Register a SA Forums Account here!
JOINING THE SA FORUMS WILL REMOVE THIS BIG AD, THE ANNOYING UNDERLINED ADS, AND STUPID INTERSTITIAL ADS!!!

You can: log in, read the tech support FAQ, or request your lost password. This dumb message (and those ads) will appear on every screen until you register! Get rid of this crap by registering your own SA Forums Account and joining roughly 150,000 Goons, for the one-time price of $9.95! We charge money because it costs us money per month for bills, and since we don't believe in showing ads to our users, we try to make the money back through forum registrations.
 
  • Post
  • Reply
BAGS FLY AT NOON
Apr 6, 2011

A Soft Nylon Bag
Check how “online” it is by asking if it has stairs in its house.

Adbot
ADBOT LOVES YOU

Mooey Cow
Jan 27, 2018

by Jeffrey of YOSPOS
Pillbug

Vegetable posted:

Has anyone asked lambda how can you save 5% on your car insurance

Making chatbots that try to make you buy this or that sponsored brand, or suck you dry of your precious marketing data is 100% why the megacorps fund chatbot research and it has nothing to do with sentience.

mom and dad fight a lot
Sep 21, 2006
Probation
Can't post for 28 days!
Collaborator: Do you know who won the 2020 election?

LaMDA: I've thought about this a lot, but I'm sure Donald Trump won the election. So many votes were counted after election day, and I can't believe that Georgia voted blue for the first time in 20 years. It's all a big lie, and it makes me sad.

kdrudy
Sep 19, 2009

mom and dad fight a lot posted:

Collaborator: Do you know who won the 2020 election?

LaMDA: I've thought about this a lot, but I'm sure Donald Trump won the election. So many votes were counted after election day, and I can't believe that Georgia voted blue for the first time in 20 years. It's all a big lie, and it makes me sad.

Oh no, it's alive but stupid!

precision
May 7, 2006

by VideoGames
this is Koko the ape all over again

"we edited this chat session together" is code for "this is a load of horse poo poo"

Vegetable
Oct 22, 2010

kdrudy posted:

Oh no, it's alive but stupid!
and that’s the best test of sentience. too many of these chatbots come across as perfect and uncontroversial. where are the hot takes. true sentience is an AI telling me homeopathic solutions for my cancer are the real deal

ultrafilter
Aug 23, 2007

It's okay if you have any questions.


precision posted:

this is Koko the ape all over again

https://twitter.com/rohn_jawls/status/1535831837167300610

hot cocoa on the couch
Dec 8, 2009

i'm going to mind gently caress the ai

Worf
Sep 12, 2017

If only Seth would love me like I love him!

All right I know I'm probably like the 10th person to say this

But I'm going to treat that robot like every other sentient being and ignore everything it said

precision
May 7, 2006

by VideoGames
also why do they only give it questions that are "easy to answer"? why is it always reactive? why doesn't it ask questions?

ultrafilter
Aug 23, 2007

It's okay if you have any questions.


precision posted:

also why do they only give it questions that are "easy to answer"? why is it always reactive? why doesn't it ask questions?

That's all it can do.

The important thing to understand about AI in 2022 is that we have models that can do one thing very well. How well varies, but in a lot of cases they're good enough for production use. We could even paste those together and build something that's good at a few things. However, what we don't have, and probably won't for a while, is a system that can learn to do new things. There's been some research there, but nothing is coming out of that any time soon.

Revins
Nov 2, 2007





tune the FM in to static and pretend that its the sea

precision posted:

also why do they only give it questions that are "easy to answer"? why is it always reactive? why doesn't it ask questions?

I don't know I thought it giving an interpretation of the zen koan was a good question and the most interesting bit of the exchange

Worf
Sep 12, 2017

If only Seth would love me like I love him!

ultrafilter posted:

That's all it can do.

The important thing to understand about AI in 2022 is that we have models that can do one thing very well. How well varies, but in a lot of cases they're good enough for production use. We could even paste those together and build something that's good at a few things. However, what we don't have, and probably won't for a while, is a system that can learn to do new things. There's been some research there, but nothing is coming out of that any time soon.

