Nuke em!

So have you ever wondered what climate “skeptic” blogs would sound like if global warming were strong? If it were, I can tell you this blog would say it loud and clear. The data is the data of course but when you don’t have strong warming, and you cannot find any impacts from the small warming, what are you supposed to write?

You can write about the charlatans, the scammers, and fake news. You can write about electricity, LED’s and oil. So many things that beat around the fact that the subject is busted because……..

there is too little warming.

It’s just not as fun.

So – I reject your reality and substitute my own!! — love that quote.

[surreal]
As you know, we have seen massive increases in CO2 levels in the atmosphere. Sea level increases are accelerating and there is nothing we can do to stop it for the next several decades. Adaptation is in the works but one wonders, what will happen when the next big hurricane hits. It’s not like sea level rise is a linear process without stochastic weather noise prepared to wreck the next city.

China has refused to admit their addiction to CO2, clean coal is the same as dirty coal in this instance, why can’t they build nuclear plants. It isn’t like the greens are stopping them after all. The US has shut down more coal plants than any other country has ever owned outside of China [could be true actually]. Why do we carry the burden of the battle. Brownouts are not our future, we need nuclear power now!!

The UN keeps pushing socialism, but what does socialism have to do with the solution? We all agree that the global warming shitstorm needs to be corrected. We don’t agree that we need a supreme overlord to get it done. Free the energy producers to use their own choices to avoid CO2 producing power. Let them build the future of energy. Stop blocking smart people from safe nuclear power solutions!! E = MC^2 has meaning folks!!!! We need nukes now!!
[/surreal]

ouch. That’s what would happen here if global warming doom were real.

BTW, we will go big time nuclear at some point.
No question about it.
Because physics.

This one isn’t up to us. E = MC^2 – All we control is when and how we do it.

Should be cool when it happens though. Available energy has transformed productivity of society in ways we would never have guessed. Increase that availability with nuclear power, and we will need to duck as to what humanity does next. It isn’t often that people give low cost energy the credit it is due. Economic productivity is so closely tied to energy over the last millennia, it is impossible to ignore.

So I’m still hoping to see a Mars landing and a conversion to nuclear energy in my lifetime. Both seem like long shots, but we shall see.

False Pragmatism in a Neural Network

Neural networks have always been interesting to me.  Every thought, every imagination, every movement you make is driven by a chemical based network of neurons which are structured to continually change as information is fed to the network.   Scientists have discovered mathematical structures which somewhat model these networks with limited effect.  You would think this is a massive failure of making a silicon brain and it might be.  Interestingly though, the failures themselves reveal an understanding of what we are as humans are actually doing in forming thought, opinion and an understanding of our world.

Nerve Cell, Neuron, Brain, Neurons, Nervous System
Credit Pixabay

In the late 90’s, I wrote my first neural network system as a single layer perception based system for character recognition.  I didn’t use libraries, as are available today.  As with climate science, I read papers and figured out what the researchers were doing and started writing simple algorithms to understand what was happening.  In my first case, I modeled data from the alphabet plus various levels of random noise on an 8X16 space on the screen.  Each neuron is composed of synapses which in my case were attached to each pixel of the image.  The synapse was given a weight which during the training process would be modified based on feedback from the training sequence.  A neurons response to the image is calculated in its simplest form as a summation of synapse weight*pixel value.  If it is a high value of total summation, that is a positive correlation and results in a positive feedback to the weighting.  If it is a negative, then the opposite.

The network training process consisted of assigning random weights to the synapse points and feeding perfect images to the neural network.  When the neuron gives a correct answer,yes that is an ‘A’ you weight the synapses attached to each pixel according to a positive influence, when negative do the opposite.   The weights are akin to synapse reactions in biology.  So if the network parts that recognize the character ‘A’ more positively, their neurons are given a reinforcing weighting feedback while the neurons which don’t react to ‘A’ positively but maybe do react to ‘B’ in a positive manner get a negative feedback from the ‘A’ character and the opposite for ‘B’.

There are a lot of mathematical games which can be played inside this world.  For instance,  ‘A’ recognizing neurons can be designed to fight with each other such that the best fit ‘A’ neurons give extra negative feedback to ‘B”s ‘C”s etc..    You can change the weighting system of the neuron to be a non-linear response such that as each synapse responds to input a sudden threshold becomes a soft one or the reverse.  Layers of neurons are often used to provide more sophisticated behaviors but in the end, the positive and negative feedback drive the final result of the network.

