Monday, September 5, 2016

Fast and Slow Thinking: Why sophisticated language makes us human.

There are essentially three different systems in the brain.

1. Fast thinking neural nets (emotions), systems that integrate sensory signals and put out a signal that indicates something else. We can't 'see' inside these systems, we can only believe their outputs are appropriate or not.

2. Slow thinking memory that uses language and ideas to make estimates and calculate future possibilities.

3. You. The thing that decides what to think about.

We know almost exactly how systems 1 & 2 work.[1]  We've given out several nobel prizes for these topics. We reproduce them in the lab and we make them better every day. We use artificial versions of them to make our lives easier every day. (Have you used Google search lately?) And they'll continue to get smarter, faster and more reliable.

We also recognize that system 2, which relies on language, is what makes us most diffrerent fromt he rest of the species on this planet.

However, we have little idea of how system 3 works. Maybe that's best. If we don't have a system 3 then we don't have to worry about the systems 1 & 2 becoming 'super intelligent' and taking over the world.  They remain black boxes that can 'think' faster than we can, but have no motivation.

Thanks for reading,
 -Dr. Mike


[1] Thinking, Fast and Slow, Daniel Kahneman, published by Farrar, Straus and Giroux; Reprint edition (April 2, 2013) 
  • ISBN-10: 0374533555
  • ISBN-13: 978-0374533557

SETI - A new proposal to search for extraterrestrial intelligence.

Why don't we see any aliens?

A significant problem is the vastness of space. Despite piggybacking on the world's most sensitive radio telescope, Charles Stuart Bowyer said, the instrument could not detect random radio noise emanating from a civilization like ours, which has been leaking radio and TV signals for less than 100 years. For SERENDIP [A3] and most other SETI projects to detect a signal from an extraterrestrial civilization, the civilization would have to be beaming a powerful signal directly at us. It also means that Earth civilization would only be detectable within a distance of 100 light-years.[89]
"With available instruments we are unlikely to detect Earthlike planets or civilizations," Airieau said. "This sort of detection will not come within our realm for another few decades." c. 1998.

The only thing that overcomes the vastness of space is the vastness of time.  The best way to detect alien civilizations is to send something to where they are, to get closer to the signals, then they'll be easy to detect. This is what we have been doing and will continue to do as long as we can.  We've been sending robots out to our solar system and beyond.  Let's assume this continues for millions of years. What would be the state of the galaxy after those millions of years?

There would be a robot in every star system.

Image result for curiosity rover pictures
We build killer robots, won't everyone else?
What does this imply? This implies that there is an alien robot (actually many alien robots) in our solar system. There are certain fundamental limits on how these robots can communicate. We can make some general observations of how these robots and their communications systems would be designed.

First, we can assume that they will use the most power efficient transmission systems possible. There's only a certain amount of energy in the universe, so the aliens would like to use it as efficiently as possible.  We then know how they would choose to send electromagnetic signals by some calculations that  Professor Emeritus David Messerschmitt has made in [1].  This paper assume s that the only signals that go between star systems are electromagnetic.  In our case we can relax that assumption and assume that the signal receiving civilization actually designed the signal sending apparatus (the robot) and thus has control over both ends of the system.  Professor Messerschmitt looks at the consequences of this in a newer paper [2].  He doesn't quite get all the way to robot to robot communication (or as he likes to call it, starship to starship communication) but, we can use the formulation in the paper to make that estimate.

Second, we can assume that they will try to have their robots remain undetected, as if they were detected there's a good chance they could be subverted or destroyed.  This means they are going to be fairly small, but not too small as they have to have enough power to actually get there and stop.  Let's assume that they would make themselves small so they'd be harder to detect. What are the limits on size detection? Right now we can detect 90% of the asteroids that are one kilometer in diameter, but soon. with advanced signal processing techniques, we'll be able to detect asteroids that are 45 meters in diameter. So let's assume these robots could be 10 times smaller than this (why we haven't seen them.) This would limit the robots to about 5 meters in size.  With this limit we can calculate the power that these robots would transmit per bit.

Since we are assuming that the robots will be talking to each other and there's one at every star, then the average distance they will be sending data is about 5-10 light years. The two antenna will be about 3 meters in diameter.  The transmitted power per bit will be 46 watt-hours per bit.  (See [3] Table III in [2], the only difference is the receive antenna will be 100 times smaller than one on a planet.) This then sets the limits on what we need to use to detect to find these robots.

We can use these limits to propose a new Search for Extra-Terrestrial Intelligence.  I like to call this the Search for Alien Killer Robots (SAKR.) These assumptions point to several new methods of conducting SETI.  We can look towards all the closest stars to try and pull out these minimal energy signals from the light of the star. Another idea to detect these robots is to send our own robots out towards the nearest stars then look back towards the solar system.  The signals will be much larger since we are much closer to the transmitting robots. And if any robots are found, we can then go get them.

One proposal would be to send some wide band receivers toward each of the stars and have them listen in the direction of the star as well as in the direction of the solar system. I believe that if we sent enough of these detectors we could also have the largest synthetic aperture radio telescope in (or out) of the solar system. This array of detectors could be used to look at many other possible signals as well as being used as a transmitter to talk to those potential starships we'll be sending out.

In future posts I will go over the costs of running these new SETI searches, how to optimally design the detection systems [4], and what else we could do with such a system if we build it.

Thanks for reading,
 -Dr. Mike


[1] Optimum end-to-end interstellar communication design for power efficiency, David Messerschmitt, UC Berkely, published in astro-ph.IM on 28th July 20132. arXiv:1305.4684v2

[2] Design for minimum energy in starship and interstellar communications, David Messerschmitt, published in astro-ph.IM on 29th March 2014. arXiv:1402.1215v2

[3]
TABLE III
EXAMPLES OF ENERGY REQUIREMENTS AT THE FUNDAMENTAL LIMIT


Parameter
Starship
Civilization
Units
Tx antenna diameter
3
300
meters
Rec antenna diameter
300
300
meters
Distance
10
1000
Light years
Received energy per bit
8
8
photons
Transmitted energy per bit
0.46
0.46
Watt-hours


[4] Insterllar Communication: The Case for Spread Spectrum, David G. Messerschmitt, UC Berkeley, arXiv:1111.0547v2 [asro-ph.IM] 2 Dec 2011.





















The Buddha’s not there.The Illusion of Truth

​ An Ode to Existence in the Prophetic Style. [The Multitheist’s Lament] Buddha looks inside himself and sees the void, the void is nothing....