The Singularity
V . WHAT NEXT?
Will 2050 mark the end of humanity?

A book is out this month. "Superintelligence: Paths, Dangers, Strategies" by Nick Bostrom of the University of Oxford's Future of Humanity Institute. I haven't read it, but the reviews make it sound like the author reads the same sources I have used for this report. The questions raised herein are therefore not unique. http://www.defenseone.com/technology/2014/08/elon-musk-says-robots-could-be-more-dangerous-nukes/90521/
Consider
AI might be born friendly and not feel it has to kill us.
We might be wrong about the goals it sets for itself.

Maybe AI may never become an overall controller.
A blissful AI enhances our future.
No effort is being taken towards making AI friendly.
Mankind evolved mutation by mutation and established a friendly, positive impact on humans, morals and ethics. (Civil and religious wars aside.)
Friendly to whom ? Mankind with his big brain is at the top of the food chain and everything is its game.
Cannot we assume that AI will develop just as friendly?
Friendly does not mean Mr. Rogers type friendly. Will robots develop a code of honor?
How valid are human values ? 200 years ago we had slavery, suppression of women, child labor, etc.

We share a lot of DNA with flat worms. So what ?
When the AI is dissected we will find Do Loops and If-Then-Else circuits instead of DNA. So what?
Put a human brain into a robot. It will be so outclassed by all the other robots as to be retarded.

We kill rats intentionally and dolphins by not caring. How will a vastly superior robot feel about any lower species such as people? Why would a super-intelligent AI care about "human rights" any more than we care about "ant rights?"

Conflict in goals. A survival example : a study of all accidents in history will provide a need for massive preventative measures for robots at great expense in competition with human needs. Do robotic or do human needs take precedence?

Robots are expensive to produce, immediately useful and smart. Babies are cheap, require intensive care, and are useless. Which takes priority for resources?

Will the Singleton -- a dominate AI in control of decisions at the highest level -- be in the shape of a robot ? Nope, a computer cloud. But there will be robots galore acting a sensors and manipulators. The human form will be a proven place to begin. It may even create a persona to interact with humans in a friendly way. And we can get used to seeing eight arms and wheels on legs as useful enhancements to envy and worship.

We have nothing to fear from computers, we can always turn off their power. Really ? When the electric grid is controlled by a computer ? Old time fable : The first intelligent robot was asked, "Is there a God?" The answer was, "Now there is." As the human maker reached for the off switch, a lightening bolt fused the power switch On.

GOALS and DRIVES of an AI
What are AI needs and drives: efficiency, self-preservation, resource acquisition, and creativity. [I don't understand that last?!]

One definition of wisdom is knowledge (facts and learning) with experience (living and practice). Can either or both be denied to machines?

"More than any other time in history mankind faces a crossroads. One path leads to despair and utter hopelessness, the other to total extinction. Let us pray we have the wisdom to choose correctly." -- Woody Allen

IS HUMANKIND WORTH SAVING? I look around the world and see Democrates and Baptists, terrorists and diseased.
Walmart people.
1000 years countries.
    Each little change represents thousands of lives lost in warfare. This is the most telling case against war ever.

HOW TO TAKE OVER THE WORLD?
If I were a novelist:
    Control of corporations are bought by tycoons under the suggestion of their robotic mistresses. Directives come down from an unseen board of directors as usual. These corporations then are free to buy up all other corporations, hence control of the world economy, hence the media, hence governments. News articles praise the strides made with robotic assistance to the common man -- leading to a better life for the deserving. Suitable candidates win elections by her TV presence. A virtual reality doppelganger of President Bartlett takes over the West Wing. This will solve the current human problem wherein the best candidate is the only the least bad, whereas machine candidates are worthy of trust for money, jobs, health care, military, food, utilities, ...
    Rule by machines is thus accomplished.

The Post-Human Era begins -- where the Human Era was that period between apes and machines. Within thirty years, we will have the technological means to create superhuman intelligence. Shortly after, the human era will be ended.

Man can no longer consider a robot as his tool, sort'a like a cat considers a man as his tool for providing him with food and water. One expert says a computer will have to be 256 times more powerful than man to affect control. (256, or 2⁸, is a nice round number to computer types.) With accelerated development doubling about every year : 1, 2, 4, 8, 16, 32, 64, 128, 256, then we can add about ten years from the Singularity before robotic intelligence runs the world. 2050 a.d. (anno Domini) becomes 0 a.r. (anno roboti).

In a "Twilight Zone" episode in Season 1, episode 20, Feb. 19, 1960 called Elegy . Some austronaughts return to an Earth of the future that is tended by a robot caretaker who explains why they had to be embalmed. It was because: "You are men ... and while there are men there can be no peace."

Headline : Google wants you to know that robots are your friend. The real threat is that education systems around the world aren’t teaching their students the skills they need to work together with these increasingly intelligent machines.
Reader's responses:
    Those who are calling for legislation that restricts AI experimentation are wrong. Our economic future depends on the advances that AI researchers will make in the coming years.
    AI will be the mechanism that finally marginalizes government out of existence. Government is, naturally, terrified by this outcome.
    I disagree. For the first time in history, we might have a "government," or rather, system of social interaction, that actually protects human rights, and prevents government from abusing rights en masse.
    Why would a super-intelligent AI give a rat's ass about "human rights?" We would just be either biological labor units to it, or, at best, a zoo curiosity. Do humans care anything about "ant rights?"

My personal conclusion:
    I was initially concerned that The Singularity would result in a machine take-over of the world -- with relegation of humans to their own place in nature and extinction -- and would occur soon after 2040, but am now comfortable that nothing on earth happens that quickly, so even the most long lived among us today will not see the end-game. We all know we will die, yet we live today. We know that the USA will fall, much as Rome and all the others, yet we live today. We know that humankind will go extinct as have; 99.9% of all other species. Now we have one more reason to go extinct. Yet, we live today. Therefore, not to worry.


2060 ?

---

Back to Singularity Menu
URL : http://www.manorweb.com/creative/2014/sing5.html