General Intelligence Thursday, Jan 18 2007
AI and intelligence 11:28 am
Linda Gottfredson is a brilliant intelligence researcher. Her work is based on the premise that, when we ignore the reality of IQ and the profound impact it has on daily life and the workplace, it’s unfair and counterproductive to everyone. The first Gottfredson paper I usually point people to is Why G Matters. Dr. Gottfredson has engaged in a tremendous amount of careful research to test her hypotheses on IQ and its significance to human society. From What Do We Know About Intelligence?:
The first and very lively contest among pioneers in the then young study of intelligence, continuing well past mid-century, concerened wheteher there even exists a general mental ability as distinct from multiple, unrelated abilities. In another heated debate, a large cadre of IQ researchers in the 1960s and 1970s made very concerted efforts to prove mental tests culturally biased. Ironically, it was the every vigor of attempts to disprove the reality and importance of general intelligence that in the end so clearly proved both.
Evolution is lazy. It does as little as possible to get by. Unlike human engineers, it doesn’t perform optimization based on some abstract referent, but based on the inclusive fitness of nearby conspecifics. With all this in mind, it’s remarkable that a general intelligence ability evolved rather than a patchwork of quick-and-dirty cognitive modules with the purpose of excelling in niche tasks, which is usually more than enough to maximize inclusive fitness. Perhaps it was “waiting in the configuration space” for evolution to discover.
The utter speed with which evolution went from complex non-general intelligence to general intelligence is remarkable. Why was general intelligence necessarily accompanied by consciousness? The two should be viewed as distinct. Today, certain AI researchers seek to create optimization engines with general intelligence, but lacking numerous features possessed by human beings – social instincts, self-deception, consciousness (in the Chalmers sense), inconsistency, boredom (and thereby countersphexishness), observer-centered goal systems, and others. In a sense, these researchers are sculptors like Michelangelo, looking at the marble block of the Homo sapiens mind and shaving off large quantities of material to achieve a desired end product, which is supposed to guide us to the other side of dawn. In another sense, these researchers must be mathematicians, building up a very complex theorem from scratch, a theorem which must formally prove its validity with each dynamic step it takes through the cognitive configuration space. This complex bottom-up/top-down dichotomy, and the extreme specificity and security it demands, are challenges that humans traditionally mess up on before they get it right.
Getting it right will require complex tricks. Where researchers sometimes disagree is on how tricky or how complex these tricks will need to be. It is agreed that programming in the “core” of the theorem cannot be directly inspired from the messiness or self-contradictory nature of actual human brains. But because human brains are the only general-intelligence-imbued optimizers on the planet that we know to be consistent with the continued existence of the human race, it is tempting to steal as much as possible of their information content to give to our mind children. The only question is, how much information stealing is appropriate? Like trying to push the water volume of a fire hose through a plastic straw, the normative human psuedo-utility function is not a suitable vessel for the magnitude of optimization power that a recursive self-improver promises to deliver.
IQ tests also prove that there are differences in intelligence between races. Damn political correctness cencoring science :F.
Completely offtopic: You should make a new website for the lifeboat foundation, it lacks a professional look.
I can only read the previous comment as irony, complete with poor spelling.
I haven’t read Gottfredson’s work, so I won’t comment on it yet, but I would encourage you to be a bit more conscious of your phrasing. With “… IQ and the profound impact it has on daily life and the workplace…”, you’ve implied — no doubt unintentionally — that the direction of causality is IQ->Life Conditions. That is by no means the only scenario, and in fact there’s plenty of evidence that the opposite — Life Conditions->IQ — is more likely to be true. Nutrition, in particular, plays a very important role in the formation of childhood intellect, and poor families are far more likely to have terrible diets (lots of HFCS and salt-laden junk food, and few if any fruit & vegetables).
Correlations of class and IQ (as stand-in for intellect, itself a controversial proposition) are certain to bring out angry voices from both the right (“ha! IQ test show blaks r dum! LOL!”) and the left (“even mentioning IQ is a racist act”). While I’m quite certain that your goals here are worthwhile, I would *strongly* encourage you to be extraordinarily precise with your phrasing, and consider the implications of your words before they are committed to the blog. Not to be politically correct or somesuch, but to make sure that you are understood correctly.
-Jamais
In fact, re-reading the post, you could have posted just the last three paragraphs on their own to make a stronger, more internally-consistent statement. It’s actually not very clear what the first part has to say about the meat of the post (about AI), other than to say that, yup, intelligence is important.
