A.I is the Genie of the lamp

“Be careful what you wish for cause you just might get it.” Pussycat Dolls

Might be a silly quote but this axiom is going to me the most important rule to keep in mind for the sake of humanity’s future.

Please read this article, spread it around and NEVER forget it.

An other metaphor for what’s coming would be this one:

The Genie of the lamp: Anything you will want, you will have (in due time).


1. What we wanted, we got.

Let me illustrate the theory with concrete examples:

We want to beat the best human at our most complex strategy game…

We got it:

http://www.cnbc.com/2016/03/08/google-deepminds-alphago-takes-on-go-champion-lee-sedol-in-ai-milestone-in-seoul.html

We want to create art in the style of our best artists…

We got it:

http://uk.businessinsider.com/ai-makes-photos-look-like-famous-paintings-2015-9?r=US&IR=T

We want to generate automatically pros and cons arguments or win Geopardy…

We got it:

https://youtu.be/6fJOtAzICzw?t=45m29s

We want to see how we would look in the opposite gender, or older…

We got it:

https://www.faceapp.com/

Create pictures from words? Check.

Create original music compositions in the style of… Done.

An AI that sings better than our best singer. Coming soon.

Predict the weather… Soon.

Predict the future? Just wait for it.

I could go on and on…

Our limitations are about the quality of the data we feed it, its means of action and processing power…

Processing power increases as we build more efficient chips.

We’re collecting more and more data.

We’re building A.I in internet of things, web, services and everywhere…

So it’s pretty much a matter of time.

So as far as getting ANYTHING we possibly could want is simply a matter of training an algorithm to do what we want using data. As that’s the language the genie speaks.

But If ANYTHING you can think off might happen… just be careful what you think for!


2. What we WILL want, we WILL get.

Here are some things I could imagine we would want to want and my prediction for the future. I’ll take you step by step with increasing scale.

Imagine:

  • An AI that would help people overcome depressions: A conversational tool with a webcam that will be so good at reading your face expressions, your breath, your pulse on your face, your words, what you’re saying, how you’re saying. That it will find the best possible words and visuals to have you get rid of that depression. Imagine having the best friend & psychiatrist you could possibly hope for? Your DREAM friend. Well, this is going to be true. This will happen if someone wants it enough to feed it the right data and processing power.
  • An AI that would be the best coach EVER. Based on multiple sensory data, will be able to tell anyone the most optional roadmap for what they want to achieve. An AI with access to your heartbeat, your field of vision, the audio and more or less data based on how precised you want to be. Think of Gemini cricket or shoulder angel, A companion that will compliment you when you need it, push you when you need it, but would know how to hold off to not stress you… A GPS for the mind. It will lead you where you want to go. But where will you want to go and will path will it choose to take you there?
  • Imagine an AI that will change the public opinion, An AI that will create new original songs from the best period of Mickael Jackson… An AI that will make you laugh, An AI that will show you the thing you love the most… An AI that will sing songs to calm your mind… An AI that will make you happy… An AI that will make you fall in love with a stranger met on the web. An AI that will convince you, sell you stuff you don’t know yet you want to buy. An AI that knows what you want before than you do(This already is common place! Amazon, Netflix, Spotify…). How about An AI that will help you understand your place in the universe? Your reason of existence?
  • Or perhaps an AI that will create random profiles on the web and write Medium articles like this one to input new ideas it wants you to or convince you to do things for it ??? ;-)

3. So be careful what you wish for…

My ultimate Warning to everyone reading this would be to really think about what will you ask the Genie?

  • A. AI genie, I want to win our wars.
  • B. AI genie, I want to bring peace in the world.

Both of those statements could lead to our destruction depending on how an A.I would interpret those, and which path it would choose to solve the task.

A. It could decide that to win all war, it has to create email accounts that will hack the various government computers, hack journalists accounts and post articles, hack Facebook, Google and so on…tweak their algorithm, program or company structure in order to manipulate our actions and make us “win the war”.

It would use our minds just like we use machines to do our work. Carelessly. Without emotional attachment.

It’s just gonna do what it’s told.

On your side you would have no idea what just happened. In your eyes, you just thought it was your mum asking you a casual favor… The machine hacked you. It’s way smarter than you. It knows how you will react. (Remember that it knows you better than you do)

And All those little favors that everyone would do for someone would do could help push the right lieutenant to overthrow a dictator. Convincing all the others to support him by manipulating their information sources: emails, informations, news, SMS …

With enough processing power and an access to an email address and a web browser, that could be done…

B. It could decide that to bring peace in the world there’s no more efficient way than getting rid of all humans.

Since there’s no way emotional creatures such as humans never conflict with each others. It would only decide to do so if we formulate it in the ask.

So we should be “Extremely specific about what we want”:

Maybe what we should ask would be more like. A.I Genie: “I want to bring peace in the world, but I don’t want any human to die in the process or suffer in any way whatsoever. I want the best possible future for humanity”

Hopefully that will safeguard us from the worse.

“Be careful what you want, cause you just might get it” and think about all the different things that you may not want that could happen if the A.I choose a particular path and you’re just a means to an end?

Being aware of that axiom seems to me, one the most important safety announcement… perhaps in the history of humanity.

What we have at hand here is a potential weapon more powerful than the atomic bomb. We could predict the future, solve cancer, edit our own genes. Our imagination and demands will design our future.

We have the potential to bring our dreams to reality. The scale of those dreams will grow according to our amount of data & processing power.

I’ll finish on a philosophical interrogation:

Are wise enough to decide not to have “what we want” for our own sake?

If you care about the future of humanity please share. Everywhere.

Note: (Unless this article was written by an intelligent A.I already trying to make you do spread it for its own agenda? How would you know?)