Contribute to our
AIShield Fund to make this program real.
This is an ongoing program so you may submit suggestions to
programs@lifeboat.com.
Lifeboat Foundation AIShield
By
Dan Berleant and other Lifeboat Foundation Scientific Advisory
Board members.
OVERVIEW
To protect against unfriendly AI (Artificial
Intelligence).
We support the
Friendly
AI proposal
by the Singularity Institute for Artificial Intelligence.
BENEFITS
Every dollar contributed towards the creation of Friendly AI will
benefit
an almost uncountable number of intelligent entities because of a domino
effect. A wide range of problems that face mankind, from disease to
hunger
to energy supplies, will benefit from this research.
RISKS
Risks from AI stem from two basic premises.
- The singularity is on its way. Soon (as noted by Vinge, Moravec,
Kurtzweil, etc.). Therefore, current models of how AI affects society
will become obsolete and we can only guess about what comes after the
singularity. In a practical sense that's what "singularity"
means.
- Murphy's law: if something can go wrong, it will. Therefore, we
need to be concerned about whatever plausible dangers we can guess at
that might occur after the singularity. We can't really ascribe
probabilities, high or low, to the dangers because we don't know enough.
We don't know enough because after the singularity models will tend to
break down. So, we need to be creative and simply try to guess all the
dangers we can...and protect ourselves from them.
Risks from AI may arise from the HCI (human-computer interaction)
paradigm that takes place. Such paradigms include the following, and
more than one of them might occur simultaneously.
A. The tool paradigm. In this paradigm, AI will serve humanity as a new
kind of tool, unique in part due to its post-singularity
power.
B. The prosthesis paradigm. Here, AI will gradually integrate with the
human body, producing cyborgian "people" with qualitatively greater
capabilities than regular people.
C. The competition paradigm. According to this view, robots will
ultimately have their own agendas which would most likely conflict with
ours.
Here are risks deriving from these paradigms.
A. Risks from the tool paradigm. These risks include the following. AI
could be embedded in robots used as soldiers for ill. Such AIbots could
be ordered to be even more merciless than ordinary soldiers. AIbots
could eliminate the emotional need for individuals to be social
organisms (consider e.g. Asimov's "spacer" colonies) leading to social
collapse. AIbots could reproduce ad infinitum, mining the ecosystem as
directed by humans until it collapses, doing so a order of magnitude
more efficiently than humans are already doing so ourselves. AIbots
could make it unnecessary for humans to work, leading to a species not
constrained either genetically or culturally to do anything useful,
resulting in deterioration of the race until deterioration is total.
B. Risks from the prosthesis paradigm. Risks from this paradigm include
these. The aging and breakdown of the human body might be substantially
delayed by replacing failed components, leading humans to get
progressively more mechanized instead of the current tendency to get
progressively older. Something approaching immortality could cause
population pressures that might destroy the ecosystem, lead to
widespread genocide, sensitize the species to devastating epidemics,
replace the younger and possibly more fecund elements of humanity with
cyborgs unable to reproduce or unable to genetically change, install
dictators who persist indefinitely, etc. Integration of AI and
supporting mechanisms into the human body could lead to an end to
natural selection based on health, until humans are so dependent on high
tech, artificially intelligent medical support to stay alive and
functional that there is no way back to being non-cyborgian. Cyborgians
could become progressively mechanized until the biological component
becomes a functionally unimportant component, at which point we will no
longer be human and may be said to have ceased to exist.
C. Risks from the competition paradigm. These risks are a perennial
favorite of apocalypse-minded sci-fi authors. The robots make their
move. Humans run for cover. The war is on, and it's them or us, winner
take all. Alternatively, the takeover is so successful that humans can
do nothing but hang out waiting until the AIbots eventually roboform the
earth to make it suitable for them but, as a side effect irrelevant to
the AIbots, unable to support higher biological life (oxygen is bad for
robots, so they'll get rid of it). A third possibility is that
nano-ai-robots (nanaibots?) take over, creating the nano-nightmare "gray
goo" scenario in which the gooey little bots destroy not only the
ecosystem but even invade human bodies for their own purposes
(germbots?), besting our immune systems and possibly sending us all to
another plane of existence.
RESOURCES
First Armed Robots on Patrol in Iraq by
Noah Shachtman, Wired Magazine - August 2, 2007.
Of Rats and Superempowerment by John Robb, Global Guerrillas
-
March 4, 2008.
VIDEOS
Actroid DER2 fembot just released by Japan's Kokoro.
Dancing robots (worth about $4 million in total).
Einstein Robot.
| |
