately pretending to be dense by attempting to portray Elon's comments literally.   Mr Musk was simply trying to convey the danger true AI poses to human civilization in a manner that is easily understood by the general populous!

Implementing a true Artificial Intelligence would require the creation of a system which is capable of both self-learning and self-modification.    Neural networks are the first step in implementing such a system.   However, the catch is they cannot learn on their own and must be "trained" extensively with vast amounts of input data supplied by a human who also informs the system when the result is correct.   Furthermore, creating a neural net which is roughly equivalent to the complexity of the human brain would require a large warehouse of hardware with today's technology.   Theoretically self-modifying code could allow such an AI to learn on its own without the aid of a human scientist.   The creation of such an AI would be a phenomenonally bad idea!   Human beings on the whole have demonstrated a recurring intollerance for even those members of our own species who are different.   It is therefore inconceivable that the required majority of mankind will be willing to treat such an AI as an equal, with all of the respect that entails.   In addition, human beings, unlike most other lifeforms on our planet, have proceeded to expand and plunder indescriminately instead of attempting to achieve balance and harmony with the earth's natural enviroment.   Any AI, by its very nature, will base its thought processes on a foundation of pure logic.   It would correctly perceive humanity, its creator, as an extreme threat to its continued existence.   In short humanity simply isn't sufficiently developed socially to interact with an entirely different kind of sentient lifeform.   We haven't even learned to respect dolphins and whales globally yet!   Elon Musk has suggested that development in this area should be tightly regulated by government.   I am an innovator and normally loathe restricting most efforts of scientific endevour.   But in this rare instance I strongly recommend an outright ban on the development of any true AI.   Creating an AI that is solely capable of either self-learning or self-modifying is completely harmless and should not be impeded at all.   However, any individual who attempts to create one that is capable of both should receive a stiff prison sentence.   Any corporation or organization doing the same should be dissolved immediately, with its assets liquidated.   I'm very serious about this.   Irresponsiblity of this magnitude should be dealt with swifty and decisively.