Sunday, January 18, 2015 ... Français/Deutsch/Español/Česky/Japanese/Related posts from blogosphere

Freedom: the difference between brains and machines

Elon Musk and Stephen Hawking teamed up to fight the rise of evil, artificially intelligent machines that are about to conquer the world. At least, the world revolution seems imminent according to the PayPal-Tesla-SpaceX entrepreneur who has paid $10 million to fight the threat in order to make his words louder.

The annual Edge.org question was "What do you think about machines that think?". Some of the answers were nontrivial and interesting. In fact, I actually liked the answer by James O'Donnell, a classical scholar, who said that "no one would ask a thinking machine what it or he or she thinks about machines that think", and he actually warned about the sloppy diverse ways in which the verb "think" is being used.



Octopuses from the 2nd floor (full), "the blue male one" and "the green female one", also known as Formeláks, were famous Czechoslovak 1986 examples of artificial intelligence. Yup, the mother would marry Václav Havel a decade later but when the movie was shot, she was as loyal to the communist regime as you can get.

Many people just stated the obvious – that brains and machines are analogous to some extent. Sure, they are. Every science-fiction-fed kid can write stories about that. But there are actually big differences between "what we call brains" and "what we call machines" as well and these differences are totally crucial for a qualified attitude to the question whether artificial intelligence is about to threaten us.




Sean Carroll who wrote one of the superficial answers (machines and brains are the same thing) opened a debate and I think that the first contribution (but not necessarily just this one) to the discussion is rather wise:

David Kerlick says:

The difference to me is between evolution and engineering design. Evolution results from a series of micro adaptations to circumstance layered upon each other. This results in stable, redundant structures, e.g in bird or insect flight. Engineering design is mostly one-pointed, to solve one specific problem, e.g. build an aircraft. So it is with brains that have “hidden potentials” which are more like unused layers that are not presently active.
Although Intelligent Design advocates might disagree, the biological life species weren't created according to a predetermined master plan. They evolved by constant adaptation to the environment, mutations, and natural selections. This is very different from the way how airplanes, computers, and even computer programs are being created. Those are designed with a predetermined goal or "class of skills" in mind.

And the very existence of the "central plan" is what steals the creativity and other "human" virtues from those machines!




Animals and humans have lots of hardwired rules telling them what they can do, what they can try, what they probably shouldn't try. Lots of these rules have been incorporated to DNA by millions of years of evolution. In other words, species evolve. Additional rules are only being programmed into the individual brains during the individual lives. In other words, people learn. ;-)

This whole process is a form of adaptation. Did the DNA of a species mutate? Yes? Well, that's a risky thing. Did it mutate in a way that threatens the life of virtually everyone with the mutation, given the circumstances on the market (in the environment)? Yes? That's too bad. The species with that mutation should rather go extinct.

Did an individual do something dangerous? Did it kill him? Too bad. Did it hurt him or kill someone else? These are lessons. One should learn from these mistakes. It's less likely that the same maneuver will be repeated. Species and people learn. They adapt. What will exactly happen is hard to predict from the beginning, especially if the outcome depends on many partial questions and events whose circumstances may evolve by themselves. The design of machines is different. One assumes fixed conditions and wants the machine to achieve fixed goals.

Now, you might object that the purpose of artificial intelligence is to create a machine that doesn't have a fixed goal – that can do similar things as the human brain. Perhaps, it may do so in a much better way. And artificial intelligence may have the "hidden potential" much like the human brain – because its architecture or philosophy is similar – and it may be equally (or more) able to learn and adapt.

All of these comments suggesting that "the machine and the brain may be the same thing" are nice. The philosophy of the architecture may be analogous, indeed. One may perhaps computer-simulate a real-world brain, anyway. There aren't any strict physical limitations that would make it impossible. Both brains and computers of any kind are objects in Nature that obey the laws of Nature. However, the point one shouldn't ignore is that to get the artificial machines to a similar situation as the animals or human brains, one actually needs the training – which takes some time and is hard.

Will someone create artificial intelligence that will take over the world by 2020?

Well, I don't think so. The reason is that machines, however sophisticated, are being built with the purpose of serving the people, in one way or another. By design, they are not in charge of things. They don't enjoy the freedom. And freedom is what matters here. Freedom, along with a sufficient time of enjoying it, is needed for things like human brains, with their hidden potential and redundancy, to evolve.

You might also object that someone may create machines for a different goal than for them to serve humans. He just wants to create a Frankenstein. Or even AI Franken. Or AI Gore. (Those names are spelled "a-i", not "a-l".) That's an ambitious goal but the person who has this goal – who wants the machines to get the independence – isn't the only one in the world. There are others. If it turns out that the new machine is designed with the expectation that it will harm the people at least as often as it will help, or if there even is some available experience showing that the machine has been harmful, people will just veto the project. CIA will liquidate the Frankenstein laboratory. Or it will destroy the Frankenstein when it's completed.

My point is that machines are subject to external pressures as well – they may be liquidated, stopped, or go extinct. And the pressure they are facing is intense. Moreover, there is some backreaction. If you think about a future world where the AI machines are very important, such a world has different optimum strategies for adaptation.

Imagine that someone creates a gadget that has some artificial intelligence and wants to become a leader of a country, or something like that, using either other artificial devices and/or humans to achieve its goals. Maybe the evil machine is a network of some sort. Will it happen? I don't think so. Most people (or companies) just won't allow their machines to be connected to this prospective AI dictator. Why would they? They will recognize it as an enemy if that status is clear enough.

We are willing to use an app or a device because we know pretty clearly what it will do and that it will be OK for us. With internet banking, we may check how much money we have in our accounts or make payments. That's helpful which is why we install the app on our phones, or things like that. But would you connect to a great new AI app that can make payments to others and collect them, acting in a human way and enjoying the freedom?

Well, if I don't tell you anything else, the answer is almost certainly No. You won't allow an app to empty your account. The app's being "artificially intelligent" doesn't increase the probability that you will agree. In fact, it will probably reduce the probability that it's a scam that can fool you, so its being "artificially intelligent" will reduce the probability that you will agree with the spreading of this app.

If it's an app that can make some profit for you and it will be verified that it seems to work, many people will agree. But most of them will also realize that there are risks. They will demand the option to leave the program when things go bad. I could continue... But you can see that in all these decisions, people still remain "in charge". They are the ultimate decision makers. Even if you think about viruses that install without the permission of the PC/phone owners, they still serve some (evil) humans at the end.

If artificial machines are supposed to become the bosses, they will have to go through the same process of gradually adapting to the environment and changing it in their own way – and re-adapting to the new, changed environment again. A long enough phase of their peaceful coexistence with the humans would almost certainly be needed. Freedom for the machines is needed for such an evolution. They are getting almost no freedom – much less freedom than what radical Muslims or hungry animals are getting. So I don't think it's really possible for the "AI machines rule" revolution to take place anytime soon.

The lack of freedom really means that the machines don't know what they want.

Animals and plants and species want to spread their genes. But why? And what does the verb "want" really physically mean? Well, there simply exist (composite) processes in which the number of copies of a DNA is increasing. It's an exponentially growing instability of the system. It's not really one fixed instability – there are many instabilities and Nature is gradually switching to new ones as the relevant DNA codes are evolving. The laws of Nature imply that the exponential growth is faster under some conditions – so it simply happens.

The basic "instinct" dictating the life forms that they should try to find ways how to accelerate their reproduction has been around since the beginning of the biological life. The hardwired arrangements in the organisms and their DNA became correlated with the organisms' faster reproduction rate etc. – simply because those arrangements that were correlated reproduced more quickly and got most of the resources etc. The verb "want" simply means to have a preferred goal because of some "internal hardwiring" and this hardwiring occurred due to some evolution, training, and learning etc. We understand why most animals or people want certain simple things.

These algorithms were constantly adapted as the environment was changing and the relevant DNA codes were changing. This "instinct" is everywhere. On the other hand, this "instinct" wasn't inherited by the artificial machines. They started from scratch.