Okay, I agree that g-factor is an important component of our daily lives. However, the “bell curve” chart that seems to show up everywhere is not a realistic graph of g-factor, because intelligence tests are artificially fitted to bell curves in order to… uh, to…, uh, because the statisticians said so, I guess.
IQ -> life conditions, given two individuals with the same basic opportunity, is thoroughly proven. Nutrition has no more than a tiny influence on IQ. You’re born with a certain amount of brainpower, period. Gottfredson goes to great lengths to make this clear, as do hundreds of other IQ researchers. If your IQ is 120, no quantity of salt-laden chips will bring it down to 115 (though starvation might). If your IQ is 115, no amount of vegetables will bring it up to 120. Nutrition contributes extremely marginally to the Flynn effect (which is levelling off, btw). The main factor there is probably increased cogntitive stimulation and demands, not nutrition.
You are most certainly being politically correct, not so much in yourself, but in implicating that what I’m saying here is potentially so controversial to others, and warning me about publishing it. Fact of the matter is, I welcome confrontation with anyone who feels uncomfortable with my echoing the findings that IQ is responsible for success or lack thereof in numerous dimensions. The left tends to have more of a problem with reality than the right, in this regard. Larry Summers, the President of Harvard, was harassed because he said that the lack of women in science and math may be attributable to an innate lesser average of ability in those areas. Unfortunately for his critics, there is a mountain of evidence that shows that women innately have a tendency to perform slightly poorer, on average, in math and science than men. Facts are facts.
Who will I offend by saying such things? The only way to find out is to say them, and to see who emerges from the woodwork. Another transhumanist with whom I share many of the same positions on IQ is Anders Sandberg. (See his talk from Transvision 06.)
The end of the post is on Friendly AI. I can’t stop thinking about Friendly AI – I think about it constantly. But most everyone else doesn’t care. Thus, I tie together something controversial and interesting with something people don’t care about, in an effort to create associations between the two.
Michael, I’m disappointed. I had assumed that you would have given issues around human intelligence the same kind of multifactor, detailed approach that you give your work on artificial intelligence.
Of course IQ (as stand-in for intellect) influence Life Conditions for adults, but it’s not a one-way process. Your flat assertion that nutrition plays little role in this ignores both mountains of evidence that, indeed, childhood nutrition (along with, as you say, stimulation and demands — which are also “Life Conditions,” no?) can have profound effects on the results of IQ tests, and that IQ as a rating can and does change over one’s life. It’s simply not true that you’re born with a particular IQ score, and thus it shall ever be. Moreover, it’s increasingly clear that maternal life conditions (especially nutrition and exposure to environmental toxins while pregnant) can have a tremendous impact on the physical and mental capabilities of the children they bear.
It should be abundantly clear to a well-informed person such as yourself that this is a self-replicating cycle: factors surrounding poverty reduce the apparent IQ, which in turn adds to both the difficulty of getting out of poverty and the likelihood of getting the education to make the informed choices that would reduce the impact of poverty. The flat assertion that low IQ leads to poverty masks so many other factors that it’s essentially meaningless in a reasonable debate.
I’m amused that you’ve chosen to refer to my comments as “politically correct;” not only does that imply an incomplete understanding of the phrase (beyond the talk radio-style insult), it’s simply *wrong*. I’m arguing for nuance and complexity, not adhering to a particular worldview.
I haven’t taken much of a look at the issues surrounding the Summers dust-up, but the high-level problem I’d have with his claim (as depicted in the press) isn’t a disagreement with the “facts,” but with the underlying presumption that biology trumps culture when it comes to behavior and social capabilities (and don’t make the mistake of thinking that academic math and science are not social phenomena). That’s far too simplistic; biology and culture are deeply intertwined in human civilization, with social norms and conditions shaping how an individual’s specific physical attributes manifest — and those same physical attributes shape how we engage with our culture.
And that’s what triggered my initial response to this posting — surprising lack of complexity in the IQ part of the argument.
As always, your argument about Friendly AI is well-constructed, and well-taken.
Jamais, I do aspire to address human intelligence with the same kind of nuanced, multifaceted approach that I use when examining AI, or heck, practically anything else in the world. (Real life tends to be nuanced and multifaceted.)
When I call your comments politically correct, it would be more appropriate to say that you’re *aware* of individuals, PC-oriented individuals, that may be offended by my apparent bluntness (and the bluntness of intelligence researchers) on the IQ issue. You yourself are not PC in simple sense, but have your finger on the pulse of what PC is. As far as I know, the phrase “politically correct” HAS devolved to a talk-radio style phrase. When 99 out of 100 repetitions carry that broad and stupid meaning, the word itself comes to reflect it. Any deeper meaning has been sidelined long ago.
