1968: "Robot Future"

This forum is for the individual discussion thread that goes with each new comic.

Moderators: Moderators General, Prelates, Magistrates

User avatar
Soupspoon
You have done something you shouldn't. Or are about to.
Posts: 3665
Joined: Thu Jan 28, 2016 7:00 pm UTC
Location: 53-1

1968: "Robot Future"

Postby Soupspoon » Fri Mar 16, 2018 3:44 pm UTC

Image
Title text: "I mean, we already live in a world of flying robots killing people. I don't worry about how powerful the machines are, I worry about who the machines give power to."

An anagram of "Melon Usk"? :P

kiochi
Posts: 5
Joined: Wed Sep 25, 2013 10:24 pm UTC

Re: 1968: "Robot Future"

Postby kiochi » Fri Mar 16, 2018 3:51 pm UTC

*whom

User avatar
Soupspoon
You have done something you shouldn't. Or are about to.
Posts: 3665
Joined: Thu Jan 28, 2016 7:00 pm UTC
Location: 53-1

Re: 1968: "Robot Future"

Postby Soupspoon » Fri Mar 16, 2018 4:07 pm UTC

kiochi wrote:*whom

Given that I had to tap back in the text I can't easily copypasta from the other tab, I did have to just now check* whether it was me or the original that was wrong. It was (is still, at time of posting) the original. Other E&OE!


But remember, kids, "if you can replace the word with 'he', it is 'who', but if you would write 'him' then 'whom'".

It doesn't help that "…about he the machines give power to," is only marginally more awkward than "…about him the machines give power to," for other arguable reasons. ;)


* "…check just now…"?

User avatar
somitomi
Posts: 647
Joined: Fri Nov 06, 2015 11:21 pm UTC
Location: can be found in Hungary
Contact:

Re: 1968: "Robot Future"

Postby somitomi » Fri Mar 16, 2018 4:12 pm UTC

It seems strange to worry about "swarms of killer robots" while living in a world where cruise missiles equipped with nuclear warheads exist.
—◯-◯

User avatar
markfiend
Posts: 500
Joined: Fri Jul 06, 2007 9:59 am UTC
Location: UK (Leeds)

Re: 1968: "Robot Future"

Postby markfiend » Fri Mar 16, 2018 4:17 pm UTC

somitomi wrote:It seems strange to worry about "swarms of killer robots" while living in a world where cruise missiles equipped with nuclear warheads exist.

Not if you're living somewhere (Pakistan for instance) that is prone to "flying robots killing people"
advanced, forthright, signifficant
pronouns: he/him

Draconaes
Posts: 42
Joined: Fri Jan 04, 2013 7:20 pm UTC

Re: 1968: "Robot Future"

Postby Draconaes » Fri Mar 16, 2018 4:19 pm UTC

somitomi wrote:It seems strange to worry about "swarms of killer robots" while living in a world where cruise missiles equipped with nuclear warheads exist.


I guess swarms of robots are reminiscent of swarms of insects, rodents, etc. Being chased down and shredded by a multitude of dangerous critters seems more viscerally terrifying than being instantly vaporized by a fireball.

rmsgrey
Posts: 3457
Joined: Wed Nov 16, 2011 6:35 pm UTC

Re: 1968: "Robot Future"

Postby rmsgrey » Fri Mar 16, 2018 4:24 pm UTC

Humans have evolved for thousands of years to be good at getting other humans to not kill us. Our history of getting robots to not kill us is much shorter, and our history of getting self-aware robots to not kill us is nonexistent.

We've got a fairly solid grasp of the risk involved in dealing with other humans (with or without various machines) but there are some pretty big unknowns when dealing with self-aware robots (starting with what kind of self-aware robot we're dealing with). We're about as informed as Jon Snow on the matter...

User avatar
cellocgw
Posts: 1947
Joined: Sat Jun 21, 2008 7:40 pm UTC

Re: 1968: "Robot Future"

Postby cellocgw » Fri Mar 16, 2018 4:30 pm UTC

rmsgrey wrote:Humans have evolved for thousands of years to be good at getting other humans to not kill us. Our history of getting robots to not kill us is much shorter, and our history of getting self-aware robots to not kill us is nonexistent.

We've got a fairly solid grasp of the risk involved in dealing with other humans (with or without various machines) but there are some pretty big unknowns when dealing with self-aware robots (starting with what kind of self-aware robot we're dealing with). We're about as informed as Jon Snow on the matter...