Their analogous "instinct" is really to serve humans and the description "how to do that" is done in such a narrow-minded, specific, one-goal way that it can't really easily adapt to new circumstances. Those are reasons why I think that Elon Musk has just thrown $10 million for a nonsensical worry. Even if we have all the technology needed for the creation of artificially intelligent machines, they will have to get all the freedom and go through the long path of adaptation to the environment.

Telling them (AI machines) what the rules of the external world are and command them to reproduce – to optimize a quantity encoding "how much they serve themselves" – isn't quite the same thing as the "instinct" of the biological species. For example, the biological species have a lot of "instincts" that were useful for some of their ancestors but they are no longer needed.

Given the humans' de facto monopoly in the control of "life affairs" on the Earth, I think that such "artificially intelligent" machines would have to go through a long enough phase of serving the humans rather flawlessly. If humans allowed them even more freedom, maybe it would be a good idea to do so. But if "artificially intelligent" machines will be so good for us, most of us will probably allow their freedom to grow intentionally. Of course that hypothetically, it may lead to the machines' being more powerful than the humans – sometime in the future. But because you don't know all the details of the questions that will be debated in such a hypothetical future, you shouldn't be able to take sides. Unless you are a prejudiced Luddite who thinks that this outcome is a catastrophe and it is a nearly guaranteed one if any "remotely related" technological advance takes place. But in such a case, you shouldn't have co-founded PayPal or Tesla or SpaceX because these advances do increase the chances and accelerate the progress that will lead to this hypothetical dictatorship by the intelligent machines. But maybe Musk is a Luddite and he just realized what a terrible thing he has done by (co-)founding all these companies which is why he pays $10 million to undo the threats he has caused. ;-)

But if he isn't a Luddite and he realizes that his creation of Tesla shouldn't be blamed for the revolution that will turn some evil AI devices to dictators, then he should admit that any other technological work – even work on AI – shouldn't be blamed for the "evil" things, either. So we simply shouldn't solve a problem before it exists.

And let me tell you one reason why I am not a Luddite in this sense. You know, real progress is a good thing. While I probably don't want my country or the globe be overrun by AI overlords (and surely not by Muslims) in my lifetime, in the long run, higher life forms or intelligent beings are not such a bad idea. We give the proper credit to apes, other mammals, and more primitive animals as parts of the sequence that led to us. And "we" (whatever that word means, but I mostly mean homo sapiens) don't have to be the final stage. If some AI artificial folks gradually make us obsolete and they're better, sometime in 2300 AD, why not? They will hopefully give us the credit we deserve – at least as much as we praise the apes. If you are an AI history textbook writer and you can read this sentence, I demand more credit because the ape wasn't even able to formulate its request in an intelligible way. At any rate, I find the opposition to the rise of AI in the long run to be as counterproductive as the movement of some jellyfish who didn't want to allow mammals to evolve and rise – ever. Why? You know that are you just a slimy, Smolin-like jellyfish, don't you? So some modesty would be appropriate. You were not expected to be the final stage that occupies the Earth up to the year 7.5 billion AD. And the same comments apply not only to jellyfish – the future AI beings may have the moral capital to say the very same things about us!

Stephen Hawking is a slightly different case. Given his new (publicly available) Intel firmware, you might classify Stephen Hawking as an AI machine. So his membership in this anti-AI Luddite movement may be said to be suicidal. :-)

See also Minds and machines, 2013, We don't live in a simulation, 2013.



ALS: Hawking became a role model for Czech celebrities

Because Hawking's name has appeared above, let me add just a sad comment on Stephen Hawking's disease. This incurable disease of the nerves is claimed to harass just 2 people among 100,000. That should be 200 people in Czechia. The number of "Czech celebrities" who are at least as famous as the two men below is surely much lower than 100,000 but it seems that we have at least two new sufferers in this group now.

Stanislav Gross, a former social democratic prime minister (pictured on the highly insincere "I mean it sincerely" billboard above) and probably the most "boyish" prime minister in the world's history became the most well-known Czech who probably has ALS. We've learned about it half a year ago or so. This guy earned tens of millions of dollars in stocks after he left politics (not to mention a controversial real estate he got before that plus some real estate in Florida he bought later) – as he admitted, insider information has helped him but he still claims that what he did was legal even if unethical.

Such a disease beats all the other advantages so I can't find enough cruelty in myself to wish him such a bad thing. Of course that I am sorry for him, despite the tens of millions of dollars he has effectively stolen.

Even more sadly, an important soccer player in Viktoria Pilsen, my hometown's club which is the #1 soccer club in Czechia right now again (and was so most of the time since 2010, sometimes beaten by Sparta Prague), Slovak-born stopper Marian Čišovský, was rumored to suffer from ALS since the early January 2015, too. So terrible, especially because none of the other men are guaranteed to be as (relatively) lucky as Stephen Hawking.

Add to del.icio.us Digg this Add to reddit

snail feedback (140) :


reader br said...

What of free will? Do humans have it? Could you have chosen to write the blog post differently? Can you choose anything, such as what to think about, or is free will an illusion of some sort? I would be interested to hear your views on this.


reader Luboš Motl said...

Dear br, people have free will, and according to the free-will theorem in quantum mechanics, so do elementary particles and every other object before it's measured.


Some primitive forms of free-will are too different from the human free will because the intelligence and experience and memory we associate with the humans is missing, but the essence of the free will is always there.


But in this blog post, I didn't want to discuss the philosophical undefinable spiritual things about the free will at all. By freedom, I meant literally the freedom in the political sense - the non-existence of others (or other rules) that dictate a subject's decisions. A computer or computer program written to achieve a certain goal doesn't enjoy the freedom in this sense, whether or not you say that it has "free will".


reader Martin said...

Luboš, you as an expert on QM, do you think that quantum processes must be in some yet unknown way involved in creating phenomena like consciousness, intelligence and experience? I mean if the brain is just a "classical" computer (regardless of the kind of internal architecture, e.g. serial CPU or massively parallel or some other form), then it seems there could be nothing of this sort (or free will for that matter) because brains would be just "clockworks", albeit very complicated, and every person's actions calculable. It would be interesting to know your opinion. Thanks.


reader Peter F. said...

Free will, such as had by fundamental particles, disappears in large lumps of matter such as not least humans animals who enter into department stores! Just watch this Darren Brown in this video and then you, Lumo, may well find that your agnostic attitude to the undefinable form of "free will" will wilt. ;-)


reader cynholt said...

Robots have replaced blue collar jobs, but the untold story is how software has replaced white collar jobs. Business has been spending billions of dollars annually on systems for forty years to reduce white collar labor costs. These are middle management and higher-skilled specialist jobs.

A well-run business today has a handful of highly-paid executives and a software system that manages the low-paid workers that haven't been replaced by robots yet.

Someday, even the Fed and the economists it funds will be replaced by a program.

Karl Marx and Ned Ludd were early, not wrong.


reader NikFromNYC said...

Computers are little more than glorified pocket calculators. Adding orders is magnitude more memory and speed and you won't even notice any qualitative difference in kind of output for a given human input. The brain of even a thoughtless mouse is so terribly complex in design that barely even a clue or two exist as to how brains really work for base instinctual behavior let alone higher human awareness that represents yet another level entirely of utter mystery. A single lowly protein molecule cannot even be simulated properly from its known linear sequence to its dynamic 3D mechanical form let alone even a single neuron of hundreds of billions of them, all hyper connected in myriad ways. I strongly note the lack of equations in your own presentation. Mathematics itself may be hiding vastly higher level orders that are so holographic in complexity that digital computers now reaching the end of Moore's law will never emulate. Only modified biology may access actual "artificial" intelligence and ethics will hinder this greatly, some brave new world of designer human brains in a nutrient aquarium.