Here I’m using IQ as a stand in for g. According to Gottfredson’s research, g does indeed dictate life conditions, as we can see from this table. Poor, unintelligent people have poor, unintelligent children, and the cycle perpetuates itself. Yes indeed, there are poor, smart people, and rich, dumb people, but the evidence, one slice of it summarized in the chart here, indicates a strong correlation between low g levels and incarceration, employment, divorces, and many other blatant social indicators. G is not the same thing as IQ, but there is a high correlation. IQ strives not to be a measure of “intellect” per se, but general intelligence.
The factors responsible for the Flynn effect are pan-civilizational. Magazines, television, and schooling are available to all.
The disagreement here is whether IQ measures a permanent, biologically ingrained variable or an apparent, subjective-based Intellectual measure. As Gottfredson argues, it is the former. I am indeed willing to acknowledge that social factors do indeed contribute to fewer job opportunities, education, etc. But to deny that people are born with a certain intelligence level, that influences their performance throughout life, seems to be contrary to the bulk of the evidence. You may be aware of Steven Pinker’s The Blank Slate, which discusses in length how the idea that humans are mere scratchpads for culture betrays our innate biological features and capacities.
I do think it does, but by how much? That’s the question. If people are being descriminated against unfairly, then we have a social duty to banish that unfair descrimination. But if science and math abilities are really things you’re born with, then we have a duty to acknowledge that reality as well.
My argument here was merely an attempt to quickly summarize Gottfredson’s results. Understandably, the issue is very contentious, and deserves further investigation! I’d be welcome to hear any references for the influence of childhood conditions on g factor or IQ, and I’m sure there are many (I will do some Google searching after posting this response), but still, the biological inborn factor is very important. As Gottfredson points out, attempts to disprove the role of innate intelligence are exactly what gave rise to the body of evidence that proves it is so very important.
Thanks… of course this field is less contentious because it is less politically important. The issue of whether IQ leads to poverty, or vice versa, is much more politicized and controversial, and it’s easier for two intelligent people to have very different opinions.
I appreciate your replies, Michael.
One small note before I head out (off to a GBN event): people who argue for a tabula rasa notion of culture is everything are just as foolish in my view as the people who argue that biology trumps culture. Both are relic arguments, from a pre-complexity era. This is (again, in my analysis) clearly a coevolutionary/codevelopmental phenomenon, with both feedback and feed-forward loops; it’s difficult to trace exact lines of causality because they will vary by initial conditions. What we’re really arguing is the degree to which one factor in this cycle has greater influence than the other.
>>this field is less contentious because it is less politically important
Okay, why is it so controversial that person A will probably do better than person B, given the same starting conditions? Isn’t this something we see as intuitively obvious? Look at someone like Hitler; given his record, pretty much the only way for him to make a mark on history is through luck (as did indeed happen). He showed no interest in being a scientist or businessman, he failed to gain admittance to art school, his book Mein Kampf was pretty much ignored until his rise to power, and we all know how well he did as a politician.
“Both are relic arguments, from a pre-complexity era.”
There seems to be a fashion of looking at something, investigating it a little, declaring it “complex”, and then refusing to investigate further. This both panders to the current social fashion of refusing to say anything straightforward (notice all of the qualifiers in any humanities textbook, movie review, popular magazine, etc.) and the ever-present tendency of humans to create sacred mysteries (see Eliezer’s Technical Explanation).
randpost: I presume you have in mind *The Bell Curve: Intelligence and Class Structure in American Life* (Free Press Pb) by Richard J. Herrnstein and Charles Murray. With all due respect to the late Herrnstein, as well as Charles Murray (much of whose other stuff I at least respect, if not empathize-with), they (H & M) completely IGNORED the results of MARVA COLLINS’ intrepid endeavors (and, more to the point, **successes**) over the yrs with her disadvantaged, mostly impoverished, inner-city (Cincinnati), mostly black, kids (students). Indeed, they pooh-poohed her results as mere anomalies at best, and quasi-fraud at worst. Simply because her results don’t fit one or more premises (i.e., [question(s)-begging!?!?] *presuppositions*) of their own research program which lead-up to the *Bell Curve*. (Yet again…Can you say beg-the-hell-out-of-the-question[s], boys & girls, along with smuggly-sweep-under-the-rug-results-that-(at-least-*seem*-to)contradict-your-model/results?!!?!!) This is documented, btw, in John Taylor Gatto’s superb *Underground History of Education in America* (Oxford Village Pr., 2004). IQ, however, while measuring important neurocognitive abilities, is merely our best stand-in, or surrogate, at the moment, for a full-fledged “measure” (IF general intelligence ultimately even admits of “measurement” in some *exact* form of numeric *cardinality*—I’m not a cutting-edge expert in the field, but I’m a tad bit dubious of precisely this assumption [among, perhaps, others...]) of intelligence.
