Technological Singularity and Friendly
Artificial Intelligence: Computer
Programs, Minds-in-General (Artificial Intelligence), and the Human
Brain At some point, brains made out of
components besides biological neurons will exist. What will that
mean? Will an intelligence that computes with silicon think at the
same rate at one that computes with proteins? Heck no. And that's
just the beginning of the huge differences between human brains,
computer programs, and artificial intelligences that you should
know about. Although no true Artificial Intelligence exists yet,
it seems smart for us to more closely examine the ways in which
they will be different from humans, so we can formulate policies
for its arrival sensibly.
Cooperating with New Intelligence (Artificial Intelligence)
As Moore's Law continues to accelerate and cognitive scientists are
beginning to uncover the underpinnings of general intelligence, the
creation of true Artificial Intelligence is becoming a real, near-term
possibility. Since true AI would be able to think at speeds millions
of times faster than humans, custom-design new cognitive components,
and recursively self-improve, it's important for humanity's well-being
that the first AI built possesses a robust and self-reinforcing benevolent
morality. If we fail to prepare and just let the chips fall where
they may, the result could be a recursively self-improving AI with
no empathy for humans. Beyond
Defaults: Transhuman Intelligence (printable)
One of my personal catchphrases is: "never expect the default
to be optimal." Just because it's there doesn't mean we should
accept it. The default is just what happens to be around when we get
here, but why not try doing better? We apply this reasoning very well
in certain contexts and very poorly in others. For example, some people
assume that Homo sapiens is an optimal species, or, complain
about Homo sapiens without offering potential improvements
or alternatives. The latter is what this essay is mostly about.
Forecasting
Superintelligence: the Technological Singularity (offsite) Oxford
philosopher Nick Bostrom defines a "superintelligence"
as "an intellect that is much smarter than the best human brains
in practically every field, including scientific creativity, general
wisdom and social skills". At what point will the creation
of such an entity become technologically possible? Must we know
every detail about how the human brain works before we are able
to build an artificial intelligence? Is superhuman artificial intelligence
really that much more difficult than human-level artificial intelligence?
Futurism at the intersection of computer science and cognitive science.
Deconstructing
Asimov's Laws (offsite) This is my contribution to AsimovLaws.com,
the Singularity Institute's web project whose release was timed
to coincide with the appearance of the "I, Robot" film
in theatres. "I, Robot" featured a bunch of robots spontaneously
running amok when their programmatic constraints (Asimov Laws) collapsed.
Sadly, there are people who still take Asimov Laws seriously. People
that don't take them seriously are really proud of themselves at
seeing the laws as fictional (as Asimov did), but their proposed
alternatives are often equally useless. The study of AI morality
needs to be turned into a scientific discipline, built on the foundations
of cognitive science and information theory.
Singularity activism:
Who
Are Technological Singularity Activists?
Long paper describing what "Technological Singularity activists"
are. Singularity activists are people trying to create a smarter-than-human,
kinder-than-human intelligence, because we view the human mind as
one particular type of mind, a mind potentially subject
to improvement along intellectual, emotional, and moral axes, among
others. We figure that transhuman intelligences, properly constructed,
will be far more capable of coming up with ideas to help
people with their problems than we humans are. This is not worship
of some speculative higher being, but straightforward common sense.
We are not so arrogant to assume that the human species represents
some ideal of inferential, moral, or intellectual optimality. We
must reach beyond the default human template for creative solutions.
Who
Cares About the Technological Singularity?
A list of prominent scientists and futurists concerned with the
arrival of the so-called "Technological Singularity".
Includes familiar names such as Stephen Hawking and Sir Martin Rees.
Considering the controversy surrounding futurist concepts such as
the Singularity, I consider it worthwhile to point out people who
take it seriously. These are straight-faced academics seriously
discussing topics like superhuman intelligence and the Technological
Singularity. Former flights of fancy are now acquiring scientific
foundations!
Achieving the Technological
Singularity (printable)
The Technological Singularity offers a huge humanitarian opportunity,
allowing us to tap the power of transhuman intelligence and recursive
self-improvement and direct it towards good ends. A successful Singularity
would signify our safe passage through possibly one of the most
difficult transitions in human history. There is no better way to
maximize human freedom, open up as many possible opportunities,
and heal as many ills as through the creation of something inherently
benevolent and significantly smarter and faster-thinking than we
are. The Singularity could be humanity's biggest and last life-threatening
challenge.
A Short Page Introducing
the Technological Singularity (onsite)
This is my example of a page I think that everyone concerned with
the Technological Singularity should have. It says a little bit
about myself and how I got into futurism, followed by a description
of the possibility of transhuman intelligence and some links. The
Singularity is an idea just starting out - much like nanotechnology
was in the 80s. In order to accelerate awareness about this important
idea requires that we get the word out. This page took two minutes
to make, please consider making your own.