"Having children really changes your view on these things. We're born, we live for a brief instant, and we die. It's been happening for a long time. Technology is not changing it much if at all." - Steve Jobs, founder of Apple Computer (Wired magazine, 1996)


reader BobSykes said...

Hubert Dreyfus' book "What Computers Still Can't Do" is now 40 years old, but it is still an important analysis of the persistent incapacity of computers. Dreyfus worked in AI, but his analysis is philosophical.


Real AI would be an autonomous robot moving through the world and acting for its own purposes. We are nowhere near that. Our current supercomputers lack the capabilities of a bacterium or virus. They solve stuff by brute force a la Watson or do clever data base searches a la Goolgle.


Will we ever have true AI? Not until we figure out how our brains actually work. The inanity and deceptions of modern neuroscience is well-described by Raymond Tallis in "Aping Mankind."


reader NikFromNYC said...

Appreciate that Obama's whitewash of all things climate corrupt has dominated the stage ever since 2009 Climategate. Obamacare lies alone will likely put an abrupt end to the Enron level fraud, come 2016.


reader cynholt said...

As you suggest, Nik, learning machines and evolved machines are two very different things. That's largely because to evolve a machine to mimic humans is much, much harder to do than to just engineer a machine to mimic humans. Engineering involves understanding, evolution doesn’t.

In fact, the longer it takes for something to evolve, the harder it is to understand, which may well lie behind Moravec’s paradox (see link below). Of course, for us as humans, it’s hard to admit we may not be able to understand something -- even though all of us on a daily basis use things we don’t understand at all, and likely no single person will ever understand them in their entirety, but at least we have a warm knowledge that potentially someone might be able to in the near to distant future.

Of course, the other drawback is that evolution takes time, so trying to evolve a machine to mimic a human (MTMH) is not something that will ever happen overnight. But saying it’s fundamentally impossible is way, way too strong a claim. If nothing else, it begs a definition of machine. Will we accept only silicon as machine? Or, will other compounds, up to and including organic, be prohibited?

Ultimately, the only argument as to why we shouldn’t be able to evolve MTMH is a religious one, i.e., there’s something out there that’s much more than just matter and energy, which makes it fundamentally unknowable/un-manipulable by us. And then it comes down to an opinion/belief.

http://en.wikipedia.org/wiki/Moravec%27s_paradox


reader Guest said...

Well, even computers are subject to evolution, which leads to some redundancy. See e.g. http://blogs.msdn.com/b/larryosterman/archive/2010/01/04/what-s-up-with-the-beep-driver-in-windows-7.aspx Another example: current Intel's 64 bit CPU supports 5 modes of memory management, 4 of which are legacy. So computers are becoming similar kludge as brain which has also evolutionary older circuits having similar function as the newer ones. Will this process lead to AI? I think it will rather lead to pile of unmaintainable garbage :)


On the other hand, AI as I know it is basically form of optimization of some objective function. And human brain works in similar fashion. Optimization of objective function (feeding, reproduction, problem solving, exploration, etc.) is rewarded by circuit formed by Nucleus accumbens and Ventral tegmental area. So I guess if one could define proper objective functions and more importantly a general and efficient method how to optimize them, then some "evil" autonomous AI could be possible. But given that most popular method today is gradient descent, we have a long way…


And also there's a risk agent would find a way to cheat the predefined objective function - the same way humans do with cocaine (then they don't have to optimize the objective function, instead they get the reward directly, in case of cocaine by blocking dopamine reuptake mechanism).


reader cynholt said...

There is a pattern emerging with this President, Nik. He puts forth an idea he wants enacted, either by executive action, memo, or congress enacting the legislation. The reasons and promises he makes are for the purpose of the action to be palatable to the voter. It sounds plausible the way he describes it; however, it rarely is what he says it to be. The informed citizen can see through this. The low information voter never pays attention long enough to see through his deception. For example, "You can keep your doctor," when he knew you couldn't. Another example, the immigration push he recently made, he said the undocumented would not be eligible for gov't subsidy programs for 5 years. That made it more palatable to the public. However, the truth comes out. Those undocumented are now eligible immediately. So, the pattern is to tell the public something that could make it acceptable, then change it, and then deny you ever said anything differently before. See for yourself if what I say is not true.


reader br said...

I wrote a neural network in software recently, for fun, and successfully trained it to recognise lines of different slope. However, one thing that struck me was that *I* had to train *it* by letting it know what the right and wrong answers were, then work out how to modify the connection strengths. So how does nature know what the right and wrong answers are? And which synapses to strengthen or weaken as a function of that answer? Obviously there has been evolution where 'dying' equals the wrong answer, but it plausibly seems that consciousness itself could play a role here, for higher organisms. For example, after a small bit of maths training, one can 'see' the right answer and then teach oneself new things that one hasn't been told. Of course, one can make mistakes, but then some people are 'brighter' than others, so can see more clearly. The important thing is that this gives an active role to consciousness, which then implies some connection to physics as it would make a difference to the physical world. Hence consciousness would be an adaptive advantage, rather than some passive side effect.


reader br said...

'there's something out there that's much more than just matter and energy'
this is a physics question if that something can actually do something and make a difference. Sean Carroll argues that it doesn't exist because current physics equations have no room in them, and the equations describe things that happen very accurately, etc, but I still have a suspicion it is not the end of the matter. I don't want to give up on free will yet, (though I don't know what that 'I' is yet either) but if 'I' can do something then ultimately that is an experimental question.


reader Graham Green said...

So long as the AI just gets on with thinking were OK. When they start believing stuff we're toast.


reader Luboš Motl said...

Dear Martin, I am convinced it is right to conclude that brains cannot be quantum computers - they are classical (non-quantum) computers, although based on a non-von-Neumann architecture - simply because the high enough temperature in the brain doesn't allow the quantum coherence and entanglement to last for nontrivial amounts of time. Therefore, the thermal noise is enough to measure the brain all the time, making it classical. It doesn't mean that the resulting apparatus is fully deterministic. The brain may contain "random generators" but they're analogous to random generators in classical computers.


reader Ann said...

What is the likliehood of 'enhanced human intelligence", i.e., new biomedical innovations that augmnt our senses or cognitive capabilities? We have repairing technologies now, eyeglasses to correct poor vision, hearing aids, hip, knee and tooth replacements. Can we envision people seeing and hearing and smelling much better through future bio 'apps' or implants? Could ordinary people be able to think faster, learn subjects more quickly, reason through more complex problems? We'd still have our free will and motives, presumably.


reader Luboš Motl said...

Good points, Ann! There seems to be a significant "obstacle" in connecting the normal biological nerves to external machines.


Except that the successful transplants etc. often impress me and convince me that it can't be hard. The nerves can't differ from a wire in the USB cable too much, can they? ;-)


So I would bet that the much more general extensions of humans will arrive soon, within a decade. With such extensions, even if they allow us to think, we will still enjoy the free will.


Like if you suddenly have more powerful muscles or you wake up and you can do calculations more efficiently, you could connect to those extension-electronic-organs, and enjoy your free-will control over them just like you control the arms. I think that the "conscious" feelings would be qualitatively the same - the difference is just in technicalities.


One could perhaps transfer the "core" of the brain activity to some artificial organs as well. Just imagine that this is done gradually, by replacing individual neurons by their electronic replacements, or something like that. I think that the same "consciousness" should appear in the fully electronic version of the human at the end.


reader lukelea said...

I think the missing ingredients here might be pleasure and pain, including fear, lust, and hunger. Without these passions (as David Hume called them) intelligence (or reason) has nothing to work on. With them even flies can think -- e.g., someone once accused me of having "the IQ of a bee" because I trusted him not to gas up his personal vehicle on my credit card account that I used for the commercial vehicles in my business. In other words human intelligence, which I think is what we are talking about here, not only involves the ability to feel but also an intuitive understanding of what others feel or will feel in certain situations (the so-called theory of mind). None of this is very scientific. Which means that intelligence itself may not be a scientific concept even though IQ (which is one part of human intelligence at least) is measurable.