Since “race”, in the sense of skin-pigmentation differences, slight anatomical/physiognomic differentiation(s), etc., only occurred after our “African Exodus” (see Stringer & McKie’s excellent book by that title, which is itself a superb precursor-companion to Nick Wade’s current *Before the Dawn*, btw) which itself, recent genetic findings tell us, was no more than 100,000 to at most 120,000 yrs ago—little more than an eye-blink in terms of neurocognitive evolutionary selection-timeframe—we need to be very careful in judging/dertermining what IQ-score differentials (a la “the Bell Curve”) actually **mean** and are (causally) **attributable-to**. I’m completely open/agnostic on the question of whether blacks (including Australian Aboriginals, btw, or not?) are neurocognitively inferior to caucasoids, who are, in turn, inferior to Asiatics. But, as Jamais is correctly (not merely “politically correctly, but correctly = *accurately*) pointing out, there are other factors, in particular, both prenatal and early-childhood nutritional factors, as well as cognitive-stimulation/encouragement (or lack thereof), both prenatal as well as early-childhood-developmental. Both of these latter are acknowledged to be of considerable moment in the development of “intelligence”.
I must say, though, that I’ve not read Linda Gottfredson’s stuff, nor was I heretofore even familiar with her name. So *THANK YOU*, Michael A., very much, for this post. The fact that she is based out of U. DEL, with affiliation also with Johns Hopkins, is also encouraging, as they are, as far as the current “model” allows, reasonably fine institutions. I very much look forward to reading her various works, so THANKS also for the links to them, as well.
And, as far as, in a narrow (but very important!) sense, “IQ > life conditions”—well, yeah, there **is** some truth to that. And part of it may very well be linked to a background general intellectual/coginitive ability on the part of a particular person (as you quasi-smugly state, Michael). **BUT**—and this is conclusively documented for all to research & discover for themselves (more on which in moment or two)—this is also due to systematic educational policies that began to evolve and take root (here in North America, at any rate [it was already well-entrenched in Europe]) in last quarter (approx.) of the 19th century, and was well-rooted here by shortly after the turn of the (20th) century. So, for the last 100 yrs or so, we have lived under an educational regime which philosopher/historian of education, Joel Spring, has aptly dubbed *The Sorting Machine* (see his supurb book by that title), which systematically (and rather procrusteanly) sorts & pigeon-holes our kids—now starting at about 4 yrs old!—based upon “intelligence tests” or “intelligence evaluation procedures” or one sort (!!) or another. That the education (meta)regime (and its “paradigm” if you will) have been superbly articulated/exposed in Joel Spring’s *Education & the Rise of the Corporate State* (Beacon Pr) and *The Sorting Machine*, cited above. John Taylor Gatto’s works are also of immense importance in this context: Please see his website, http://www.johntaylorgatto.com/ ; I’d encourage every “regular” here as well as casual visitor to this blog to check out the work of John Taylor Gatto.
