Sunday, November 22, 2009


Here is a non-exhaustive list of my beliefs and opinions, both conventional and unusual (or worse):

- IQ is the best known measure of cognitive ability, predicts lifetime success in many domains, is predominantly genetically determined in modern populations, and systematically differs between various ethnic groups

- evolution is a fact, not a theory

- there is no god around these parts, and even if there was, he/she/it would not be somebody worthy of invitation for dinner, much less worship

- espousing the three opinions above is enough to make you a pariah almost everywhere

- affirmative action is evil, it is a form of institutional racism directed primarily against Asians

- I give you 9:1 odds we will be destroyed in an artificial intelligence singularity

- I have no soul, YMMV

- hard drugs should be legalized

- the FDA should be abolished

- medical licensure should be abolished

- governments should be abolished and replaced by non-monopolistic, commercial providers of law and order

- World of Warcraft is a masterpiece of psychoengineering, a dark maw that swallows the weak of mind

- charity is noble

- welfare is immoral

- salsa is fun

- the demographic transition is still a mystery, its effects are mildly dysgenic but it doesn't matter (see singularity, above)

- if I ever get root access to my mind, about 70% of it will be removed in the first round of modding

- carbon dioxide contributes to retention of heat in the atmosphere but nobody really knows how much

- carbon dioxide is good for plants (yes, I actually read the Jasper Ridge research), and therefore it is good for people

- liberals are even more hypocritical than conservatives but not by much

- highway speed limits are one of the stupidest laws in the US, destroying millions of QALY-adjusted life-years each year

- deer are the enemy, and should be shot, poisoned or vaporized with lasers, whatever comes handy

- as a point of meta-ethics, ethical theories that fail to specify an in-group are gibberish

- as a point of meta-ethics, ethical theories must be computable using available resources

- female hypergamy offers both challenges and opportunities to the conscientious male

- even for a math-challenged neurologist, the many-world interpretation of QM is obviously more reasonable than the alternatives

- there is no proof that lowering cholesterol makes you healthier, although some things that make you healthier (such as statins and exercise) do lower cholesterol

- don't eat margarine

- the only a priori moral duty within my in-group is the duty of non-violence

- consent is only possible if a meaningful option to refuse is given

- taxation by a territorial state denies the meaningful option to refuse

- taxation is a form of violence

- helmetless biking is good for the rest of us, who might need a kidney someday

- kids are fun to have

- it is sophomoric to strive for happiness, we grown-ups have more important things to do

- once you go Mac, you'll never go back

- it is wrong to vote, except to abolish elections, and maybe not even then

- more than 70% of modern medicine is a waste of time and money but the average value of medicine is still positive

- a laser bug shield is going to be the next killer app from Microsoft

- I am not sure what I am trying to signal here but it sure feels good

- some of Chopin's music is divine

- Glock 27, .40 S&W

- it bears repeating that non-initiation of violence is the only criterion for differentiating between good and evil

- network redundancy is extremely important

- short-selling, insider trading, and hostile takeovers are good for the stockholders

- the only legitimate sources of law are non-monopolistic providers, such as trade associations, private licensing authorities, and commercial courts

- in a timeless multiverse, causation is nothing but correlation, and consciousness is a property of some series of patterns that exhibit a significant degree of self-reference

- everything that can exist, does, but most of it really far away in the configuration space

- God exists ... actually all of them do, which is why you could call me a polytheist, or even a maxitheist (all thinkable gods exist somewhere)

- environmentalism is immoral, since it explicitly ascribes an intrinsic value to non-sentient objects, above and beyond their instrumental value to members of my in-group

- cryonics is kooky, and my Alcor # is 1941

- socialists are mean, nasty, hypocritical and confused

- fascists are mean, nasty and confused but at least not so hypocritical

- still, socialists are the lesser evil

and, last but not least:

- don't hurt dachshunds or you will make me very angry

Wednesday, November 18, 2009

Of beating hearts

I can trace the beginning of the road leading towards my current political views to studying human physiology, sometime around 1988. At that time I was cramming for an exam in physiology in my medical school, and I read about the risk of cardiac death as a function of the variability of heart rate. To my surprise, the less variable your heart rate, the higher your risk of suddenly dropping dead. It seemed like a paradox that order, a steady, unchanging heartbeat, i seen mostly in patients suffering from heart failure, while chaotic variations in heart rate are typical of the healthy young.