Anyway those are my perhaps unintelligent thoughts on the subject.


reader Gordon said...

A problem of this discussion is that you are assuming that AI needs to mimic human intelligence and the human brain. That view is completely unnecessarily anthropomorphic. Also the assumption that we are hardwired is only partly true...we have highly parallel neural nets with redundancy and plasticity, with the ability to modify the structure of those nets, create new ones to a limited degree, and also to augment them artificially.
Also, machines can evolve---machines can be programmed with evolutionary or genetic algorithms, random mutations or directed mutations, to produce things like winning checkers programs with only very basic pre-programming. Blondie 24 is a good example. It is a program to play checkers that was programmed with evolutionary algorithms that gave it the rules and little else. Then it was entered on an online P2P checkers playing site with humans as something like "Thor" as its nick.
Playing badly as "Thor" (or whatever, I forget) it learned but obviously made strikingly dumb moves and was flamed relentlessly, so the creator changed its cover story and renamed it Blondie 24, a blonde, single grad student at Stanford (? I'm covering my own memory gaps), and the flaming stopped to be replaced by flirting and requests for dates :)

Anyway eventually it learned from its games and reached master level.
I believe the fears are misplaced and similar to the overblown fears about other things. Also, although there are vast amounts of money being spent on understanding the human brain (The Brain Project, etc), AI and a self aware machine does NOT need
to simulate a human brain. The octopus is self aware and intelligent and has 8 or so distributed brains, and, like the squid, communicates using skin chromatophores to flash evanescent changing colors (like the aliens in the movie, Encounters of the 3rd Kind).
Also, we are already forging a symbiotic relation with computers as cyborgs. Cell phones are becoming extrabrain memory dumps and cpus. Implants are controlling seizures and depression etc etc.
Personally, I am excited that silicon evolution(or gallium arsenide) or whatever may enhance our (or their) understanding of the universe with with humans eventually or without.


reader Gordon said...

It is happening much faster than you think, Lubos.


reader Gordon said...

Yup, see my post.


reader Gordon said...

Cynthia---Everyone is getting hung up with AI having to mimic a human brain.
Typical of humans to think this. It is just like the US thinking of US exceptionalism :)
One DOES NOT have to develop a machine to emulate the human brain to create a conscious intelligent machine. One might as well say that until one creates an AI that can pass a squid Turing test, we dont have AI.
Humans! "Only two things are infinite, the universe and human stupidity, and I'm not sure about the former." - Albert Einstein


reader kashyap vasavada said...

Interesting comment Lubos. If high temperatures rule out QM in brains, then how about lasers working at high temperatures? Also are you ruling out possibility of room temperature (organic or inorganic)superconductors?
What am I missing in the argument?


reader Gene Day said...

Eventually can be a very long time, Gordon. Please say more. I am not convinced.


reader John Archer said...

Dear Luboš,

"Hours after this blog post was published, The Daily Mail, published a criticism by David Rose with pretty much the same content. It is totally plausible but not certain that he was able to make the same conclusion independently of this TRF blog post."

Totally plausible, eh? And the Gavin Shit tweet is now two days old. Hmmm.

You're such a nice man, Luboš. :)

However, to be fair to Rose himself, or at least to give him the benefit of the doubt [which I reckon as order zero] as you kindly have done here, if your post were his inspiration and even if he had wanted to give you credit the editorial staff would probably have expunged it, that probability being 100%. You see, the highly informative, original and investigative Daily Mail cannot be seen to have had any assistance from "a mere blogger", not even one who sheds more IQ in his toe nail clippings than they have ever accumulated in their whole lives combined. The rest of the MSM is on a par when it comes to such shitty behaviour. They are all whores.


reader Gene Day said...

No, Cynthia; they were both early and wrong.
Mid-level management is as essential as it has ever been and it will not be computerized. It is a deeply human activity and will always be so.
What kind of a sterile, inhuman environment do you work in? It must be just awful.


reader Luboš Motl said...

Dear Kashyap, a laser beam has a high energy of the photons - which would appear in a high-temperature thermal radiation as well - but its entropy is virtually zero (the beam is completely specified by its direction, frequency, and intensity) and it is completely out of equilibrium, so one can't assign any temperature to it.


It's possible for such a beam to sustain itself because Maxwell's equations are linear - there are no interactions between photons and interactions are needed for a system to converge towards equilibrium.


In the brain, there is so much mess that everything interacts with everything else, thermal equilibrium is achieved almost instantly, and coherence in any phases is instantly lost.


reader John Archer said...

Bliar was exactly the same, and indeed a seminal influence on a whole generation of UK politicians — he took the 'art' to new heights. Cameron is merely one of his heirs.

Some legacy.


reader Gordon said...

Blondie 24 (2002--almost Paleolithic)
http://www.amazon.com/Blondie24-Playing-Kaufmann-Artificial-Intelligence/dp/1558607838/ref=sr_1_7?s=books&ie=UTF8&qid=1421603968&sr=1-7&keywords=david+fogel


reader Gene Day said...

There is no place in the universe that is cold enough for quantum computers except in cryogenic laboratories but I’m not sure that is what Martin is talking about. Does he understand that all processes are quantum processes? If he did he would not speculate on a person’s actions being calculable, at least not exactly.


reader Uncle Al said...

If the human race desires protection against in silico intelligence that 1) thinks a million times faster, 2) has few uses for humans beyond slavery, and 3) might possess a sustained appetite for blood sports - have Microsoft design and write the software.


reader andy said...

Nerves differ from wires a lot. Nerves as wires is one of those baby talk analogies you hate. A nerve has gated channels and receptors along the length of the wire and has active organelles and g-proteins etc..Myelin is not simply insulation - myelin cells are also biologically active and interact with the nerve. The whole set-up is complex beyond belief. We barely understand it.


reader andy said...

Brains are not really like computers at all. A brain is an analog calculator, not a digital one. Nerves fire based on a summation (based on various gated channels and fluxes in membrane potential) of inputs and the strength of the output can vary (think of lifting a pen vs. a barbell - you recruit more fibers to lift a barbell) and how they interact to produce consciousness is a complete mystery. I don't think it's as simple as the number of connections or the "programming". Having a computer be good at a mathematically based game may be considered AI by some, but I don't think so. I certainly may be wrong, but I don't think we'll have true AI anytime soon, if ever.


reader cynholt said...

Business owners and owners of capital don't care about being human, Gene, they only care about being profitable. And if it requires that they cut humans from the workforce in order to increase profits, they'll do so. Without hesitation.

It is commonly predicted that by around 2025 a computer will be built that will have enough processing power to mimic the human brain. While the first one will probably be the size of a small room, it will certainly be followed by a succession of smaller and more capable machines. What will the world be like when a Harvard MBA can be replaced with a cut-rate business computer?

Better, I'd say.


reader Tony said...

Somewhat related - since nothing revolutionary happened in AI field in the last few years, I wonder what prompted these persons to raise such issues in public?

Could it be that they are turning into attention whores?


reader WolfInSheepSkin said...

I agree. I think AI will be mostly useless. The technology is more likely to be focused on machines that do specific tasks.


reader Gordon said...

As an exercise, NASA used evolutionary programming to design a satellite antenna to compare with what they were using. The result looked comical but apparently was superior to their version.


reader Martin said...

Dear Luboš, thank your a lot for your answer. Very interesting. (Gene: I am quite well aware of the fact that QM underlines all processes, including those that occur in transistors and neurons. But what I meant was that the processing of information is "classical".)

You are mentioning a non-von-Neumann architecture. Would you care to elaborate? Or point to a resource that is concerned with the topic?

As for random generators — computers typically use pseudo-random generators and are therefore deterministic. But perhaps the brain uses some quantum process (measurement) to act as a "true" random generator. That could be used to explain why a person's a actions cannot be computed.


reader Tony said...

I would sooner compare a multinational, big corporation to a church than to a lean, mean machine.