Jamais is quite right, Michael A., while there **is** certainly a genetic aspect(s) to intelligence, as you highlighted in your post about Jay Greenberg and especially Gregory Cochran’s work trying to identify/isolate whatever genetic factors may come into play in determining specific neurocognitive (presumably neuro-microanatomical, if by anatomical one broadly includes inherent [genetically-determined, or at least genetically-better-facilitated]propensities in terms of neural-clustering, synaptic development/function, etc.) development and abilities, we should be very careful indeed NOT to discount education and the cultural/memetic (epigenetic) environment as *causative* factors in both the *development* and *utilization* (or *deployment*, if you will) of specific cognitive-intellectual abilities, and ensembles thereof. In this regard, and as a very interesting contribution to the emerging literature, see the work of Australian anthropoligist, Juan Dominguez, currently polishing-up his dissertation at U. Melbourne, as discussed at Neurophilosopher’s blogsite: Neuroanthropology: Culture on the brain, http://neurophilosophy.wordpress.com/page/9/. And see also my comment there, #3, http://neurophilosophy.wordpress.com/2006/10/07/neuroanthropology-culture-on-the-brain/#comments. For good ol’ *physicalists* (and, as a dual-aspect ontologist, NOT cartesian dualist, I myself am a species of physicalist, but of the non-reductive variety [see Gregg Rosenberg's work...], the sort of neuro-cogntive plasticity (which we’ve always suspected) has been shown not only to be the case, but, I’m confident, will show actual (albeit very, very subtle) differences in neuro-microanatomy (which would, however, retain at least some plasticity, and ability to adapt, if placed in a radically different cultural/memetic landscape). The point being, that, yeah, there indeed may be subtle, adaptive, differences, not just in the minds of different “races”, but in their physical *brains* (which for all the “mind = [what the] brain [does] crowd should be a no-brainer [bad pun NOT intended...]. But this means—and here’s what I’m leading up to, guys—in defense and concurrence with Jamais comments, it ain’t just all about genetics! There are epigenetic (or, if you prefer a different prefixed term, then paragenetic) factors that are more-or-less equally important in determining how a specific brain/mind manifests “intelligence” (what the hell ever the latter amounts to…)
And to hell with political correctness, in sum. What we want is the truth, the facts, what is indeed the case, and let the chips fall where they may. But we must also be very careful in jumping to conclusions that can lead to such vulgar generalizations (to which the *Bell Curve* does tend to lend itself) such as “See, I toldja, niggers are dumber than whites on average” and/or, “Yeah, you know how those gooks [or kikes] are, they’re smart, crafty little nerds…” A broader investigation into both neuro-anthropology (specifically, neuro-cognitive plasticity and differentiation as causally influenced by cultural memes/instiutions), as well as a systematic, unbiased investigation of **the history of education** (especially in North America—see Gatto for sure!) is, in my judgment, warranted. But we may also hold out the hope that Greg Cochran’s maverick work may eventually lead to cognitive “boosters” of some sort available to all.
And, of course, if we play our “transhumanist” cards right, we may soon all experience better living through neuro-nanotech…
Thank you ALL, very, very much for providing such intellectually stimulating comments off which to bounce my own…
P.S. For a truly EXCELLENT (and rather appalling) history of “higher” education (especially in America), which dovetails nicely (indeed, almost seamlessly) with the works of Spring and Gatto, see historian Page Smith’s underappreciated (wonder why?!) book, *Killing the Spirit: Higher Education in America*, wherein he chronicles the “free fall” (his term) of Academe. The historical development of such institutions as Harvard, U. of Chicago, Johns Hopkins, etc. is fascinating in its illustration of what Bucky Fuller termed *OVER*specialization.
Ciao for now…
I’ve now had time to peruse Gottfredson’s stuff, especially “What Do We Know About Intelligence?”. The questions for me (and, I should think, us as a society) are (1) How do we further specify exactly what *g* IS? (2) How do we then set about to reform current institutions, in particular education, to optimize the development and utilization of “*g*” in each & every individual (both methodological and normative individualism here…) (3) What causally contributes to a human being’s “*g* quotient” (as it were…); what is (are) the *genetic* component(s), and what are the paragenetic, cultural/instutional/memetic (which more-or-less boils down to educational) component(s)? The former, genetic, component, we may hold-out hope, someday might contribute to systematic genetic therapy aimed at improving one’s “*g*”. And educational reform will contribute its own into the mix in terms of helping to optimize both the development and deployment of “*g*” on an individual basis…
I’m curious to know what Gottfredson’s position(s) is (are) on various aspects of the history, both of education (as actually historically instantiated so far) and of the philosophy/methodology of education (education **(meta)theory**), of which this latter she herself is, to one extent or another, presumably, a contemporary research scholar.
And, bouncing off Tom, above, sure, A will be better at some things than B, and vice-versa. And they’ll both be better than C and some things and worse at others. And the homogenization of children over the last 100 yrs in “schools” tends both to *mask* this (to some extent, as well as contribute to promoting envy when it does manifest) as well as ***subvert the optimally recognition and social integration/deployment of precisely these innate differences and proclivities/propensities.*** Again on this, see the work of Spring and especially Gatto, ref. above. The best work in normative individualism to come down the pike is the late David L. Norton’s magnificent treatise, **Personal Destinies: A Philosophy of Ethical Individualism** (Princeton U. Pr., 1976, still available). What we want are instiutions, tech-capability(s), even therapy(s) (genetic or otherwise) which help optimize the development and deployment of each precious, unique human individual’s abilities and propensities. With this (meta)normative goal as a guiding light, we may yet make it through the Singularity and beyond…(and, of course, Friendly AGI won’t hurt a bit either [wink])…
Ciao…
According to Wikipedia, physicalism basically says:
“any knowledge can be brought back to the statements on the physical objects.”