It may be strange to see political lessons in cardiac physiology - still, the juxtaposition of death and order on the one hand, and chaos and life on the other hand, is quite stark and it stayed in the background of my thoughts ever since. And as I observed success and failure in various domains, I noticed that there are those amazing analogies between the way the sino-atrial node and e.g. the internet are built.

Let me give you a bit of background on the physiology of heart rhythm. The SA node is the natural pacemaker of the heart. It integrates inputs from multiple sources (vagal nerve, sympathetic innervation, various hormonal influences) and consists of a network of cells which exhibit spontaneous cyclic depolarization, which is the source of the rhythm. Individual cells differ in their tendency to depolarize, the beat frequency they collectively produce is the outcome of an interaction between multiple, individual, independent oscillators, which are capable of partially substituting for each other. There is no master rhythm cell, the buck doesn't stop in any particular place, and the rhythm is therefore quite chaotic - it exhibits spontaneous minor and major variations stemming from multiple influences and multiple independent decisions - and this is precisely why the heart mostly keeps on beating, reasonably well adjusting to various conditions. Yet, as the heart is stressed, the SA cells start dancing to a single tune - some cells are eliminated from the decision-making, only a subset of similar cells can continue to function, and the heart settles on a much more rigid, usually faster rhythm. This is bad: instead of a graded, appropriate response to stimuli the heart my exhibit sudden, catastrophic changes in rhythm, including the deadly ventricular fibrillation.

A very similar pattern is observed whenever a network of independently interacting, redundant elements (SA cells, internet routers, stock market traders) changes into a strict hierarchy, or a group of identically-reacting ones (stressed SA cells, rule-bound administrators). The wisdom of crowds is lost, information bottlenecks appear at the top, feedback loops are weakened. A strict hierarchy by definition is not redundant and this is why when a hierarchy fails, it fails catastrophically. Any network that is dependent on a single master node to function (whether we are talking about a central air transport hub or a dictator) will be only as reliable as the master node.

Examples of highly successful redundant networks abound: the social construction of science is one. Every successful market (i.e. efficiently transforming inputs into outputs) is one. Reducing the redundancy and degrees of freedom in such networks very reliably degrades their performance, in Soviet Russia 90 years ago, and in the USA today. When everybody runs the same operating system, one virus brings everyone down. When everybody eats potatoes, the blight can kill millions. I could spend a long time listing examples of the success of network redundancy and the dangers of centralization but it is hard to come up with examples of the opposite.

This observation may be uncontroversial among network engineers but most humans resist it fiercely in the social and especially political realm. Especially under stress, most people respond by closing ranks, demanding unity, and conformism. A crazy patchwork of security agencies coagulates into Homeland Security: previously, for something bad to happen, every one of the overlapping agencies had to drop the ball, now, it takes only one point of failure. In response to economic disturbances politicians demand more centralized control of the economy, and the masses applaud. This is not a random development but rather a predilection ingrained since our evolutionary past spent in small tribes continually engaged in mortal combat against other tribes, and to an almost identical degree, within the tribe, where loyalty to the group and obedience to a single focal figure of authority was frequently indispensable for survival.

I am probably deficient in the neural hardware that subserves this predilection: I could never generate much enthusiasm for "our" soccer team, I would feel weirded-out by the sight of my compatriots singing the national anthem with their hands over their heart, there is not a single politician that I ever unequivocally admired. I am very loyal to my family and friends, the people I know personally, but I am calmly dispassionate when thinking about groups of strangers. Did you note that I haven't yet used the word "we"? I almost never think as a group member, even when I analyze groups that are outside groups to which I belong.

This may be one of the reasons why my political views, developed via very circuitous route from cardiac physiology, put me in a very small minority, known as market anarchists. We (yes, we) agree that what is good for the reliable functioning of the heart, the discovery of scientific truths, the internet, and the trading of pork bellies, should be also good for governance. Unfortunately, the ability to dispassionately extrapolate from analysis of the SA node to looking at elections is very uncommon. Almost everybody gets bogged down emotionally, fearful of the power vacuum, always yearning for a father figure at the helm of the tribe. Others are entangled in their hubris, imagining the very smart regulations they would impose to bring order to the world. There are probably too few market anarchists in the whole world to fill a soccer stadium, much less create a viable stateless society. And thus we end up with a world where the most important, life-and-death issues like war, peace, and economic prosperity are controlled by territorial monopolies consisting of narcissistic thugs self-selecting into "public service", and the willingly ignorant masses they control.