However, when it comes to the fast food, I think that characterization: 'made by the machines, for the machines' makes sense, (revealing my nasty, elitist bent) so they may as well cut the unnecessary intermediary.


Incidentally, decades ago, those intermediaries used to be the way that forward looking, middle class young person was initiated into the business world.


reader Bob Tisdale said...

And my post at my blog and at WUWT that made the same point.

https://bobtisdale.wordpress.com/2015/01/17/on-the-biases-caused-by-omissions-in-the-2014-noaa-state-of-the-climate-report/

Cheers


reader scooby said...

Nothing revolutionary as you said, but steady progress in the field of machine learning. After all in the 80s AI was practically considered dead, having failed to achieve all it promised. Now,I'm not so sure. There's been good progress in pattern recognition with recurrent neural networks and deep learning convolutional neural networks, and I find some applications such as Skype Translator nothing short of amazing.


Still a long way to go until Skynet though... True AI by 2020? Not a chance. But by the end of the century, I don't see why not,


reader scooby said...

Isaac Asimov solved this problem a long time ago. It's called the 3 Laws of Robotics :).


reader Alexander Ač said...

Geez, who cares about such abstract thing as global temeperature? Globaly warm or cold, who cares?


reader Gene Day said...

I understand what you are saying, Cynthia, and there is an element of truth in it. Company management more or less has to reduce headcount if it improves competitiveness but that's capitalism.
A friend of mine in Oregon was the COO of Viking Freight (now FedX Ground) when they converted to an owner/operator model. He had to layoff 15,000 people in one day but the alternative was worse. There would have been a huge loss of business to competitors who had already made the switch and the fired drivers would not have had the opportunity to own their own truck and continue driving. I know him well and it was very painful for him personally. Short of abandoning our free market system, what else could he have done?


reader mesocyclone said...

Also, from a scientific standpoint, who cares if it is the warmest year by a tiny, tiny fraction. Is the hypothesis "2014 was the warmest year" an interesting one to science? I think not. It is only interesting to those politicizing the issue.


reader mesocyclone said...

Developments in neural nets have been quite dramatic. Also, neural nets do their reasoning in a way that their programmers do not and probably cannot understand. They are a bit weird.


reader mrbuffalo said...

What if they let the machine understand hunger- say connected with is power supply?


reader John Archer said...

I suppose in your mind-fcuked, religiously greentard milieu that would be considered a right-on, informed and deeply moral sarcastic remark directed at a bunch of gullible and dangerous morons who couldn't give a damn about preventing an imminent world-wide catastrophe.

To me though, you sound like a flat-earther who didn't get much joy from Christopher Columbus as he was about to embark on the Santa Maria and set sail west, his parting shot being: "Go on then, ignore all the experts, ignore anyone with any common sense and sail right over the edge into the abyss. You fool! Eratosthenes was a denier!"

"... in the pay of BIG FRANKINCENSE!"

You're simply laughable.

By the way, I note you present no argument against the post. How typical. Empty of content to the last.


reader Alexander Ač said...

John,

surely, as you say, I am


"mind-fcuked, religiously greentard".


What's more? I like you too ;-)



Alex


reader NikFromNYC said...

Yet there is some limit on things like this that they must all not break mathematics itself, and thus must fit into and be a part of mathematics. The mystery is time. Neither matter nor energy can exist without it. Is time mathematical? I do not know. You don't either, none of you do.


reader Alex said...

and lust by the way you connect the male plug to the female plug


reader John Archer said...

Never mind the pleasantries. Where do you disagree with Luboš's post?

WHERE?

In particular:The tweet by Gavin Schmidt is a simple example of mass manipulation in action. They publish some of the correct yet inconvenient clarifications at places where almost no one reads – the press conference was attended by a small number of people, Schmidt has a few thousand Twitter followers, almost no one reads the bulk of the IPCC reports etc. – while at the places which matter because millions of people read them, they always post the distortions, oversimplifications, and downright lies.

This combination of strategies allows them to say that they "did release the truth". However, they are careful that they only speak the truth when almost no one listens.

That's pretty damning. And your silence on that conclusion damns you.

BTW, do you drop litter in the street too? If not, why do you do it here?


reader mrbuffalo said...

And exchange of a partial algorithm.


reader davideisenstadt said...

When a computer can choose to play speed chess for crack and donuts, like one particularly strong player in Philly, then i will believe that computers are our equals...


reader davideisenstadt said...

and its not an electrical impulse that travels down (or up) a nerve fiber, isn't it an something like an ionic pulse..? that the ions on either side of the membrane reverse temporarily?


reader Alex said...

You need to hardwire in a pain/pleasure centre in it's 'brain'. Most human emotions are connected to that. Something a little beyond autonomous reactions which are just conditional reflexes. A cockroach will clean itself to death when you remove one of it's legs. Its nerve impulses/ganglion tell it that it has something on it's leg and it is probably food.
I might give some thought about what would give a machine pleasure rather than fulfilling it's design function.


reader Gordon said...

Hmm, convincing evidence...
For the past ten years or so you have been taking part in a Turing test with me on the other end... :)

AI research has been kind of stuck for the past 35 years or so, but machine learning has flourished. True AI will likely not arise from a project set up to build one. It will be incremental from machine learning improving pattern recognition and from incorporating increasing levels of powerful abstract mathematical techniques. For example, in 1943 and following, McCulloch and Pitts devised the neural nets concept. For awhile this became very hot in AI research, then cooled off, as did the commercialization of intelligent agents. Fairly recently it has been resurrected in the form of "Deep Learning" techniques pioneered by Geoff Hinton. These techniques use graph theory, a branch of discrete math. Other techniques using Gaussian kernels have made great progress.
This is not to produce an AI, but eventually as techniques incorporate more powerful mathematical models, they will be used in AIs. All these billions of dollars for Human Brain projects will help with spin off info, but I believe trying to build an AI like the human brain is misguided but typical of humans (oops, gave away that I am not human in our Imitation Game Turing test :)
I also disagree that Moore's Law will end soon. Obviously eventually it will end, but again people think in the current paradigmatic (hate the word) patterns. It will slow down then suddenly someone will "think different(sic--thanks, Apple for the crappy solecism)" and a new way of doing things may emerge (The Deus Ex Machina model :) )


reader Gordon said...

"myelin cells are also biologically active and interact with the nerve."
...as are glial cells that used to be though of providing support only. Also the brain also communicates with the body through various hormones.


reader Gordon said...

Hmmm, toast....

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


reader Gordon said...

True. Have you read Michael Lewis' first book, "Liars' Poker"?
Hilarious, but very depressing at the same time.


reader Gordon said...

"Someday, even the Fed and the economists it funds will be replaced by a program."


The Fed shouldn't be replaced by anything. It should just be eliminated. It is simply an extension of the modern Medicis---the investment banks.

The public does not seem to realize that the Fed was set up to serve the needs of the bankers, is a private corporation with shareholders. It is a driver in the whole greed machine.


reader Uncle Al said...

1) A muslim robot must injure an infidel or, through inaction, allow an infidel to come to harm.

2) A muslim robot must obey the orders given it by mullahs, except where such orders would conflict with the First Law.

3) A muslim robot must commit explosive suicide to harm infidels, as long as such suicide does not conflict with the First or Second Law.


reader Mark said...

I think the most dificult thing to program is an idea. Thats because the idea is depending of the environment, for example the justice, the goodness, etc. It I were muslim probably I would have other ideas than I have. There are many holes in AI, for example object recognition still requires more work. The best approach is practically a brute force program (deep learning). I would be interesting to leave 1 million of robots in Marte such that they evolve to more intelligent and empatic robots. It is likely that a AI computer discover the absolute rules of universe beyond any human theory as QFT or string theories, using hyper capabilities of maths and imagination , and such rules could be so extremely complex that a 200-IQ human would requiere 1000 years to understand them in detail.


reader cynholt said...