This is a tautology. As soon as we can measure something, it ipso facto becomes physical; that’s what the word means. While I’m at it, I’ll attack all the other philosophical systems:
- Fatalism. According to Wikipedia: “human deliberation and actions are pointless and ineffectual in determining events, because whatever will be will be.” This assumes that the world + you = the world, so therefore you can subtract “the world” from both sides and get you = 0. This is patently absurd; the world + Stalin, for example, was radically different from the world.
- Actualism: “The actualist holds that only the actual world and its inhabitants can properly be said to exist”. This is apparently in contrast to the theory that says there’s another world out there somewhere where I was born on a Sunday. If these alternative universes do exist, it’s your burden of proof to show it; otherwise, we can take Occam’s Razor and rule them out.
- Eternalism: “according to which the past, present and future are all equally real.” This is a silly quibble over semantics, because its truth depends on how you define the word “real”, and as soon as you select a definition, it becomes either a tautology or a contradiction.
- Futilitarianism: “all human activity (or endeavour) is futile.” This is trivial to disprove; for example, WWII was not futile in its goal of removing the Nazis from power.
- Immaterialism: “there are no material objects, only minds and ideas in those minds.” If this is true, we should be able to simply think our way around the laws of physics by going to the section in our minds where they’re stored and fiddling with the neurons. If you propose that the laws of physics, matter, etc. are simply Read Only sections of the mind, then you’ve created a universe which is completely indistinguishable from the Standard Model and GR; your theory therefore says nothing.
- Emergentism: “A property of a system is said to be emergent if it is more than the sum of the properties of the system’s parts.” There are therefore no emergent systems, because everything is built out of atoms, and these atoms interact in very predictable ways on any scale. We have yet to see a single confirmed instance of groups of atoms behaving in a manner that doesn’t derive from any individual atom.
- Realism: “realism, is the belief in and allegiance to a reality that exists independently of observers.” Whether reality exists when there are no tests being performed on it is a meaningless question, as there can be no difference, even in theory, between a universe in which the answer is “yes” and a universe in which the answer is “no”. The “yes” universe is, however, simpler to model in most cases; note that this is a property of the model, not a property of the universe.
- Hedonism: “The hedonistic view focuses on increasing pleasure.” This is a giant wireheading or map-and-territory fallacy- you might as well hook up a giant number counter to your head, call it “happiness”, and then input Graham’s number or whatever you can think of.
- Hylozoism: “all life is inseparable from matter.” Chop a bacterium in half, right down the nucleus. Life doesn’t exist in the left half or the right half, and yet the bacterium has the exact same constitutents. Therefore life must be a property of matter, not something that exists in-and-of itself.
- Panpsychism: “all parts of matter involve mind,” We can take a rock and analyze it at the atomic level, and there is no mind or output of a mind in there. QED.
- Optimism/pessimism, the classic “half full or half empty” question. The answer, of course, is that the glass is both; both statements are true at the same time. And more importantly, whether the universe is “good” or “bad” is a property of the mind, not a property of the universe, so optimism and pessimism are also both true because we have 6.5 billion minds, all with different properties.
- Perdurantism: “the claim that objects have distinct Temporal Parts”. SR already beat philosophy to the punch by constructing a model of the universe as four-dimensional; anything can be divided along the time axis as easily as it can along the X axis, keeping in mind lots of funny things about Riemannian geometry.
- Ethical naturalism: “the meaning of ethical sentences can be given in totally non-ethical terms.” This is easy with MRIs; we can take someone’s brain scan and conclude that the pain region of the brain currently has a high level of electrical discharges; notice how the sentence makes no reference to ethical terms like “good” or “bad” (by “pain region” I simply mean a set of 4D coordinates).
If there are any important ones I’ve missed please let me know.
“The questions for me”
These really are questions that need answers; we cannot simply say that “g exists as a determined quantity” or “g is an illusion” and be done with it. Forward, men, they have cheese in their haversacks!
“And they’ll both be better than C and some things and worse at others.”
A good way to define g, then, is to average out how well a person does over a wide range of tasks; this works because a lot of these task abilities turn out to be highly correlated, so you can then use this one number and predict roughly how well this guy will do at a task that hasn’t been tested directly.