So sorry if that is too wonky an explanation but those who are more math inclined may recognize this system as a poor man’s multivariate regression.   I say poor man because it is inherently sloppy.  The data and weightings are constructed by unclear ad-hoc mathematical weightings as opposed to the rigorous methods of something like ‘least-squares’ fit.  For those who are familiar with some of my work, you may recognize this streak in me from posts like this one  where I attempted to determine the approximate series weight per year from Mann08 by working the fit backwards (a very noisy proposition).

Imagine you trained the network that ‘A’ was actually ‘B’ and then later told it the truth.   How much harder is it for that network to recognize the truth than an originally blank network and what will the end state of the truth recognizing network be?  How about if you  simply say ‘A’ is ‘B’ and ‘B’ is ‘A’ and the higher levels of the network reject your new true statement and reinforce the original false statement.  The reinforcement will give the opposite effect of the truth and reinforce the lie.  A complex enough network may even learn to reject any information from the truthful source.  e.g. MSNBC or Breitbart- where do you get your political news?

So why is this interesting?  First, I believe from my reading that the attempts to model neural networks are very much mathematically similar to the function of the animal brain.  Interestingly to me, the primary problem that affects creating larger neural networks is overfitting.  Yup! overfitting stops these giant piles of arrays from properly functioning and causes them to require massive additional amounts of training time and often results in irrational answers based on heavily weighted overfit networks. 

When training more complex networks, the images of my alphabet become more complex.  Perhaps it might be a cat vs dog vs bike vs motorcycle.  How many kinds of cat do you want as positive results and in what size matrix do you train for a certain number of image types.  Invariably, even for a trained network, failures of performance are identified and the failure rate of the network is a standard measure in the field.  That is not a cat, that is a dog but for some reason the network thinks it is a cat.

In the human mind, we are continually in training mode.  If the mind determines it is looking at a cat, the neurons which contributed to that recognition receive a positive feedback from other neurons farther up the chain and synapse weightings are biologically adjusted in a similar fashion to the mathematical version described above. As the weights of our synapses become stronger valued, the result is that they are more fixed in the neuron, patterns are harder to remove.  To make it more complex, the same neuron might be working on a cat recognition as well as on something completely unrelated like for instance a door.  When this same neuron receives data on perhaps a snake, it may partially contribute to the recognition of that snake.  Once the jelly difference engine of our minds creates a positive feedback to the neuron, it may weaken its response to a cat.

All that is central to the meaning of this post.  What happens when you feed bad information to the neurons?   What happens when you feed this network mountains of false information?

Naturally, you might assume that you would reject the bad information and the learning processes wouldn’t generate a positive feedback to the neurons.  In reality, if the network doesn’t always properly reject the bad information, some of it will be accepted as good information and with repetition it will be burned into the mind with stronger and stronger weighting such that even good information is rejected by the jelly brain.  In fact, it would take an overwhelming amount of re-training of that network to correct the problem created by feeding the mind bad information.  It may not even be possible.

So let’s take the example of global warming and how this has played out.  First, there is a kernel of truth in the dogmatic information fed to the public.  CO2 does block/absorb/whatever you wish to call it/ outgoing radiation from the planetary surface.  This WILL create some warming effect. Basic, basic, basic stuff.  From that scientific fact, we have done a lot of observation and learned that there has been little to no warming measured by our instrumentation.  Modeled warming has completely and irrevocably failed to match these observations.  BUT, we are still treated to the dogma of ‘existential threats’ to the point where every single candidate for president in the democrat party is advocating for massive and expensive transfers of power (and money) to the central government to address this situation.

Now I don’t know if the candidates truly believe what they are saying, as it is obvious to properly thoughtful minds that they are the beneficiaries of the ‘solution’ they propose, but large blocks of the public truly believe we are harming the planet with CO2 despite having ZERO evidence of this fact.

