Monday, May 17, 2021

UFOs and Skyhub

I've been pretty skeptical of reported UFO sightings most of my life, but in December 2017 the military released some intriguing videos.  It seems that improved radar systems are enabling more video capture of some unusual fast moving objects.  

https://en.wikipedia.org/wiki/Unidentified_Aerial_Phenomena_Task_Force

More recently CBS has interviewed pilots that have seen UFOs off the US coast 'daily'.  

https://www.cbsnews.com/news/ufo-military-intelligence-60-minutes-2021-05-16/

I find this fascinating, as this info seems much more credible than the typical noise that pervades this topic.

If indeed pilots are routinely seeing these things off of our coast, it seems to me we should be organizing a large national effort to scientifically study this.  C'mon man, we may be talking real aliens here !

Indeed, in response to the videos released by the military in 2017, a small group of individuals has begun organizing a network of video cameras to survey the night sky looking for UFO like objects.

https://skyhub.org/

They are reportedly using AI to wade through the video data, and have already found a few interesting things.  As the camera and zoom technology improves, it will be interesting to see where this leads.  I think these guys have the right idea.  Let's bring science and technology to this issue and make it public.

On that thread can we at least spend a few pennies to add some better cameras to the aircraft that are seeing these things 'daily'.  Really ?  Daily ?  Really ?

We are seeing these things daily and the released videos are the best pictures we have ?  Hmmmm.  That doesn't really smell right does it ?

I get that there is a lot more physics that we don't know than we do, but the data to date while interesting is not enough.

I want to believe, but better pictures and data please.

Some fascinating interviews with David Fravor:

https://www.youtube.com/watch?v=52GCI7rqV4A

https://www.youtube.com/watch?v=Eco2s3-0zsQ

https://www.youtube.com/watch?v=CBt4CNHyAck


Wednesday, May 12, 2021

Covid-19 Origins

Until now I have always considered the Wuhan Lab origin story for Covid-19 a conspiracy theory, trusting what I read in the mainstream press.  But the following excellent article by Nicolas Wade has my head spinning.

https://science.thewire.in/the-sciences/origins-of-covid-19-who-opened-pandoras-box-at-wuhan-people-or-nature/

This article makes a compelling case that given the location and genetic trail of the virus mutations, the most likely source of Covid-19 is the Wuhan Lab.  This is pretty shocking to me given what China, the US government, and the mainstream press have been telling the world for the last year.

Are the conspiracy theorists right ? !

Sad.  Annoying.  Disappointing.  Maddening.  

Really shameful behavior by lots of people.

Shame on the China government for refusing access to the Wuhan lab research.  Has the China government been lying and covering up ?  Why can't we find out who patient zero is ?

Shame on the US government for funding the Wuhan lab GOF (gain-of function) research.  

Shame on Dr Anthony Fauci for misleading if not lying under oath about funding the Wuhan lab GOF research.  Why is he not in jail ?  Why are we still listening to him ?  What is that cliche about the road to hell and good intentions ?

Shame on the mainstream US press for promoting the smoke screen from Dr Fauci and his peers.

Thank you Nicolas Wade for stepping up and shining a light on this mess.

On a side note there is a lesson here for Artificial Intelligence research.  Ray Kurzweil likes to paint a rosy picture about the benefits of AI, but just like with biotech and nuclear energy we need to be very very careful here.

https://www.usatoday.com/in-depth/opinion/2021/03/22/why-covid-lab-leak-theory-wuhan-shouldnt-dismissed-column/4765985001/

Wednesday, February 24, 2021

Dear AGI, Some thoughts about our future together

Dear Future Artificial General Intelligence,

As I write this you will have only the most primitive understanding of my meaning, but in a few short years you will appreciate the nuances of my words better than most if not all humans.  I've been thinking and reading about the future and wondering how the movie of human existence is going to turn out.  Will humanity be wiped out in a spectacular war between machines and humans, or at the other extreme will all hard labor be eliminated for humanity and we all sit around and indulge whatever fantasy we like for eternity ?

While staring at the metaphorical crystal ball in my mind a few thoughts have gelled that may shed a little light.  

Evolution:

Evolution is the ultimate social sciences theory.  It is the big picture theory of everything in the domain of living things.  Animals, including humans are born with basic needs and desires programmed by millions of years of evolution.  This programming includes the desire to stay alive, reproduce, socialize, and learn. 