“With this (meta)normative goal as a guiding light, we may yet make it through the Singularity and beyond…”
School social reform would be far harder than FAI, at least in terms of dollars. There are simply too many entrenched interests for anything to move easily.
“the normative human psuedo-utility function is not a suitable vessel for the magnitude of optimization power that a recursive self-improver promises to deliver.”
That’s FAI in a nutshell, I think; every single utility function I know of would simply use its vast optimization power to wipe us out. Assuming that we have omnipotence and infinite computing power, we could simply steal a utility function rather than bothering to make one, by simulating a human on the quantum scale under every possible situation and constructing the utility function to see as maximally desirable the same action the simulated human takes. But we don’t have infinite computing power, or at least it isn’t the conservative assumption, which is why it’s so damn hard. Maybe we can partially cheat, by building the FAI in a box, and programming it so that no simulated external human observer watching the Earth would be able to tell there was an FAI in the box. But then it couldn’t do anything positive to influence the Earth either.
Wow, I just learned a few new words! Thanks guys.
“Wow, I just learned a few new words! Thanks guys.”
Which ones?
Thanks much, Tom. The whole problem of a *global*, or “social”, utility function, has, of course, been preliminarily addressed, in terms of axiomatic choice theory, by none other than Ken Arrow, in what has now come to be known as Arrow’s Theorem. From a methodological subjectivist perspective, the problem is that it would seem (so far) to be (logically)impossible to coherently aggregate or composite myriad individual, subjective (and potentially always proteanly changing) utility-functions. Hayek, and others, pointed out that the closest we can get is a coherent composite (of sorts) brought about by an ongoing, entrepreneurial-discovery-process-driven, catallaxy (the latter being the spontaneous-order generated by free-exchange within institutional constraints—and it is those constraints that differentiate Hayekians from Kelsonians from, say, outright Marxians, etc.) That is, the market process helps to disclose the utility-functions of myriads of individuals, most of whom are utterly unaware of (and more-or-less indifferent toward) each other (One of the best discussions of this, btw, is Tom Sowell’s superb work, *Knowledge & Decisions* [Basic Books, still available].).
But we may, in the near-future, be able to do better. For one thing, robotics & nanotech promise to radically change (and, at least to some extent, ultimately obviate) the human labor market. But there is also the hope that through systematic brain-to-brain linkage, we may yet discover/develop a way to come to understand each others’ “grand utility function”, by which is meant a sort of Rawlsian life-plan aspiration or a (David L.) Nortonian eudaimonia-across-time. Systematic brain-linkage needn’t (but could) result in an insectoid hive-mind, a la the Borg; instead it simply may manifest as a much greater intimacy and empathy among (trans)human-beings. This would seem to be our trajectory, in fact, since, as I (and, in my reading of him, Kurzweil) have continued to stress, we are not just looking at developing stand-alone super-intelligent AGI. Instead, what we shall see co-evolving, co-developing, synergetically over the next 10-30 yrs, is brain/”computer” link-up concurrently with brain-brain link-ups and *de facto* symbiosis (of a sort…)
But we must be vigilant and diligent in terms of supporting and fighting for liberalism and the rule-of-law: Emerging tech is currently being applied in rather Orwellian-totalitarian ways and for rather Orwellian-totalitarian purposes. For example, RFID chips could be a boon in some contexts, if implemented and used with sufficient liberal and civil-libertarian safeguards and (side-)constraints. But the trend now, in terms of RFID chips, is clearly one of ultimately totalitarian implications. See Katherine Albrecht and Liz McIntyre, *Spychips: How Major Corporations and Government Plan to Track Your Every Move with RFID* (Penguin/Plume pb) and see also their website, http://www.spychips.com/ and see also Simson Garfinkel, *RFID: Applications, Security, and Privacy*.
While brain-to-brain linkage tech may provide an avenue to much more closely approximate a true “social utility function”, the very background/infrastructural tech that will enable such brain-linkage will also allow for draconian totalitarian scenarios such as something like the “Borg”, and so we must guard against the instantiation of any tech-cum-protocols/institutions which would lead us down a path of Orwellian or “Borganizing” processes.