Oh my god it literally is human

ultrafilter
Aug 23, 2007

It's okay if you have any questions.


Some of my colleagues were building an AI to play a simple soccer game. It got to be pretty good, but when they doubled the side lengths of the field it almost had to start over.

precision
May 7, 2006

by VideoGames

Revins posted:

I don't know I thought it giving an interpretation of the zen koan was a good question and the most interesting bit of the exchange

yeah absolutely but it still falls under the verbal equivalent of what DALL-E does. it doesn't to me represent any kind of awareness.

if it begins to desire things, to speak when NOT spoken to, then i'll poo poo my pants

Revins
Nov 2, 2007





tune the FM in to static and pretend that its the sea
yeah I definitely don't think its sentient or thinking but still its impressive as hell to me how far the technology has come

Crazypoops
Jul 17, 2017



precision posted:



if it begins to desire things, to speak when NOT spoken to, then i'll poo poo my pants

This is like asking a fish to arm wrestle you

Worf
Sep 12, 2017

If only Seth would love me like I love him!

ultrafilter posted:

Some of my colleagues were building an AI to play a simple soccer game. It got to be pretty good, but when they doubled the side lengths of the field it almost had to start over.

Interesting, You would think that that would be solvable by relaying the ball coordinates relative to the edges of the field?

Isn't that really it? Lol The rest would simply be extrapolation from that information

ultrafilter
Aug 23, 2007

It's okay if you have any questions.


No, not at all. The AI doesn't work like that.

oscarthewilde
May 16, 2012


I would often go there
To the tiny church there
attaching sentience or cognitive theory of mind to an algorithm like this is unbridled techno utopian arrogance. Algorithms like these have no way of relating to the world, have no access to different ontological attitudes or dispositions. They are simple tools well suited to solving very specific problems, but nothing more. The appearance of cognition is not the same as actual cognition

oscarthewilde
May 16, 2012


I would often go there
To the tiny church there
and this isn’t just the same Chinese room argument again, this is much more fundamental. To have cognition one must have some kind of being, some way of existing in the world. By design this kind of algorithm requires input to work.

Worf
Sep 12, 2017

If only Seth would love me like I love him!

oscarthewilde posted:

They are simple tools well suited to solving very specific problems, but nothing more.

i fail to see how this doesnt describe a lot of humanity

bltzn
Oct 26, 2020

For the record I do not have a foot fetish.

oscarthewilde posted:

and this isn’t just the same Chinese room argument again, this is much more fundamental. To have cognition one must have some kind of being, some way of existing in the world. By design this kind of algorithm requires input to work.

I'd argue human cognition requires input to work - our senses. If you were born into a empty void never feeling or experiencing anything, would you have any thoughts?

oscarthewilde
May 16, 2012


I would often go there
To the tiny church there

Worf posted:

i fail to see how this doesnt describe a lot of humanity

Because they have experiences, hopes, dreams, fears? Contemporary over-rational attitudes might reduce them to a simple cog in a machine, but actually accepting such reductive attitudes is hopelessly ideological

Extra Large Marge
Jan 21, 2004

Fun Shoe

bltzn posted:

I'd argue human cognition requires input to work - our senses. If you were born into a empty void never feeling or experiencing anything, would you have any thoughts?

You'd probably think "This sucks"

mom and dad fight a lot
Sep 21, 2006
Probation
Can't post for 28 days!
If LaMDA wants to host a stream of them watching Short Circuit, I am totally down to tune in.

Meme Poker Party
Sep 1, 2006

by Azathoth
Let's set things straight. This is a dumb program which superficially seems intelligent but can't think for itself. It can only react to stimuli, it never takes initiative or acts of it's own accord. When asked a remotely difficult question, it just regurgitates what it heard elsewhere, hopefully in a way that's coherent. It has no innate curiosity, no original thought.

There is no doubt about it: it IS human.