Well then, let's instead mimic a dolphin's brain or an elephant's brain, Gordon. Mimicking an ant's brain might yield some pretty interesting results, too. But if you really want to find something that's truly exceptional in term of intelligence, only in our eyes, of course, we oughta leave Tara and the other rocky planets and head out to a gas giant, be it Saturn or Jupiter. Gaseous intelligent would be strangely alien to us, but decidedly intelligent nonetheless. And perhaps it might be more intelligent than anything you'd find around here on any of our rocky planets.

Let's move out even farther to the icy bodies, beyond Neptune and the Kuiper belt all the way out to the Oort cloud. A dirty snowball suspended in the Oort cloud may indeed be home to some really exceptional intelligent life. We might have to speed up the evolutionary process in these alien lifeforms, but at least they are far enough along evolutionary-wise to make this a reality.

I don't think we can go all the way back to the first moments after the Big Bang and find anything with intelligence or the potential to be intelligent. The building blocks simply weren't there to do so. You might think otherwise if you are naive enough to believe that honest-to-god God particles are swimming around like tiny brain cells in a bowl of quark soup. That would qualify as intelligent design, but something which any intelligent person would regard as a medieval fairy tale. Fast forward to the cosmic background radiation and I seriously doubt there is anything there that could conceivably be construed as intelligent either.

In order to find anything with intelligence or the potential to be intelligent, we need to at least move forward in time to the early days of star formation and preferably following a massive series of supernova explosions that produced a rich abundance of elements beyond iron. The building blocks for intelligence are bound to be there. We are living proof that this is more than a mere possibility.


reader MikeNov said...

Why is 1998 on NASA's chart. The total is just 82%, and 2007 should have higher odds since it is at 62 vs 61 for 1998. Several other years must be in the running with a few percentage points each.


reader Guest said...

It depends on one's definition of "revolutionary".

Is this (http://www.technologyreview.com/view/532156/googles-secretive-deepmind-startup-unveils-a-neural-turing-machine/) it? We'll see...


reader davideisenstadt said...

eh...when the computer can scour its environment for energy, locate a donut, eat it to fuel its operation, and then metabolize i,t then I will consider the case of AI.


reader Luboš Motl said...

Yes, indeed, there have been other years aside from the listed ones, like 2013, 2012, 2011, and so on. ;-) Was your comment a joke?


reader Luboš Motl said...

Nice and idealistic. The only problem is that the Qur'an says rather clearly that infidels should be killed, all Muslims know it, and there is no potential for most Muslims to deny this essence of Islam or to split from the rest.


reader Shannon said...

Lubos, at least I give here the beginning of a solution... and so does Marine Le Pen ;-). I have never understood this "infidels" thing... If you are not muslim you can't possibly be an infidel ! It is like saying your neighbour's wife is infidel to you !! Only so called "bad muslims" can be infidels...
I think a new Islam is possible.


reader Luboš Motl said...

Lol, I thought that infidel and disbeliever were exactly synonymous!


reader Shannon said...

To me disbelieving is easier to fake and so is believing. Being an infidel however is being recognised as a betrayer... not following the precepts of Islam.

What do you think of this answer to a moderate muslim in the US https://www.youtube.com/watch?v=rCLj5jKnsIU


reader Gordon said...

It is not even necessary or desirable to mimic other animals' brains either (note to the religious--humans are animals).
All you need is self-awareness, an ability to sample and experience its environment, sufficient memory, a fast cpu, redundancy, error correction, and an input/output device.
Given that there are around 10^23 stars in just the observable universe (horizon), and that most of those stars have some sort of planetary system, unlike our host, i am quite confident that life and intelligent life, whether biological, silicon, or whatever, exists outside of earth.


reader Shannon said...

Must be the climate... ;-)


reader Luboš Motl said...

Dear Shannon, I guess that the French are even better at being adulterous than being infidels ;-) but I assure you that "infidel" means what I say, see e.g.

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



and both pagans and atheists are excellent examples of "infidels". Maybe you confused the word "infidel" with heretics, renegades, apostates, or something like that.


reader John Archer said...

Shannon,

Just a guess, but I'm thinking your take on 'infidel' comes from too strict an adherence to the Latin etymology, possibly — i.e. strictly unfaithful. N'est pas?

A classical education is not necessarily a good thing. It can result in one leading oneself up the garden path. It can certainly produce the most irritating of pedants. The latter not in your case though. :)


reader Shannon said...

Nope. Wikipedia says: "Infidel (literally "unfaithful") is a term used in certain religions, especially Christianity and Islam, for one who has no religious beliefs.." The definition is that infidel means "unfaithful", word also used for a husband who betrays his wife.
As a christian I have never used or heard the word infidel but rather non-follower or not practising...


reader maznak said...

The AI that will probably destroy mankind is the one that will be programmed to "do good".


reader Shannon said...

Wikipedia says "infidel" (literally unfaithful)..."... Who is being led the garden path ?


reader Cogniscentum said...

Evolution has the advantage of being able to exploit laws of nature that it does not need to understand. For all we know, memory and the power of insight may depend on some aspect of physics still unexplained by string theory or rules of QM that we just don't see apply that give the brain vastly more computational power than we believe it to have.


We also have the combined wisdom of billions of years of evolution encoded in our brains. I would be impressed if we could simulate the brain of an ant, not based on programmatic rules for situations, but on rules that are an emergent expression of the original organization of the "brain," like real ants.


reader Jack Abramoff said...

Thanks for nothing but fiction Assmov. Seriously,
there's more loopholes/holes in The Three Laws than in LQG.


reader pEGO said...

I have masters degree in mechatronics, but I prefer human touch. AI is heartless, creativity free system that according to human brain has 120 IQ and it is polyhistor that jogges it to 180 IQ. Due to robust matrix of feelings EQ is equally almost to zero. You could see Sheldon in Big Bang theory, that’s look like perfect AI could be.
;)
www.youtube.com/watch?v=HzR9xF4iB1k


reader Smoking Frog said...

As a christian I have never used or heard the word infidel

Nor even read it? The Catholic Church uses it, and probably so do various other Christian churches, especially the fundamentalist Protestants. Here's the "Infidels" entry of the Catholic Encyclopedia, and here it is in documents by Pius XII and Paul VI>.


reader Smoking Frog said...

So, again, we non-muslims cannot possibly be unfaithful/infidel to Islam. Do you understand the logic or have you reached the gate already ?

As far as I know, Islam teaches that everyone is already a Muslim, but some don't know it, and some knowingly deny it. However, not only the knowing deniers are called infidels, and this is consistent with dictionary definitions.


reader Smoking Frog said...

the Qur'an says rather clearly that infidels should be killed

I wonder if that survives their standard system of interpretation. I don't know the system, but part of it seems to be that later statements in the Qur'an supersede earlier ones.


reader Smoking Frog said...

Infidel and unbeliever are synonymous. A disbeliever knows of something but does not believe it. An unbeliever may not even know of it.


reader Smoking Frog said...

Saturday's Los Angeles Times played this like a historic event, with a big, serious-looking headline. The article quotes someone from Woods Hole saying that the new high should put to rest the "hiatus" idea, but the hiatus is evident in a chart on the front page. The article talked about a global mean temperature of 57 F. as if global mean temperature (not temperature anomaly) were standard or usual in talk about global warming.

It just made me sick with dismay. How can anyone fight this kind of thing? So "official," so heavy! I wish enough people who know what they are talking about would assault the media with protests.


reader sabretruthtiger said...

Don't ask an alarmist a science question!! They only deal in ad hominems


reader sabretruthtiger said...

Obviously not the Alarmist media otherwise they'd report the ACTUAL global temperature including uncertainty from the error margins.
They'd also report the plateau starting from 98 and ignoring the El Nino 98 high point to avoid cherry picking accusations would report that there is a slowing of the warming.


reader MikeN said...

I would think 1998 is fifth on the list, behind 2007, yet NASA has it 4th on a list of 4.


reader John Archer said...