Then again, I have it on good authority that Jon Snow is actually John Connor in disguise. The last time he went into his wayback machine, a couple parameters got messed up.
https://app.box.com/witthoftresume
Former OTTer
Vote cellocgw for President 2020. #ScienceintheWhiteHouse http://cellocgw.wordpress.com
"The Planck length is 3.81779e-33 picas." -- keithl
" Earth weighs almost exactly π milliJupiters" -- what-if #146, note 7

User avatar
gmalivuk
GNU Terry Pratchett
Posts: 26529
Joined: Wed Feb 28, 2007 6:02 pm UTC
Location: Here and There
Contact:

Re: 1968: "Robot Future"

Postby gmalivuk » Fri Mar 16, 2018 5:26 pm UTC

kiochi wrote:*whom

*whomst've
Unless stated otherwise, I do not care whether a statement, by itself, constitutes a persuasive political argument. I care whether it's true.
---
If this post has math that doesn't work for you, use TeX the World for Firefox or Chrome

(he/him/his)

User avatar
Charlie!
Posts: 2035
Joined: Sat Jan 12, 2008 8:20 pm UTC

Re: 1968: "Robot Future"

Postby Charlie! » Fri Mar 16, 2018 8:23 pm UTC

Image
Some people tell me I laugh too much. To them I say, "ha ha ha!"

macca
Posts: 1
Joined: Fri Mar 16, 2018 8:59 pm UTC

Re: 1968: "Robot Future"

Postby macca » Fri Mar 16, 2018 9:08 pm UTC

cellocgw wrote:Then again, I have it on good authority that Jon Snow is actually John Connor in disguise. The last time he went into his wayback machine, a couple parameters got messed up.

No no, you've got it all wrong. You thought Daenerys Targaryen was exiled to Essos to escape Robert's rebellion? Wrong - she's actually Sarah Connor in disguise, hiding there from Skynet...

gimmespamnow
Posts: 44
Joined: Wed Sep 14, 2011 6:35 am UTC

Re: 1968: "Robot Future"

Postby gimmespamnow » Fri Mar 16, 2018 10:23 pm UTC

Shouldn't the "Now" be inside the left bracket?

Since this was 9 years ago: Image

Mikeski
Posts: 1039
Joined: Sun Jan 13, 2008 7:24 am UTC
Location: Minnesota, USA

Re: 1968: "Robot Future"

Postby Mikeski » Fri Mar 16, 2018 10:38 pm UTC

somitomi wrote:It seems strange to worry about "swarms of killer robots" while living in a world where cruise missiles equipped with nuclear warheads exist.

Cruise missiles are quite expensive and relatively difficult to obtain. Nuclear weaponry is quite expensive and relatively difficult to obtain. Use of nuclear weaponry renders the surrounding area unusably toxic, possibly for generations.

Drones are relatively cheap and easy to obtain. Bullets are relatively cheap and easy to obtain. Use of a bullet renders the surrounding area pretty easily usable (once they drag the corpse off of it).

But the combination of the two problems, crazy AI with nuclear missiles, can be solved with a few games of tic-tac-toe, so each problem should be easily solvable on its own.

User avatar
pogrmman
Posts: 540
Joined: Wed Jun 29, 2016 10:53 pm UTC
Location: Probably outside

Re: 1968: "Robot Future"

Postby pogrmman » Sat Mar 17, 2018 9:51 am UTC

gimmespamnow wrote:Shouldn't the "Now" be inside the left bracket?

Since this was 9 years ago: Image


Except those aren’t currently controlled by AI. They’re just like RC planes, but at a really, really long distance.

Ray Steel
Posts: 7
Joined: Sat Mar 17, 2018 9:34 am UTC
Location: Yes, i'm a crook

Re: 1968: "Robot Future"

Postby Ray Steel » Sat Mar 17, 2018 10:09 am UTC

Charlie! pictured:
Someone build an AI that does what you toll it to do.


Hi guys,

that's crazy, they already there, all the machines.
But nobody talks the right language.

8-)

User avatar
somitomi
Posts: 647
Joined: Fri Nov 06, 2015 11:21 pm UTC
Location: can be found in Hungary
Contact:

Re: 1968: "Robot Future"

Postby somitomi » Sat Mar 17, 2018 12:26 pm UTC

Mikeski wrote:
somitomi wrote:It seems strange to worry about "swarms of killer robots" while living in a world where cruise missiles equipped with nuclear warheads exist.

Cruise missiles are quite expensive and relatively difficult to obtain. Nuclear weaponry is quite expensive and relatively difficult to obtain. Use of nuclear weaponry renders the surrounding area unusably toxic, possibly for generations.