Who are Singularitarians?
(onsite) A shortened version of "Who Are Technological Singularity
Activists"? One day greater intelligence will be technologically
possible, and we can't assume it will be in the form of an enhanced
human being. It could arrive in the form of an AI, and if it does,
we want to make sure that that AI has the goal system complexity
necessary to understand and value qualities like empathy, peace,
compassion, etc.
Writings on life
extension:
Immortalist Utilitarianism:
the Value of Life Extension (onsite) What makes life worth
living? Why do some people eat healthily and exercise to extend
their lifespans? Why do some people advocate cryonics? Can we quantify
moral desirability? If 80 years of happy, fulfilling living is desirable,
would 800 years of happy, fulfilling living be 10 times more desirable?
This document examines the relationship between immortalism and
utilitarianism.
Fountains
of Youth: Hacking the Maximum Lifespan (offsite)
This is a piece I did with R.U. Sirius, cyberpunk pioneer and former
editor of the magazine MONDO2000, the precursor to WIRED magazine.
R.U. interviewed myself and Cambridge biogerontologist Aubrey de Gray
for the website "The Neofiles", sponsored by Life Enhancement
Products, a nutritional supplement company. Various questions about
progress in the control of aging on cellular and genetic levels, and
the philosophy of immortalism. I even get to mention nanotechnology
and cryonics! Objections
to Immortality: Answering Leon Kass (printable)
An article I did for the nonprofit Immortality Institute in February
2003, responding to the latest outbursts of the government bioethicist
Leon Kass. Mr. Kass is basically pushing to outlaw all life extension,
and if he scares enough people, it could actually happen. Not only
could this kill billions who would rather live than die, but it
damages the overall intellectual environment which futurists are
operating within, stifling the fruition of further beneficial technologies
and the corresponding freedoms they would unlock (Virtual reality,
nanotechnology, space programs, etc.) However, Mr. Kass's specific
stance against the prospect of indefinite lifespans may help convince
bystanders of its near-term technological feasibility, perhaps resulting
in a net positive. It's really hard to say. This article was also
published in the Extropy Institute's online magazine "Extropy".
Writings on molecular
nanotechnology:
My Position
on Nanotechnology Administrative Policy (pdf)
(printable) At some
point in the next decade, possibly as early as 2006, someone will
create a molecular assembler - a submicroscopic programmable robot
with the ability to make copies of itself. This will quickly lead
to nanofactories, desktop appliances capable of synthesizing practically
any product you could possibly want based on software specifications,
for only the cost of raw materials. If these nanofactories are administrated
and deployed responsibly, we could greatly benefit from their use,
but if not, we could all be dead.
More Dangers From Molecular
Nanotechnology (printable)
This page was written as an add-on to the Center for Responsible
Nanotechnology's page on the subject. Nanotechnology, when it arrives,
will reshape every facet of human society. It will be a tidal wave
of new technological abilities for humanity - and these abilities
will be used both for good and for evil. In order to prevent abusive
or ignorant applications of such an advanced technology requires
that we visualize the dangers and prepare for them.
Writings on futurism
and transhumanism:
What is Transhumanism?
(onsite)
Short document summarizing transhumanism, introducing existential
risks and the Technological Singularity. Transhumanism is an extension
of the popular philosophy of humanism, a worldview that places human
beings and human abilities in central focus. Ethics in humanism
derive from what individual humans care about.
Accomplishments
of Transhumanist Organizations (onsite)
A series of overviews of the concrete accomplishments of transhumanist
organizations and individuals associated with them. I feel that
having a list of these accomplishments would be a good idea, because
it's convenient to see them all in one place, plus it will help
encourage people to get involved. Nine organizations reviewed so
far. Please send me an email if I missed anything.
Future Shock Level
Analysis (onsite)
Analysis and expansion of a popular classification scale for various
types of futurism. The "Future Shock Levels" scale helps
one make decisions about which futurological concepts to present
to which audiences; for example, you wouldn't want to talk about
nanotechnology to someone who isn't familiar with genetic engineering
or nuclear fusion. The interesting thing about the Shock Levels
is that there seems to be a fairly regular migration from one end
of the spectrum to the other, although this is not entirely certain.
If this phenomenon does not hold for the whole scale, it at least
holds for a major chunk of it.
Transhuman Intelligence
and Altruism (onsite)
A relatively short piece on transhuman intelligence and altruism.
A rant of sorts - winds through several related topics. I'm generally
arguing that transhuman intelligence and altruism really need to
go together, that altruism is a real thing, and how much better
the world could become if it were saturated with altruistic transhuman
intelligence. Whether or not this eventually comes to pass, of course,
will depend upon our actions in the present.