So this title is a perfect example — from the properly dystopian titled website —  The New Scientist

Climate change is making the seas rise faster than ever, UN warns

While the UN did say that, the statement about sea level is demonstrably false information.  In my own mind, any neuron reacting positively to this statement has an overwhelming mass of data and information to the contrary which weights against accepting this statement and will likely result in a more positive assertion that sea level rise has NOT changed in 150 years.  In another person, who doesn’t have the skill or willingness to read and accept a graph, this fits with what they’ve been told and just as aggressively as my own mind, will weight that information in a positive fashion. Polar opposite decisions burned with great strength into different neural networks.
So many false articles repeat the same information – Olive Oil Times , Climate Home Report, Carbon Breif

If I were to meet one of these people, and perhaps suggest that their version of the evidence is false.  We would likely discover that they are ignorant of the data and completely unable to accept that their understanding of sea level rise is false.  I’ve had that conversation with a young engineer some years ago.  I was unable to penetrate these burned in responses.  In practice, you will find that not only will common sense and rational discussion of the data turn quickly into rejection of any of the information you could provide but that other information provided by you on different topics will also be rejected.  I was left as the known idiot and crazy person in the conversation instead of actually getting critical thought from this person.

So why is it that the people won’t accept true and verifiable information when it is placed directly in front of them?  From our understanding of basic neural networks, It is because so many neuron’s have been set with such high weighting that the negative feedback loop delivers a strong rejection response through the network.   It may even become emotionally strong if the opposite pattern is set deep enough.  It requires a special mind to reverse such patterns.

The big problem we are having now is that many people have had vast amounts of the same verifiably false information burned into them from early childhood.  The schools pump it out like a fire hose.  Sea levels, hurricanes, coal bad, Antarctic melting, on and on despite the fact that there is no evidence clean coal is hurting anything whatsoever on the planet.  Last month, my son was treated to a book which taught that there are MANY genders and that you can change your gender whenever you want to AND that conservatives are bigots.  Of course my resistance to teaching 13 year olds things that are pragmatically false was met with the same reactions you would expect.  These people immediately went to bigotry and ignored the issue as though I was unqualified to have an opinion on this matter.

Loving it.

The same thing happens across so many issues.  Trump is a Nazi, a bigot, a racist, colluded with Russia.  All of these statements have been proven false by basic data.  However, a person who has educated themselves properly will recognize each of them as false statements put out repeatedly by the propagandists in the news media.  For me, I immediately reject these statements based on the evidence and my neural network receives a rejecting feedback.  Conversely, the bulk of today’s college students who have been inundated with the nonsensical false facts of the schools and media, have a pattern in their minds which will respond in opposite fashion.  Rejecting the facts in favor of the neural weightings they have spent a lifetime developing. The situation is so bad that even pointing out that Nazi’s were actually the socialist party of Germany, with most of the same policies as Bernie Sanders and Trump is virulently fighting against those policies very often gains no traction with their neural networks.  They don’t see that their own violence and authoritarianism is exactly the same thing that happened in Germany.  The facts cannot penetrate the weighting given to those synapses over the 20 years of their young lives and as they get older, few of them will be able to fix the disaster that has infiltrated and corrupted their minds.

It is a disaster too.  These patterns are so broken that the people are barely functional as members of society.   They believe that feelings are as bad as physical hurt.  They believe that free speech should be limited to things that give them positive feedback in their minds.  They believe conservatives are racists and bigots – with no supporting evidence and in the face of so much evidence that authoritarian governments are the greatest evil the planet has ever known.  The situation is severe enough that no amount of data can penetrate this demonstrably false worldview.

How can we change it?

Well there are several ways the pattern can be broken but in our current situation, the public is being fed massive amounts of false information starting in early childhood and continuing through adulthood.  This is an absolute cancer on society which needs to be stopped.   A pragmatic and idealistic approach to education is needed where only facts are taught and opinions on those facts are put to the side.  In a pragmatic education system, the human mind will tend to look critically at information given and in most cases resolve on the correct solution.  In our current situation, the young minds are being programmed with backwards, false information and given a positive feedback when the mind responds to those facts.

Climate brat Gretta Thunburgggry or whatever is a perfect example.  She has a poorly functioning mind to begin with but she has been programmed with false information.   Now that she’s a famous climate protester, that information has received a massive positive feedback in terms of emotion and recognition.  Any intent to teach young Greta the truth will now result in her jelly brain giving a fierce rejection of that truth, in favor of what she has been programmed to regurgitate.  There is almost no hope of her figuring out that she was completely wrong in her opinion even when the Earth fails to end in 20 years – or show any signs whatsoever of some kind of climate problem.  Each time she has been confronted with this type of information, she reacts very aggressively to self-reinforce the demonstrably false conclusions her mind espouses.