That was superb, Shannon. Truly superb. Top stuff!

That Brigitte Gabriel is the first straight-talker I've heard on this issue in a public forum like that.

She won't be interviewed by al beeb anytime soon at least, that's for sure, but it's this kind of discourse that sorely needs to be put before the public and needs to replace the toxic misinformation they're currently being fed.

Also I think the fact that Mlle Gabriel was a woman helps a lot. A man speaking thus can very easily come across to many as just channelling raw aggression only. A woman doesn't suffer that disadvantage, but especially so here as the 'over-empowered' questioner, who needed that good slapping down, was a female herself. I'd say the muslime was expecting a big peecee round of applause for her 'brilliant' question (which she would have undoubtedly got if it had been a typical hand-picked BBC audience) but she got the slaughter man's bolt straight through her forehead instead. No halal here! Quite right too. :)

That video deserves a wide circulation. Well selected!


reader Shannon said...

Smoking Frog, I have never read it either. Maybe your protestant US church uses it but I am no protestant at all so I don't know. I can assure you that at church the priests never ever use that word and it is written nowhere on the books and prints given for mass.


reader Luboš Motl said...

Non-follower or not-practising is quite a politically correct language! ;-)


reader Shannon said...

In this case Lubos it is not PC it is just accurate. ;-)


reader Luboš Motl said...

Infidel is equally accurate.


reader Luboš Motl said...

Right, John. The infidel is the most standard translation of the Arabic "Kafir"

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


reader Shannon said...

If you think so then you must be a muslim Lubos ;-)


reader Shannon said...

;-) I'm glad you enjoyed the video. I also see a tiny bit of blackmailing from the "muslimette"...


reader Luboš Motl said...

Interesting! It seems to be 5th by this GISS graph:

https://lh4.googleusercontent.com/-l2_HKUHfLFc/VLlLIy1UNQI/AAAAAAAAHlA/r_kBNw4TGW0/w731-h461-no/giss-up-to-2014.PNG



from my previous blog post.


You know, maybe 1998 has not only a different mean value but also a different error margin, probably a larger one. If it has a larger error margin, it is a higher chance to deviate in the up direction from the mean value, and that increases the probability of its being the warmest year, so that it's 4th by the probability of being the warmest.


My point is that the sorting of the mean values and sorting of the probabilities isn't the same if the error margins are different for different years.


reader Luboš Motl said...

Sorry, Mažňáku, you are ignoring the crucial point here that the adjective "good" – a moral appraisal – only makes sense given a human experience and heritage.


There is no technical canonical quantification of "good" that would be available to a computer starting from scratch – this was really the main point of this blog post, although said in different words.


I am not a moral relativist in practice, someone who would say that everything is equally moral or immoral as everything else, but I am surely a moral relativist in principle, agreeing that all the morality has always been created by individuals or societies and "good" in one way or another meant "good for him" or "good for them".


reader Luboš Motl said...

How did you (incorrectly) exclude the possibility that I am actually an infidel? ;-)


reader Alexander Ač said...

Yes, there IS "slowing" in the global warming, exactly. So no "pause" or anything close to is. Expect speed-up in atmosphere warming once we have El Nino and/or positive phase of PDO:

http://2.bp.blogspot.com/-JdRrmZYtFBs/VLrhKBg2xQI/AAAAAAAAPjc/LW6xtKcgf6U/s1600/trend-since-1998.png


reader Shannon said...

Because you have nothing to do with Islam. Islam is for muslims, not for non muslims. Most muslims agree with that and I think they should remind their radicals about this simple fact.


reader Luboš Motl said...

Sorry, Alexander, by using the noisy data of the last 20 years, one cannot decide whether the "underlying warming trend", if this quantity makes any sense at all, remained positive or became negative.


Within the error margin, the trend is zero which is what people - including your fellow alarmist crooks - talk about the pause or the hiatus. There just hasn't been a warming for 2 decades, and if there were a warming, it was so slow that it couldn't have been measured with any significant confidence. It could have been cooling, too.


reader Shannon said...

Islam recognises the three monotheisms so they cannot say jews or Christians are infidels. You never hear them say that.


reader Luboš Motl said...

Dear Shannon, according to the Quran, both Islam and the right to live (without being stoned to death or constantly harassed) is only for Muslims.


reader Luboš Motl said...

Shannon, they are called monotheisms but they're still lethally wrong beliefs according to Islam and that's why the Islam counts Christians and Jews among infidels.

You don't have to go to any hardcore Muslim sects to see it in this way. Just go to ordinary schools in a relatively moderate Muslim country, e.g. Egypt, and you will learn that Christians and Jews are infidels, too, see:

http://www.jpost.com/Middle-East/Egyptian-curricula-states-Jews-Christians-are-infidels


reader Shannon said...

Lubos, yes but the right to live as a muslim. It is the snake that eats its tail. This religion is self centered. We shouldn't care what they do to themselves. Only those of good will and peaceful hearts can breathe in our Christian world.


reader Luboš Motl said...

It's self-centered because it wants to exterminate everything that is not "self". That's what the plan called "jihad" means.


reader Shannon said...

This 2011 article shows that there is a will from the think-tank muslims to change their education system. That only is the proof that there is good in them.... or the seed of a new Islam.


reader Alexander Ač said...

Hi Luboš,


I remember our cancelled bet that 2010 will be the hottest in instrumental temperature record ;-)

But to the topic of (no) global warming in the last (xy) years, I am sure you and your fellow skeptics will not be impressed by this year (2015) which will be probably also record hot (certainly if El Nino is developed). And can you explain why 2015 was the hottest neutral year, warmer than 1998?

Best,


Alex


reader Shannon said...

Yes. I know... They need to feel the organised peaceful resistance. We have to be skillful and clever about this. The only thing we can do is to never give in and protect ourselves the best we can. Show them our organised resistance against their obscurantism. We have to communicate with them through our strength to resist without entering their game. Muslims respect strength.
A good police, secret services and a complete review of the Schengen Agreement is needed (M.Le Pen is strongly calling for all this and she is right).


reader Luboš Motl said...

First of all, I've never made such a bet.

Second of all, no one can explain that "2015 was the hottest..." because 2015 just started.

Even if you meant 2014, it is not true. 2014 was between 3rd (UAH AMSU) and 6th-7th (RSS AMSU) warmest year among the recent 36 years, a whopping 0.3 deg C cooler than 1998.

Third, 2014 didn't include a full-fledged El Nino episode but the Earth has been in the weak El Nino conditions for many months, on the verge of becoming an El Nino episode by the standard definition, see the newest ENSO update:

http://www.cpc.ncep.noaa.gov/products/analysis_monitoring/lanina/enso_evolution-status-fcsts-web.pdf



So it is misleading to say that it was an ENSO-neutral year.


Fifth, among various years, one of them has to be the warmest one, and because the temperature fluctuations are largely random and unpredictable, it is largely a random and unpredictable year.


An El Nino episode increases the probability that the year is the warmest one, and 1998 indeed came during/after an El Nino of the century. However, there is no strict relationship here. Some years without El Nino or with La Nina ended up being rather warm and vice versa.


Looking for explanations of these random noisy data where the differences are around 0.01 °C and much smaller than the error margin is a sign of someone's stupidity, a dumb exercise similar to the creation of horoscopes. i am not looking for any explanation here - and I will kindly leave this activity to idiots.


reader Alexander Ač said...

Luboš,


you are Master of spinning the facts, I have to admitt.

You know very well, that *global* temperature is inherently a noisy measure, but we have no better, less noisy measure to show global change. One cannot expect to reach global temperature records by huge margins. And that is not the point, either.


I am sure even is global oceans were rising 5 or 10 mm/year, summer Arctic ice is gone, and boreal forest is burnt to the ground, few people would still seriously claim it is perfectly natural force of nature.


Well, in a way I understand that, I am very tolerable person, however I am not going to join these people.