Tips for Transhumanist
Activists (onsite)
Ah, the wonders of complaining about how your friends and associates
should be acting. That's part of what this paper is about! Meant for
the audience of people who are attempting to be transhumanist activists,
this paper includes a few suggestions for making the most out of your
presence. Don't be a lazy cheerleader or ride on the coattails of
others; you too can be the "ideal success story"! (If you're
wondering what on Earth a "transhumanist activist" is, or
why they should even exist, check out the link to the Transhumanist
FAQ, three links down.) Transhumanist
FAQ (offsite)
I didn't actually write this. Oxford philosophy professor Nick Bostrom
did, and I just put it here to get you to go read it. I'm trying
to show that respected academics and intellectuals are spending
their time considering (and pioneering) transhumanist issues, because
they're genuinely important. This particular piece is about the
philosophy/technological movement I'm a part of, "transhumanism",
which originally descends from humanism and continues to rapidly
acquire sympathizers and adherents. Notice how transhumanist ideas
have been steadily seeping into the mainstream for years now.
Writings on reason and rationality:
A Concise Introduction
to Heuristics and Biases (onsite)
Rather than approaching each problem from the ground up, the human
mind uses heuristics - rules of thumb - to confront and solve problems
in a cognitively economical fashion. Unfortunately, while these
heuristics are very effective for solving certain problems, they
often break down. Since these biases are so deep-rooted and practically
universal, experience - even decades of experience - fails to remove
them. They must be isolated empirically and formally debiased.
Online interviews:
Happily
Ever After: An Interview with Michael Anissimov (offsite)
A delightful interview I did with Phil Bowermaster of Speculist.com,
a great futurist blog. And what a cool title for my first formal interview,
none the less! Phil has been working in the fields of IT and Telecommunications
for the past 15 years, where he spent 4 years in Malaysia and helped
lay a variety of communications networks. The interview is on extreme
life extension, transhumanism, AI, and the Singularity.
Seven
Questions with Michael Anissimov (offsite)
Another interview with the Speculist. Short set of questions about
the future. The answers highlight my opinions regarding the future;
I strongly emphasize unknowability, transhuman technologies, greater-than-human
intelligence, and the Singularity. (The usual.) Good intro to my
worldview, which resonates strongly with the worldviews of fellow
futurists Eliezer Yudkowsky and Nick Bostrom.
Personal miscellanea:
Statements I Agree
or Disagree With (onsite) Opinions, so many opinions. Distinct
opinions are a big part of our individuality. Even if they can't
always be completely correct, I advocate the communal sharing of
deep opinions, and oppose the withholding of opinions for PR or
embarassment reasons. Interesting fact: if we never used the sincere
opinions of others as "ammo" to degrade them or their
ideas during heated arguments, then people would be a lot more open
about sharing them in the first place. Be tolerant of each other's
deep, sincere opinions! Otherwise all conversation just turns to
small talk.
My Current Ideological
Platform (onsite)
Everyone has an ideological platform, whether we pay attention
to its existence or not. Rather than letting my platform pull at
me from the mysterious darkness, or keeping it as a secret, I like
throwing it out into the unshielded open, so it can be criticized
and discussed. My current platform revolves around some nontrivially
abstract and complicated anchors, though. Enjoy!
Quotes I Like (onsite)
Quotes I like, but don't necessarily agree with entirely. Most
of them I agree with, though. Example: "Science itself, therefore,
may be regarded as a minimal problem, consisting of the completest
possible presentment of facts with the least possible expenditure
of thought." -- Ernst Mach. Send in any quotes if you think
of any I might like!
World Peace Through World
Domination (onsite)
Ever wanted to wipe out poverty, disease, war, and aging, think
you may know how, but are worried that pesky governments or international
mega-corporations will get in your way? Look no further, "World
Peace Through World Domination" is the strategy for you.
Intelligent People with
Interesting Ideas (printable)
Throughout the course of my life, I've been influenced profoundly
by a series of contemporary thinkers. I believe that the possibilities
of advanced technology totally change how we should think about
philosophy, planning, science, politics, et cetera, so only thinkers
who understand advance technology have truly up-to-date ideas. Here
they are!
Ten Reasons to Do Stuff
(onsite)
Why try to live as long as possible, enhance human intelligence,
develop safe nanotechnology, safe artificial intelligence, or learn
more about science and technology? This document lists 10 reasons
for each.
Why Intelligent People
Fail (onsite) An interesting question, isn't it? Even though
intelligence in IQ sense is a powerful indicator of real-world success,
sometimes certain variables intervene to prevent a high IQ person
from reaching their full potential. Since I'm sure all my readers
have extremely high IQs, they might find this list interesting.
|