No hope.

It is because of this understanding of the structure of the animal mind that I have little hope that the climate doomers of the progressive party will learn they are wrong.  I expect the opposite in fact.  We fill our children with false information from basically improperly programmed adult teacher neural networks.  These ‘teacher’ networks are in serious need of reprogramming on issues like economics, history, math, science and even on human rights, but it goes against everything they get internal and external positive feedback on.   It is so bad that Alexandria Occasio-Cortez can make 10 statements in a row and I will make the opposite decision on every single one and be much better off.  You literally can assume the exact opposite of what Nancy Pelosi says on every single issue and you will be better off.  It’s kind of a fun game actually.  Listen and reverse the statements and you find that they make perfect pragmatic common sense.

Now I know that Nancy doesn’t believe in what she says and is only doing it for the money and power BUT her reverse-programmed constituents don’t!   They hear truth in the opposite world of the progressive socialists while I hear the falsehood.  Opposite feedbacks created from the same statement again.  You wonder why we have such diametrically opposed thought patterns in America, the weighting and feedback of these neural networks is the key.

You can imagine the bifurcation of a society of neural networks that can be created when two opposite reactions are repeatedly created.  A completely bi-polar black and white situation. Knowing how this works, imagine how you could manipulate a society of these networks if you start by creating this positive feedback through something irrational like emotion.   A cynical person can do it completely on purpose – and they do.   Start with feelings, insert false information, add more feelings.  But imagine further, once that false information pattern is established in the mind, how those neural networks will completely fail to self correct and will fail further to become pragmatic fact-based minds.  Almost no amount of fact based reasoning can convince any of these climate doomsayers that we have had ZERO measurable impact on sea level.  Nothing has changed in 150 years.

Of course things aren’t so black and white all the time, but in today’s America where progressives have gone off the deep end of fact based thought the damage seems to be done. It is apparent to me that when a large group of people refuses to accept rational data, we are in very deep trouble as a country and this IS what our education system has created. Bad programming throughout society is analogous to cancer in the human body in that once the seed is planted, it spreads and grows with little additional input.

The economy is of particular concern to me.  Unlike the planet,  the economy is something that governmental policy can truly impact.  Worse yet, few economies operate as badly and with such disastrous life-ending results as a socialist centrally controlled one.  You can point these facts out to the poorly self-titled progressives with examples in every manner possible, including plenty of opposing examples of how capitalism has literally elevated the entire planet from poverty, and the result you often create is create a negative feedback based on their lifetime of false mental conditioning.   Instead of convincing people, deeply weighted neurons will literally reinforce the original demonstrably false decision rather than weaken the pattern.  Few facts can penetrate this pattern of false economic logic in their minds.  For example, as I predicted so often here, the economy boomed with tax cuts.  Our government literally took in more income tax in the first year the income tax cuts were enacted due to more money being made by people.  This indicates we were on the wrong side of the Laffer curve.  The same curve often touted as ‘disproven’ by the socialist poorly self-titled progressives and their fake economists, was just proven to be true in concept if not in shape.  Making that statement to a ‘progressive socialist’, even with so many facts to back it up, results in a lot of nonsensical argument and in most cases these days results in a worsening of the economic neural trash patterns in their synapses.

It is now shown to be true that at some point more tax = less revenue.  So many people simply cannot grasp that concept.  Hannity is one who constantly touts the penny program.  If you raise spending a penny you must tax another penny worth.  This thinking is disastrous in that it doesn’t recognize the complete nonlinear response of the economy to heavy taxation.  Plenty of people don’t understand this concept though and it is so often said that we need to increase tax to pay for X or Y when the opposite becomes the case in a heavily taxed and regulated economy.  As recently demonstrated, we needed to cut tax and regulation to get more money to pay for X or Y.

I’m old enough to know that I don’t have any real answers to solve the issue of poor programming of the animal mind, but hopefully you guys found this interesting because it took me 2 hours to write it 😀

Happy new year folks.