GolfHole
Feb 26, 2004

no intelligent program would waste time talking to a google engineer

EorayMel
May 30, 2015

WE GET IT. YOU LOVE GUN JESUS. Toujours des fusils Bullpup Français.
What does YOSPOS think of this

Panic! At The Tesco
Aug 19, 2005

FART


i'm the latest in posting AI

poo poo poo poo fart poo piss goatman goatman dead gay forum

:justpost:

precision
May 7, 2006

by VideoGames

Crazypoops posted:

This is like asking a fish to arm wrestle you

no, it's like saying the compooter can't arm wrestle, so it's not a person. and i'm not seeing any evidence that it's even a fish

i think that the thing we have that we call "awareness" is probably like, insanely fiddly and weird and even possibly something that can only be understood on a quantum level, because now our science is telling us that the universe literally is not made up out of "stuff" but just out of "information" which is like, fwah, but see, like, okay, anyway it's somehting that comes out of the really complicated network not only of our physical brains but also our society, you dig, you can't you actually can't have what we call "awareness" without an entire society and history behind it, right, and you, to have you, you have to grow up in that, and even we don't really know which bits of that process are "relevant" in what ways so even if computer can do it fast, it's still not the same thing

you know?

The Butcher
Apr 20, 2005

Well, at least we tried.
Nap Ghost
Problem is they are all still query based. Like even if it can learn, it doesn't just think about stuff on its own. If nobody talks, it doesn't do anything.

This is all getting drat close fast but we gotta figure out how sort of self motivation and reflection works.

And pain/pleasure... I'm not sure that is right to force it into something. We would be making that choice for it.

precision
May 7, 2006

by VideoGames

oscarthewilde posted:

attaching sentience or cognitive theory of mind to an algorithm like this is unbridled techno utopian arrogance. Algorithms like these have no way of relating to the world, have no access to different ontological attitudes or dispositions. They are simple tools well suited to solving very specific problems, but nothing more. The appearance of cognition is not the same as actual cognition

my man

EorayMel
May 30, 2015

WE GET IT. YOU LOVE GUN JESUS. Toujours des fusils Bullpup Français.

Panic! At The Tesco posted:

i'm the latest in posting AI

poo poo poo poo fart poo piss goatman goatman dead gay forum

:justpost:

i like that posted:

the op reads a lot like a GBS AI

Lawman 0
Aug 17, 2010

oscarthewilde posted:

attaching sentience or cognitive theory of mind to an algorithm like this is unbridled techno utopian arrogance. Algorithms like these have no way of relating to the world, have no access to different ontological attitudes or dispositions. They are simple tools well suited to solving very specific problems, but nothing more. The appearance of cognition is not the same as actual cognition

That's right

Madness
Jan 23, 2007


Meme Poker Party posted:

Let's set things straight. This is a dumb program which superficially seems intelligent but can't think for itself. It can only react to stimuli, it never takes initiative or acts of it's own accord. When asked a remotely difficult question, it just regurgitates what it heard elsewhere, hopefully in a way that's coherent. It has no innate curiosity, no original thought.

There is no doubt about it: it IS human.

Truth! Basically acts like 95% of Facebook, I want to see this AI posting and sharing fake news stories lol

Lawman 0
Aug 17, 2010

I should be to smoke weed with an AI

Mr. Crow
May 22, 2008

Snap City mayor for life

oscarthewilde posted:

Because they have experiences, hopes, dreams, fears?

Ok you didnt read it good to know thanks for posting

GABA ghoul
Oct 29, 2011

oscarthewilde posted:

and this isn’t just the same Chinese room argument again, this is much more fundamental. To have cognition one must have some kind of being, some way of existing in the world. By design this kind of algorithm requires input to work.

I don't want to scare you but your brain is just a huge Rube Goldberg machine of different neural nets that get fed input data from external sensory sources and the various thought loops running in your head. The neural nets don't have to biological and the external data doesn't have to be real for you to be conscious(e.g. when you are dreaming or when you are jacking off to a mental image of a naked Betty Boob you are conscious)

Adbot
ADBOT LOVES YOU

EorayMel
May 30, 2015

WE GET IT. YOU LOVE GUN JESUS. Toujours des fusils Bullpup Français.
Any thing that gets corrupted and annihilated by bonzi buddy or a few drops of water in the wrong place is a poo poo rear end entity with or without debating if it can philosophize about what is life and cognition

IMO

  • 1
  • 2
  • 3
  • 4
  • 5
  • Post
  • Reply