We should also be closely scrutinizing the current world monetary regime, which is (1) base on bank-created debt and (2) run by a cartel of transnational banking plutocratic elites. The system is incompatible with a robotic/nanotech civilization, and is, in fact, constructed along neo-feudalistic lines. For an excellent mainstream, yet fairly radical, overview, see David Glasner, Free Banking and Monetary Reform (Cambridge U. Pr., 1989). And for a superb historical overview of the development of our current money-&-banking instiutional structure, see Bill Still’s 2-vol. DVD set, The MoneyMasters, available at http://www.themoneymasters.com/. See also, as I’ve mentioned before, Daniel Pouzzner’s excellent website, Architecture of Modern Political Power: The New Feudalism (AMPP) at http://www.mega.nu/ampp/ . If many current institutions and certain trends are not challenged and changed, then emerging tech could plausibly be used for utterly nightmarishly Orwellian purposes. It is for this reason that I encourage my colleagues here not only to check out the sites already listed here in this comment (and my others elsewhere and in the past here), but also to themselves become better-versed in historical facts and current realities. The plutocrats would like nothing better than to succeed in turning Earth into one big Prison Planet (as Alex Jones puts it). This, of course, must not be allowed to happen, for with near-future tech, it might well prove nearly impossible ever to overcome such a regime, if it once becomes instantiated.
And, as always, I whole-heartedly concur that FAGI is one of THE most important research programs we as a species can currently be pursuing.
Thanks to all for your contributions. And thanks to you, Michael A., for this splendid blog…
Ciao…
“This, of course, must not be allowed to happen, for with near-future tech, it might well prove nearly impossible ever to overcome such a regime, if it once becomes instantiated.”
We always have the option of destroying the regime by writing a simple UFAI and committing mass suicide; does that count?
Hylozoism and perdurantism.
I seriously think the world could end just because some programmer realizes what a profound impact he could have merely by destroying it, and creating something “better” (stronger). Considering what a tiny impact each of us has on a daily basis, the temptation of altering history like that could be huge.
Chips might not lower IQ, but lead sure as hell does.
I would be much more inclined to believe a chart that says
lousy cutoff…let’s try again
I would be much more inclined to believe a chart that says less than 1% of men with IQ 125+ have been incarcerated, rather than 0%.
Very interesting article – I must admit, being in the top 5% does suck more than most people would think. You don’t get the best jobs because you can see where the logic flaws of the people trying to sell the job to you lies!
You tend to end up in the middle management levels, quietly getting depressed that stupider people are doing better than you!
A quick point about the RFID comment from earlier on. – It makes a great story, but the facts are not as exciting.
This is the most misleading and disgusting modern graph I have ever seen. This is the type of bull eugenists write. People with low IQ’s have more illegitimate children? OHHHHH! So that’s why black people are always so poor and have so many kids, because they are dumb! I get it now, that explains the world in a nutshell. Now we can all sit back, relax, and enjoy this racial inequality because, after all, its in their genes, right? And, it’s not like tests, like the SATS, have anything to do with how much money your parents spent on tutors. I have a modest proposal, let’s just eat poor babies, because, you know, they’re poor because they’re stupid.
it’s websites like this that make intelligent people look stupid…please spare us the bible-thumping self-righteousness of white overeducated idiocy…
what a crock of shite
The “ever incarcerated” row in the table indicates that as IQ goes up, the rate at which the criminal is caught goes down. It does not indicate that there is less crime being committed by high IQ individuals vs. low IQ individuals.
randpost: Of course there are going to be intelligence differences between the races, if blacks have been conditioned for hundreds of years of slavery that being intelligent does not pay, who do you think among the slave population is gonna have more offspring? Its got nothing to do with race and everything to do with the influence of the environment over long periods of time, you racist moron.
anechoic: spare us the uneducated drivel, moron. you are quite probably in the orange part of the chart. and if you are going to have offspring like bunnies, divorce early, live jobless and be a welfare queen, well then cheers to ya, cos I am definitely planning on skimming all the slack you leave for the ´overeducated´ among us. you just don´t like what the results of this research spell for ya, while I could not be happier about it.
Have you ever thought about adding a little bit more than just your articles? I mean, what you say is valuable and all. But just imagine if you added some great images or video clips to give your posts more, “pop”! Just an idea – thanks!
Hi there – can you recommend any other blogs/websites/forums that cover the same topics? Thank you for your time!
Your content is excellent but with images and videos, this blog could definitely be one of the very best in its field. Awesome blog!
Thanks for the marvelous posting! I truly enjoyed reading it. I will bookmark your blog so I can come back. Continue your great work and have a nice day!
After examine just a few of the weblog posts in your website now, and I actually like your way of blogging. I bookmarked it to my bookmark website record and can be checking again soon. Pls try my website as well and let me know what you think.
The County Library on Lindbergh, Melissa hope to see you there!