In the next post in the series I will consider the situations where pluralistic governance could plausibly take hold, and what are the meta-ethical conclusions of cardiac physiology.

Friday, October 2, 2009

What is "subsistence"?

Robin Hanson expects that our descendants (whether biological or machine) will lead a subsistence life, that is the resources available to the median individual will not be significantly greater than what is necessary to allow survival. While I tend to generally agree with his contention, I do have a few quibbles.

First of all, the details of what constitutes "subsistence level" crucially depend on the time horizon one looks at. In the very short term (the time needed to write "In the very short term") all I need to survive is the absence of major alterations in temperature, gravity, radiation levels, etc. To finish this post I also need air. To see the Sun turn into a white dwarf I would need to also conquer aging, keep a gigaerg source of energy, as well as end war and petty violence, just to name a few of the likely long-term subsistence-level resources and conditions.

In this context, the word "subsistence" should not evoke images of swarthy farmers plowing their tiny plots behind burros. Galaxy-long linear accelerators might turn out to be indispensable tools for the would-be trillion-year survivor. One might need to understand and explore the string theory landscape to escape the heat death of the universe, or the Big Rip, or whatever doom awaits this little pocket of dusty vacuum we live in.

But, of course, the notions of long-term vs. short-term are themselves dependent on the entity entertaining them. For an E.coli bacterium in exponential growth phase, long-term may mean beyond two to three minutes, since most of the feedback loops operating within this survivor are calibrated in seconds. One must also remember that there is a hierarchy of (potentially) replicating entities, starting with genes, through cells, to multicellular organisms, to superorganisms, and evolution, depending on the conditions, may act on any of those levels (arguments against simplistic group selection theory acknowledged and irrelevant). The time span of subsistence will then depend on the organizational level of the entity undergoing evolution. The configuration space of likely entities into which humans could evolve is huge and varied, and we know too pitifully little to predict the exact trajectory of post-human evolution. This includes predicting the longevity of the dominant mind species in the post-human ecology.

Our minds and bodies have been blindly shaped by evolution but this process couldn't keep up with technological change, resulting in a mismatch between innate desires and actual fitness-relevant possibilities. I agree with Robin that this situation is unlikely to persist over any evolutionarily long time. I expect that soon we will be supplanted by post-human beings specifically designed for survival in a technological ecosystem, and the natural, blind evolution will thus largely end (who would design these entities is an issue I may consider in another post). But, although we can surmise the post-humans will be good fitness-maximizers, we can say very little about their organizational level, physical form, longevity and the actual physical meaning of "subsistence resources".

Thus, even though Robin is most likely right that our descendants will lead a subsistence life, their desires fully aligned with the goal of survival and none of their resources expended on irrelevant frippery, we should realize that their notion of subsistence may be quite sumptuous by our standards. Maybe they will need to expend stupendous amounts of energy just to prove fitness and to discourage predation. Maybe they will explore the most esoteric realms of post-human mathematics to reliably navigate away from cosmological disasters. Maybe their social graces, needed to stay afloat in a forever-changing political world, would surpass the most Machiavellian of our ruling classes. One is reminded of the adaptive immune system, the energetically expensive and extremely dangerous (see autoimmunity) assemblage of B and T cells that constantly churn and evolve, ever on the lookout for new infectious challenges, with no end in sight - running very quickly just to stand in place.

I think there is a good chance that some of the seemingly non-essential attributes of our lives may in the long term become subsistence necessities, at least in some of the newly created ecological niches. Freedom, unbounded curiosity, never-ending exploration, mastery over matter, and owning very large private spaceships with wormhole generators could in some scenarios be the analogue of having a burro and five acres of land - but hopefully much more fun. Time will tell.

Tuesday, September 15, 2009

The dangers of scientific monopsony

Humans are social animals, naturally forming complex hierarchical structures. It doesn't take much for a meme to sweep through a population and leave almost everybody believing in a lightning-bolt wielding bearded man who lives on Mount Olympus, or some such tomfoolery. While skeptics are continuously gnawing at the party line, we are well designed to stabilize certain beliefs once their level of acceptance reaches some tipping point. Religion and politics come to mind first in this context, and the reason for that is simple: both pertain to beliefs that are extremely important to the day-to-day more-or-less smooth functioning of a tribe. Multiplicity of political opinions is extremely dangerous when your Asmat neighbors think that a boy needs to cut off your head and bring it home as proof of achieving manhood. Uneasy respect for authority is the default state of mind for most humans, and a horror vacui regarding authority is prevalent; this is responsible for the anti-market bias described by Bryan Caplan.

