Subject:
|
Re: Design
|
Newsgroups:
|
lugnet.robotics
|
Date:
|
Wed, 7 Dec 2005 01:45:10 GMT
|
Original-From:
|
steve <SJBAKER1@stopspammersAIRMAIL.NET>
|
Viewed:
|
1113 times
|
| |
|
|
PeterBalch wrote:
> > dan miller wrote:
> > Why the negativity about constrained or simulated
> > worlds? We are not Nature; we are not constrained to work in a design
>
> space
>
> > that never changes.
>
>
> I think that the whole AI community was chastened by SHRDLU (or, at least,
> 99.9% of them were). It seemed like a huge success: "Why did you clear off
> the blue cube?" "To put the red pyramid on top of it."
I don't have the dialog verbatim - but something like:
Human : Are there any steeples in the box?
SHRDLU: I don't know what a "steeple" is.
Human : Define a steeple as a cube with a pyramid resting on
top of it.
SHRDLU: OK
Human : Are there any steeples in the box?
SHRDLU: Yes, there are two of them.
...that's pretty cool!
> It looked like a
> real human-language interface. All it needed was to be scaled up. A new,
> faster computer would do the trick.
>
> Nowadays, you won't find a single AI worker who thinks that. The received
> wisdom is that SHRDLU was dealing with a toy world in a way that didn't
> scale well.
SHRDLU was pretty amazing for it's time...but like a lot of things
in AI - once someone's done it, it's not AI anymore. Sure, SHRDLU
wasn't intelligent - and even within it's limited world, it couldn't
pass a Turing test - but as far as it went, I'd call it an important
step.
Within his little toy world, SHRDLU was pretty amazing - if you've never
played with it, you won't have felt the creepy feeling that there is
something very different going on in there.
It's not like the famous (or perhaps infamous) 'Eliza' which rapidly
shows up that it's not really working as advertised. SHRDLU really
was working - within it's admittedly limited domain.
I wonder if there are any SHRDLU re-implementations that would run on
a modern PC? I'd sure like to play with it some more.
I'm a little suprised that programs like SHRDLU aren't in active use
in environments of similarly limited scope. There must be lots of
applications for something with that level of intelligence over a
very limited domain.
> Only 1 in a thousand algorithms moves successfully from a toy world to the
> real world.
>
> It's time we admitted that it's a research paradigm that doesn't work.
> Hence the negativity.
Well, even if only 1 in a thousand makes it - there is progress - and
results that are useful along the way.
> (So whay am I spending time trying to write PlaniSim to simulate a "real"
> 2D world? 'Cos I don't learn from my mistakes :-).)
Cool! Planiverse lives!
|
|
Message is in Reply To:
 | | Re: Design
|
| (...) space (...) I think that the whole AI community was chastened by SHRDLU (or, at least, 99.9% of them were). It seemed like a huge success: "Why did you clear off the blue cube?" "To put the red pyramid on top of it." It looked like a real (...) (19 years ago, 6-Dec-05, to lugnet.robotics)
|
5 Messages in This Thread:
 
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
