I think you're thinking in too limited ways about the physical tasks, simulated
or embodied - although that may be the fault of my definition.
You don't, I would suggest, program AGi machines primarily, say, to run mazes.
You program them primarily to move towards "goals" by moving around "obstacles"
- very generally. Mazes are the particular form of obstacles and whatever
prizes they contain are the particular form of goals, that you initially teach
the machines.
Also you give them general goals as animals have like "food" or "fuel" which
can in principle be instantiated in an infinite variety of particular forms -
so they can always be open to adding new forms to their motivations.
You also I guess (I'm thinking aloud here) give them a drive to move around and
explore - possibly for its own sake. If they have a drive to explore new
territories beyond those of mazes for new forms of food, there is obviously
potential to develop a set of very different activities.
You also give them (thinking further aloud) the capacity to form subgoals -
(but Gawd knows immediately how to do it) - so that certain subgoals and
subactivities, can become primary - as play eventually becomes in animals.
Also visual or sensory exploration of the environment ("sightseeing") could
become detached as an activity in itself, not just always in service of getting
through the maze to the food. Especially if the robot has free time on its
hands.
I think any AGI machine will have to have a mind structured by a tree of
several levels of generality and particularity - and be capable therefore of
thinking in very general ways of "goals", "moves", "obstacles," "food" "paths"
etc. - and continually adding more and more particular examples of each. (Ben
& I seem to agree here) And that tree is key to its adaptivity.
----- Original Message -----
From: James Ratcliff
To: [email protected]
Sent: Wednesday, April 25, 2007 9:49 PM
Subject: Re: [agi] The University of Phoenix Test [was: Why do you think your
AGI design will work?]
So basically it appears there are 3 areas of possible AGI Test Applciations:
1. Non-Natural Language Required Tasks.
- Question Answering (little comprehension)
- specific programmed tasks
- Most AI work, vision, sound, etc
2. Natural Language Tasks (advanced parsing and comprehension)
- conversations, Turing Tests, advanced interactive tasks.
3. Embodied Tasks
Either virtual or robot.
movement and smart interaction with the environment, to complete tasks.
Can an AGI be built at all that first into area 1?
It seems too limited to me (rough delineated definitions above granted)
I think any system in one is written for a specific task, chess, maze
solving, etc.
So does it require 2 and/or 3 and which is better.
#2 I think would qualify as an AGI, if it could talk and pass the Turing test
and converse. How useful is it though, what kind of system does it create?
Only expert systems for a domain, or chatbots?
#3 Video game characters or robots are an AGI as well, but how limiting are
they if they dont have complex language skills to be given and learn ever
increasingly complex tasks to do.
Mike Tintner <[EMAIL PROTECTED]> wrote:
Nobody came back on my suggestion for a much simpler AGI test. Let's call
it
the Neo-Maze Test.
You program a robot rover or simulation robot with flexible rules to run
fairly basic-type mazes, including mazes with multiple solutions.
The test is then whether it can run very different kinds of neo-mazes (to
be
set by other programming teams) - let's say, mazes with holes-in-the-wall,
large open squares, roundabouts, spaghetti junctions, underpasses, ramps,
traffic lights and traffic problems etc. etc. - for which its rules have
NOT equipped it, but which it does have the basic raw capacity to run.
This will be a test of its ability to fundamentally adapt its approach to
an
activity to fundamentally altered environments - to adapt its rules, steps/
moves and recognized paths to its goals. That, I reckon, is the primary
requirement of AGI. (Animals can do it).
From there you can move to ever more complex, and higher-level activity
tests. But setting your first goal as passing a language test is to my mind
absurd.
P.S. Another comparable test would a Video-Game Test, where some
gameplaying
agent that is programmed to play something like Pac-Man, will have to adapt
to fundamental variations on that game, with again radical alterations to
the game's maze structure and the type of predators it must deal with, and
treasures it must find.
Or, thinking of the link screen on the Novamente site, you could have
simply
a Building Navigation Test - a robot or agent programmed to negotiate
fairly
simple rooms with simple furniture arrangements, and fairly simple
corridors - will then have to megotiate ever more complex rooms and
building
corridors, ideally the full range of modern architecture. That surely would
be a highly practical test with highly commercial applications.
----- Original Message -----
From: "Derek Zahn"
To:
Sent: Wednesday, April 25, 2007 2:40 PM
Subject: Re: [agi] The University of Phoenix Test [was: Why do you think
your AGI design will work?]
> Ben Goertzel writes:
>
>> I don't think there are any good, general incremental tests for progress
>> toward
>> AGI. There are just too many different potentially viable approaches,
>> with
>> qualitatively different development arcs.
>
> Nevertheless, I wish somebody would try to specify some that are perhaps
> not completely general. Without that, the only way to determine whether
> any
> progress at all is being made is by an analysis of internal structures --
> pointing to
> a data structure and making claims about its meaning. And the developers
> of
> the system are notoriously bad at doing this as they are too emotionally
> and
> intellectually tied to the work.
>
> I wonder at what point our ancesters became "generally intelligent"? Were
> humans
> of 10,000 years ago generally intelligent? If so, why did it take them so
> many billions
> of person-years to develop the most rudimentary capabilities that we seem
> to expect
> our artifical general intelligences to breeze through effortlessly? I
> suppose the
> real test is at what point an individual from the past would be able to
> pass the Turing
> test (or some similar thing) if born into our present world and educated
> like we were
> and I doubt any scientists could make any confident guesses about that.
>
> I think that figuring out a good working definition of general
> intelligence and
> demonstratable intermediate steps is the single most important missing
> piece
> of the endeavor.
>
> -----
> This list is sponsored by AGIRI: http://www.agiri.org/email
> To unsubscribe or change your options, please go to:
> http://v2.listbox.com/member/?&;
>
>
>
> --
> Internal Virus Database is out-of-date.
> Checked by AVG Free Edition. Version: 7.5.463 / Virus Database: 0.0.0/0 -
> Release Date: 00:00
>
>
-----
This list is sponsored by AGIRI: http://www.agiri.org/email
To unsubscribe or change your options, please go to:
http://v2.listbox.com/member/?&;
_______________________________________
James Ratcliff - http://falazar.com
Looking for something...
------------------------------------------------------------------------------
Ahhh...imagining that irresistible "new car" smell?
Check out new cars at Yahoo! Autos.
------------------------------------------------------------------------------
This list is sponsored by AGIRI: http://www.agiri.org/email
To unsubscribe or change your options, please go to:
http://v2.listbox.com/member/?&;
------------------------------------------------------------------------------
Internal Virus Database is out-of-date.
Checked by AVG Free Edition.
Version: 7.5.463 / Virus Database: 0.0.0/0 - Release Date: <unknown> 00:00
-----
This list is sponsored by AGIRI: http://www.agiri.org/email
To unsubscribe or change your options, please go to:
http://v2.listbox.com/member/?member_id=231415&user_secret=fabd7936