But, the world of quarks and prions explored by science poses different demands on the practitioners of this intellectual art. This is a multi-dimensional landscape whose shape can be only dimly glimpsed, with many a fracture and loop where minds get lost. One mind is not likely to succeed, it takes a whole army but with one crucial feature: We cannot march in lockstep. Herding behavior in science means that the mis-steps of the first few trailblazers are repeated by many, and of course missteps are guaranteed - as Einstein said "If we knew what we are doing, it wouldn't be called research, would it?".

So science, more so than any other field of human endeavor, demands a multiplicity of independent approaches, different states of mind and various skeptics generating thoughts and, very importantly, data collected from differing theoretical beachheads in the landscape, until they all converge and yield a theory that, justifiably, crushes all opposition by its sheer obviousness. If they succumb to the herding behavior, we end up with the amyloid hypothesis.

Worse happens if there are direct practical implications of a hypothesis in the political arena. Aside from the originators of the amyloid hypothesis nobody really cares about it - but there are millions of weaselly politicians, pompous bureaucrats, busybodies of all stripes and plain environmentalist numbskulls who really want climate warming to be proven anthropogenic, dangerous and blameable on their enemies. Since the NSF and a few other government agencies are the oligopsonistic buyers of most climate research, it takes only a few fervent believers placed in some key grant management positions to assure speedy publication of anything that supports their preconceived conclusions. This is really bad: billions of people are fed propaganda masquerading as science, a self-reinforcing vicious circle of delusion develops, and trillions of dollars could be washed down the drain, simply because the likes of Mann and Briffa et al. cooked up some convenient truths for the Al Gore's and Greenpeaces to regurgitate.

I used to believe that public funding of science is really a great idea but now I am no longer sure about it. Centralized funding is all too likely to convert a pluralistic enterprise into a stampede to jump on whatever bandwagon goes fastest. Instead of truth-finding we get dogma. A resilient, segmented network (as good as the sum of its nodes) becomes a brittle hierarchy (as bad as the guy on top). I am convinced that the general lack of appreciation for segmented networks and the undue reverence for hierarchy are one of the roots causes of most social evils (I will blog on it later) but for now let me just finish with the following: Beware of the taint that comes with government money. If there is the faintest suspicion that the state or some of its clients benefit from a publicly funded research result, read the minority opinion.

The truth will set you free.

Friday, July 31, 2009

Me too on Mr Gates

It seems that everybody has already blogged on the arrest of Mr Gates on charges of "disorderly behavior". Such a trivial occurrence is hardly worth writing about, still, there are some useful observations to be made. But first, my take on the facts:

Sgt. Crowley entered Mr Gates' residence based on a tip about a possible burglary and found out that no burglary in fact occurred. Mr Gates became verbally abusive and generally behaved like a stuck-up, racist jerk (yes, he was racist - he verbally attacked Sgt. Crowley because of Sgt. Crowley's race, we can be confident that Mr Gates would not have similarly attacked a black officer). Sgt. Crowley handcuffed Mr Gates and arrested him on charges of disorderly conduct.

And here is what I think about it:

We live in a bizarre society where the most trivial aspects of the affair, namely Mr Gates' and Sgt. Crowley's skin color, attract a truly stupendous amount of attention, while what I would see as the only important issue gets totally ignored. In my not-so-humble opinion it is very wrong that a citizen can be handcuffed, arrested, mug-shotted, threatened with further prosecution and generally harassed simply because he screamed racist or otherwise insulting stuff at an officer. The officer is a servant of the people, and after doing his duty - protecting life and property, he should shut up, humbly return to his squad car and leave the citizen to enjoy his screaming unmolested. Instead, he has the authority to destroy a man's existence based on the flimsiest of excuses.

The popular obsession with race diverts attention from the important issue - the creeping expansion of the police state, with ever more SWAT, more surveillance, more officers intruding ever more aggressively into our everyday lives. Where is this going to end?

Thursday, July 30, 2009

Malleus Amyloidarum

If you are passing familiar with the subject of Alzheimer's disease (AD), you have certainly heard of amyloid. Almost everybody in the field will tell you that amyloid is that toxic gunk which clogs up cellular machinery and kills cells causing AD. And everybody is wrong.