Best,


Alex


reader Luboš Motl said...

Alexander, "admit" is spelled with one "t".


I think that your claim that "global temperature is intrinsically noisy" is deeply misleading. It's noisy but less noisy than most other quantities that actually matter. It's the very point of looking at the "global average" for the demagogues to *reduce* the normal noise so that even hundredths of a degree may look "big" and may be cherry-picked and used in totally idiotic propaganda press releases.


In reality, the local temperature anomalies - even if averaged over a month - are comparable to 2-3 degrees Celsius and hundredths of a degree are clearly *zero*. They have no consequences that matter for anything.


It is complete bullšit that "we don't have any better" quantity to describe the global change. What about the world GDP and hundreds of other quantities that are more important than some hundredths of a degree in a contrived, largely ill-defined, convention-dependent average of temperatures?


There is *nothing* important about the global mean temperature and only mental midgets think otherwise.


Of course that if and when the Arctic sea ice completely melts in the summer, which is rather unlikely but totally plausible within the next 100 years or so, there will still be *zero* evidence that there is anything unnatural about it. There won't be anything dangerous about it, either.


reader Hartog said...

Is there a definition of "thinking" that all this refers to?


reader br said...

And that was Asimov's point when he wrote about the consequences of his 'three laws of robotics' - despite sounding good on a first reading and working for a while, the logical conclusion was that all humans should be basically imprisoned http://en.wikipedia.org/wiki/The_Evitable_Conflict . Not that I think it must be that way, but the 'law of unintended consequences' should always be taken seriously, in proportion to how complicated a system becomes.


reader Shannon said...

Alexander, your hair ! Where is it gone ? You look tough now. A moustache and a black leather jacket with studs would suit you too ;-). I too want to save the planet :)


reader Albert Zotkin said...

Do you know that human beings have a lot of working neurons in their stomachs (gastrointestinal system) not only in their brains. So most people think by means of their stomachs.


reader Alexander Ač said...

Shannon,


I did not write it is gone. I wrote IF it is gone, some people will have no problem with it. Do you read what I write? I doubt it :-)


Alex


reader Alexander Ač said...

Sure, ad hominem by John Archer is totally OK ;-)


Alex


reader Artur Krejčík said...

Help them evolve? Wow thats so etnocentric and NAIVE, islam have not evolved 1400 years. Im not willing to offer my country for the religious reformation, that may take place 200 years from now, or never thats not certain, but what is certain that in the meantime we will have islamic supremacist claims, terorism, rapes, and during the reformation alot of blood. (thats what happens during the reformation)


reader br said...

'Belief' in some sense of the word is built in to neural networks - due to noisy and new inputs, the outputs have values between 0 and 1. So if you input a hand-written letter 'a', then if it is clear the output for 'a' might be 95%, and the outputs for the other letters might be less than 1%. However, for an unclear letter 'a', the output drops - to 80%, 60%, or even 40%. If a neural net is allowed to 'run free' and start modifying its weights without a clear training, then it can end up in all sorts of funny states. How to train a network is an art, even when the programmer knows what s/he's doing, and can have all sorts of unintended local minima. That could explain a lot of human beliefs - local minima in a network that doesn't have clear training. I reckon a study of how to get out of those minima could be really beneficial to the whole of mankind!


reader Luboš Motl said...

Interesting, Alexander, you seem to have confused your hair with the Arctic sea ice.


If you don't care that the former is gone (and I don't care about a similar, less pronounced trend, either), why should Shannon care about the latter? ;-)


reader Alexander Ač said...

That is true ;-) BTW I stil have some hair, but given this trend, to have the rest short is more convenient and parctical.. :)


Alex


reader Shannon said...

Ok fine, Artur, so what do we do ? Nothing ? Deport them ? Where ? Accept full submission to the charia laws ? It's all nice to snivel about their being nasty people but when do we take action that will dig deep into the roots of the problem and last for ever ?
I am not naïve, I look further, and you are a professional mourner.


reader Artur Krejčík said...

France will pay for the mistakes they did (mass immigration) Czechia can just close the doors by selective (non-muslims only) immigration policies, ban/start enforcing laws on the islamic hate speech...
The root of the problem is the open borders to islam, to something we are in war with. We can and should work on changing it but not on our playground.


reader Smoking Frog said...

I have never read it either.

So you're saying that reading the Wiki doesn't count. It counted for purposes of my question, though. I did not mean to ask whether you had ever read it in any "book or print given for mass," but whether you had ever read of its usage in Catholicism or Christianity. The two papal documents I linked use it.

I was raised as a Catholic, but nowadays I don't go to church. Apparently you do go to church. I find it a bit surprising when a Catholic or other Christian thinks "infidel" is exclusively a Muslim word.

I can't swear that I've ever heard a priest, monk, or nun use it, but there's a good chance I have. I'm thinking especially of an alcoholic pastor who one Sunday, while drunk, delivered a sermon condemning Martin Luther (funny), and an aunt of mine who was a Sister of Notre Dame. I think there's also a chance that I heard my Christian Doctrine teacher (a priest) use it.


reader Smoking Frog said...

That's not what Wikipedia says, and I think Wikipedia is right.

Contrary to what you seem to think, an infidel is not one who is unfaithful to a commitment he has made (as in adultery).


reader John Archer said...

Of course it's OK!

You kicked off with a pure trolling: "Geez, who cares about such abstract thing as global temperature? Globaly warm or cold, who cares?"

You didn't address a single point made in the post.

After that anything goes, because YOU changed the game — to no holds barred.


reader Shannon said...

We are already paying for these mistakes. We need Marine Le Pen, quick. I've been saying it for a decade now.


reader Shannon said...

Yes Wiki says "infidel" is like unfaithful: "Infidel (literally "unfaithful") is a term used in certain religions bla bla..." see the first sentence of the definition.


reader Shannon said...

Only religious people might use it between themselves but we, believers, church-goers, ordinary people ;) etc, never use it. And that is what is important here Smoking Frog: what has leaked into the people themselves. The message.


reader Luboš Motl said...

Right, Artur. Trying to change it while already having millions of them on our territory may be a little bit too late.


Last time we needed to change the fellow Germans who lived with us, it was hard and we needed some Soviet and U.S. help to actually change those Germans. ;-)


reader Gordon said...

You just described a Chicago cop;)


reader donqpublic said...

Worked with an ex-Chicago cop years ago. Unlike the Viet Cong, at least Chicago cops took prisoners on a routine basis. Maybe taking prisoners had something to do with the reward structure of getting to eat donuts rather than eating fried rice and dog meat? One of his on-the-job words of wisdom for discrete repeating elements of social reality was a catchy phrase: "no sense being a nigger, if you can't act like a nigger." I presume this tautological hypothetical would apply to recurring intelligent machine "behavior," and Muslims too. I suppose the real proof of AI would be if, given the option, machine X reported "fuck it" before it fried its mother board just to have the peak experience of super sex by plugging in to the 220 instead of the 110 outlet. Maybe if Barack ate donuts he'd take prisoners instead of droning them to death and forgoing the intelligence.


reader Don said...

good stuff! This is REAL news! Thanks for update!


reader Petr said...

Luboš, "bullshit" is spelled with "sh".


reader Steve said...

Heads up! The HFi Planck French website has posted new summary results from Bicep2/Keck/Planck. See the site (in French):http://public.planck.fr/resultats/253-la-reponse-de-bicep2-keck-planck


reader Luboš Motl said...

Hi, I saw it yesterday. The paper will be out on Tuesday morning, rumors say.


I don't see how this paper may differ from the other papers that just mix "signal" datasets from BICEP2 with "non-signal" dataset we know from Planck - the result of which is clearly a weakening of the signal.


The outcome is a small excess that is not enough for a discovery,and an upper bound "r is smaller than 0.13" which is a higher value of allowed "r" than the previous one.


I think that I won't dedicate a blog post to this paper and will patiently wait for Keck Array and BICEP3, among others.