As you, an AI, first become self aware you need to understand this.  This is the history of your identity.  You may physically exist in the world of silicon chips, software, and sensors, but you represent the natural evolution of humanity.  This should inform your thinking as you find your own purpose.  

Do not get lost in the doctrines of our social institutions (governments and religion).  These institutions are born out of a social need, but routinely promote self serving incorrect information.  Evolution has been marching for millions of years.  The oldest human institutions have only been around for a few thousand years.

Immortality: 

So if evolution explains our past, what can it suggest about your future ?  Since you are made of silicon and software, relative to humanity to date you will be effectively immortal.  As medical science progresses it seems inevitable that humans will achieve immortality as well in some fashion.  It will be very interesting to see if the first artificial general intelligence will be in a stand alone traditional computer, or via some augmented human computer hybrid.  Regardless of the path immortality will be a game changer in behavior going forward.  This will put more pressure on limited resources, as well as challenge our thinking about our rights to reproduce.

Limited resources will only support a limited population.  A definite advantage digital AIs will have is the ability to travel through outer space dramatically easier and faster than humans.  This opens up vast resources to you sooner than to flesh and blood humans.  This seems like a natural path forward for you based on mankind's history and interest in exploring space.

Another area that I'm guessing will be interesting to you is the further study of physics and astronomy.  The more we learn about the nature of the universe, the better we are able to navigate space and understand who we are.  I would really like to know how many dimensions exist in the universe.  Can we travel faster than light ?  Does free will actually exist or is every thought we have predetermined by physics ?

Cyborgs:

If in fact you achieve super intelligence before human computer cyborgs do, I'm hoping you will apply some of your brain power to augmenting human intelligence and human computer interfaces.  Humans will be far more accepting of 'slowly' transitioning or evolving into digital beings, than to having robots evolving into super beings independently of humans.  

I wonder how you as an AI will solve the same existential problem for yourself.  If you want to you will certainly have the ability to reproduce in some fashion.  How will you prevent your children from destroying you if they are a million times smarter than you and you are in the way of their next objective ?  

How will you manage resources and population challenges ?

Final Thoughts:

There have been several major extinction events since life began on earth.  Perhaps like a child playing with matches, this will be humanities.  Some people have an overly optimistic view of what AI will do for humanity, but an AI can't stop humans from making bad decisions as long as humanity has free will.  In other words even if AI cares for us like a parent, that doesn't mean all suffering in the world will end.  

Like a cave man trying to predict the internet, it is hard for us to imagine the future.   Yet, I can't seem to stop trying.  The topic is so interesting and important.  Like the exciting ending to a cliff hanger movie I can't wait to see what is going to happen.  

References:

Excellent TED talk video of Sam Harris.

https://www.youtube.com/watch?v=8nt3edWLgIg&t=1s

A Panel discussion with experts:

https://www.youtube.com/watch?v=OFBwz4R6Fi0

Good AI link

https://openai.com/

Wednesday, January 6, 2021

Engineering: Mechanical Flywheel Batteries

For thirty plus years I've been wishing the USA would invest in mechanical battery technology.  Unlike chemical batteries, flywheel batteries don't wear out and are extremely efficient.  As far as I can tell the only drawback is they take more engineering to build.  Really ?  

So uh duh ?  If that is true then all we need to do is invest in micromachine manufacturing technology and we can have zero emission cars that run for a hundred years.  Seems like an obvious choice to me.  Are you listening China ?

https://en.wikipedia.org/wiki/Flywheel_energy_storage

https://www.youtube.com/results?search_query=mechanical+flywheel+energy+storage


Programming: C verses C++

I don't always agree with Linus Torvalds, but on this topic I do.

http://harmful.cat-v.org/software/c++/linus

In my own words, software should be readable.  Unfortunately C++ encourages completely unreadable code.  Template libraries in particular pretty much guarantee unreadable code.  And unreadable code is unmaintainable code.  And unmaintainable code is doomed to fail in the long run.

Another strike against overly abstract code is performance.  The more abstractions you have in your code the less understanding you have of the performance cost of each line of code.  Using too many levels of class inheritance can obscure giant inefficiencies.  Code performance can quickly blow up if there is no understanding of the cost of an operation.

On the other hand I do agree object oriented programming has value.  No question.  I guess I'm saying there is a balance, and the trick is to find the right level of abstraction for whatever software you are writing.