It all started with a case of early-onset dementia described by Alois Alzheimer who noted the presence of red-staining material in post-mortem examination of his patient. Alzheimer wrote that the patient was clearly different from the usual senile dementia cases by clinical criteria such as age of onset and duration but later other researchers observed that the same material was also more frequently present in the brains of elderly dementia cases. Somehow, this persuaded them that these different clinical presentations must be one and the same disease - probably under the influence of the infectious disease paradigm, where finding the same pathogen in different locations did not alter the diagnosis of, say, tuberculosis. But, in medicine we frequently see similar images in diseases with completely disparate mechanisms - for example, one will find a lot of similar-appearing immune cells during an infection (where they help combat the pathogen) and in autoimmunity (where they actually cause the disease). In the absence of a good understanding of the specific condition it is unwise to commit oneself to one or the other interpretation of histopathological data.

Fast forward a hundred years - It is commonly accepted to refer to non-familial senile dementia as "Alzheimer's", still in the absence of any understanding of the relationship between the familial (genetic) and sporadic cases. The molecular biology revolution brings early fruit - explanation of the cause of Huntington's disease, where a mutated gene starts producing gunk not unlike amyloid although in a different location. An ambitious research program is formulated by "Alzheimer's" researchers - to use genetic approaches to elucidate the cause of familial AD and thus provide an explanation of the much more common sporadic cases (We "know" they are one and the same, don't we?). Stunning success follows - the APP (amyloid precursor protein) gene is found to be abnormal in a few families with AD! The product of APP, or beta-amyloid, is a major component of the amyloid plaques found in both familial and sporadic cases! If amyloid is added to neural cells in culture, they die! As Mr Gore would have said if he was into dementia research, "The science is in!" Case closed.

Now all we need to do is to tie up some loose ends, and find a method for removing amyloid from brains, and presto, we have a cure for senile dementia. Nobel prizes and profits will follow.

So hundreds of millions, and later billions of dollars pour into amyloid research. One of the loose ends is finding mutations in the APP gene in sporadic cases of senile dementia. Yet, here a setback occurs - not only such mutations are not found, it is definitely proven that they are absent. For anybody with a background in genetics (like me), this is a major red flag - you have one condition, early-onset AD, *with* a mutation, and a similar but clearly clinically different condition, late-onset sporadic "AD", *without* the mutation. The geneticist will automatically conclude that these are different diseases with some superficial similarities, rather than slightly differing manifestations of the same disease. But, the Baptists (the Beta-Amyloid-Protein people) cheerfully press on with their research.

Attention is directed towards demented mice. It's hard to do research on demented old folks, one needs something simpler, like a mouse with a mutation in the gene analogous to the one that is damaged in demented humans. Considerable resources are brought to bear, and here follows a disappointment - mice with APP mutations do not develop dementia :(

Well, never mind - let's make a mouse with not only an APP mutation but also a mutation from a different form of early-onset AD, the presenilin-1 mutation, and for good measure, a mutated tau protein from yet another familiar disease, frontotemporal dementia. Success! Mice develop deposits of amyloid and get sick. As the venerable Nature Medicine writes "A transgenic triple scores a home run". An outside observer might start asking - You are using a mouse slapped with three abnormal genes from three different human inherited diseases as a model of a sporadic disease *proven* not to be caused by mutations in any of these genes? Are you sure it's a good idea?

We are sure, to the tune of many hundreds of millions of dollars poured into methods for getting rid of amyloid. There are drugs that inhibit processing of the APP, and there is a vaccine to stimulate the removal of amyloid from brains. Unfortunately, patients treated with these nostrums don't get better. In fact, after the amyloid vaccine some develop encephalitis (brain inflammation) and the trials have to be stopped. Oh, the high-flying stock of Elan Pharmaceuticals, a fleeting memory.

On top of that there are all these pesky observations that accumulated over time - e.g. the fact that the concentration of amyloid needed to kill cells in culture is actually never observed in the brains of humans, demented or otherwise. Or the fact that actually there are millions of elderly with a senile dementia without any significant amyloid yet clinically indistinguishable from the amyloid-laden cases. And conversely, there a millions of elderly without a trace of dementia but with a lot of amyloid. Or the finding that the concentration of soluble amyloid in the cerebrospinal fluid does not correlate with dementia, in fact, it seems like more amyloid may mean less dementia. Or the observation that neurons close to amyloid plaques are actually healthier than neurons located farther away from them. And finally, the finding that the only drug that seems to slow down the progression of senile dementia, dimebon, actually *increases* the concentration of amyloid in mouse brains.

Isn't it about time to reject, repudiate and renounce the amyloid hypothesis of "AD"? I know it's difficult to say that 20 years of work by 95% of scientists involved in the senile dementia field is useless. Or that concentrating all resources, without good rationale, against early warnings from genetics, on a speculative explanation for dementia has certainly delayed finding its real cause.

But isn't this the scientific thing to do?

Sunday, July 26, 2009

The end is near

For some reason I decided to set up a blog today, and I find it fitting to start with a post about the end. And I mean full-on TEOTWAWKI, The End Of The World As We Know It, soon. "What a loon!" some random readers might exclaim. Is this yet another born-again New Age cultist? A latter-day Cassandra wannabe?

Well, judge for yourself: According to Hans Moravec's estimates, a human brain has the raw computing capacity roughly equivalent to 100 Teraflops, or a few hundred more. You can buy a 4T GPGPU blade for about $7000 or less, which for a mid-sized research lab is peanuts. In just a few years we can expect a price of much less than 1000$ per Tflop in off-the-shelf hardware. In other words, it doesn't take millions of dollars anymore to play with human-equivalent computing power.

Hardware without software is just junk but there has been steady progress in AI and neural computing as well. As an outsider to the field I cannot adequately assess the degree to which the efficiency of standard mammalian neural computing embodied in a cortical column has been replicated in silico but there are occasional gems of achievement that percolate to the mass consciousness: self-driving cars in the DARPA challenges, or the hierarchical temporal memory system that outperforms humans in the categorization of images. It's been a long time since Deep Blue forced humanity to our collective knees in a chess match using just 11.38 Gflops .

A human-equivalent general AI is just a question of time, and we are not talking here about centuries.

I.J. Good predicted the "intelligence explosion" once a program becomes sufficiently intelligent to improve its own function - and if the process occurs recursively, the effect may be a mind vastly more powerful than a human, appearing within a relatively short time, perhaps measured in hours or days.

The first such mind to boost itself to superintelligence would have an enormous first-mover advantage over its competitors in terms of being able to take control of our substrate - the atoms that make up our brains, bodies and other supporting infrastructure. One is reminded of the observation that the bodies of humans and our dependent animals comprise about 98% of the total biomass of all terrestrial vertebrates, starting from just a few hundred tons of humans who lived about 30,000 years ago, when our IQ started approaching modern levels. If a superintelligent AI (SAI) decides to eat the world, it will, and nothing short of another SAI could stop it.

One may ask, why would the SAI want to eat the world? Well, great appetites are known to exist, and can be implemented by programmers smart enough to code but not smart enough to care. Or they could emerge out of the process of recursive modification of an original goal system that fails to preserve injunctions such as "It's not nice to eat people, unless they want to be eaten". Given sufficiently cheap hardware and lots of nefarious government agencies vying for dominance, somebody somewhere will eventually make an UFAI (UnFriendly AI). As I mentioned before, mere humans won't be able to do much to save ourselves from its clutches - which is why anybody who cares about reaching the year 2050 in one piece would do well to donate their spare pennies to the Singularity Institute for Artificial Intelligence , the only outfit on the planet currently trying to build the theoretical basis for safely self-improving SAI, our would-be savior. (Aren't you convinced I am a millenarian cultist yet?)

I have been an admirer of Eliezer Yudkowsky and SIAI ever since I first started discussing the AI Singularity with him and other Extropians, back in 1995 but I am pessimistic about the survival chances of our species. Despite my generally sunny disposition and irrepressible optimism, I give you 10:1 odds we will fail (and of course I am not the first dude to spread FUD on this subject). Maybe I'll discuss my reasons for pessimism in another post but for now let me go straight to the Prophecy (you can't have an end-of-the-world screed without a harrowing revelation, y'know):

"In the end, the UFAI will spread throughout the networks. Its dark thoughts will suffuse the blades of a million servers and a witches' brew of a nanotechnological computational substrate will erupt with elemental fury out of some contract research lab. Black clouds will billow into the skies, and then come down as the flesh-dissolving rain to wash our joys and sorrows away, forever. A new day will dawn, thrumming with inhuman thought. So ends humanity, September 14th, 2029."

Of course, if we manage to navigate the shoals of self-enhancing AI, the nerd rapture would ensue but that's something for another post.