Drones are relatively cheap and easy to obtain. Bullets are relatively cheap and easy to obtain. Use of a bullet renders the surrounding area pretty easily usable (once they drag the corpse off of it).

But then what we really worry about is "drones with guns" and not AI, is it?
—◯-◯

rmsgrey
Posts: 3457
Joined: Wed Nov 16, 2011 6:35 pm UTC

Re: 1968: "Robot Future"

Postby rmsgrey » Sat Mar 17, 2018 5:14 pm UTC

Ray Steel wrote:
Charlie! pictured:
Someone build an AI that does what you toll it to do.


Hi guys,

that's crazy, they already there, all the machines.
But nobody talks the right language.

8-)


Yeah, it's not AI that does what we tell it that's needed, but AI that ignores what we actually say, and does what we mean instead. Or even AI that asks us if we really want to nuke Tasmania and waits for us to sober up...

Ray Steel
Posts: 7
Joined: Sat Mar 17, 2018 9:34 am UTC
Location: Yes, i'm a crook

Re: 1968: "Robot Future"

Postby Ray Steel » Sat Mar 17, 2018 5:56 pm UTC

rmsgrey wrote:Yeah, it's not AI that does what we tell it


You must only push the right Buttons in the right Order.

8-)
P.S.:
These autocorrector drive me craaaazy!

Ray Steel
Posts: 7
Joined: Sat Mar 17, 2018 9:34 am UTC
Location: Yes, i'm a crook

Re: 1968: "Robot Future"

Postby Ray Steel » Sat Mar 17, 2018 5:57 pm UTC

o.k. It's a joke.

:D :D :D
P.S.:
These autocorrector drive me craaaazy!

User avatar
Charlie!
Posts: 2035
Joined: Sat Jan 12, 2008 8:20 pm UTC

Re: 1968: "Robot Future"

Postby Charlie! » Sat Mar 17, 2018 7:21 pm UTC

rmsgrey wrote:
Ray Steel wrote:
Charlie! pictured:
Someone builds an AI that does what you tell it to do.

that's crazy, they already there, all the machines.
But nobody talks the right language.

Yeah, it's not AI that does what we tell it that's needed, but AI that ignores what we actually say, and does what we mean instead. Or even AI that asks us if we really want to nuke Tasmania and waits for us to sober up...

Programs do what you program them to do, and always will, but right now, getting a program to do what you want is not a primitive action - it's not something you just do, it's a difficult many-step process. So, if you want to revise my sentence, please revise it to something like "Someone builds an AI that takes general orders in natural language and takes actions to implement them."

I was tempted to put "AI understands Winograd schemas" there, but I figured that might be too abstruse :P
Some people tell me I laugh too much. To them I say, "ha ha ha!"

Ray Steel
Posts: 7
Joined: Sat Mar 17, 2018 9:34 am UTC
Location: Yes, i'm a crook

Re: 1968: "Robot Future"

Postby Ray Steel » Sun Mar 18, 2018 1:08 am UTC

Yes, i know. What you talk about is self-awareness. You need it to realize that You are not alone, or vice versa.

And if You want to talk to a machine instead of pushing buttons You need an interface.
And this is a difference in degree not in principe.

:D :D :D
P.S.:
These autocorrector drive me craaaazy!

rmsgrey
Posts: 3457
Joined: Wed Nov 16, 2011 6:35 pm UTC

Re: 1968: "Robot Future"

Postby rmsgrey » Sun Mar 18, 2018 1:11 am UTC

Charlie! wrote:
rmsgrey wrote:
Ray Steel wrote:
Charlie! pictured:
Someone builds an AI that does what you tell it to do.

that's crazy, they already there, all the machines.
But nobody talks the right language.

Yeah, it's not AI that does what we tell it that's needed, but AI that ignores what we actually say, and does what we mean instead. Or even AI that asks us if we really want to nuke Tasmania and waits for us to sober up...

Programs do what you program them to do, and always will, but right now, getting a program to do what you want is not a primitive action - it's not something you just do, it's a difficult many-step process. So, if you want to revise my sentence, please revise it to something like "Someone builds an AI that takes general orders in natural language and takes actions to implement them."

I was tempted to put "AI understands Winograd schemas" there, but I figured that might be too abstruse :P


Right now, my computer does what I tell it to do - if I double-click on an icon on my desktop, my computer "runs" whatever the icon represents. If I tell my computer to format the hard drive, it will format the hard drive (I may have to insist a little, but it will). Depending what software the computer has installed/active, I can tell it to do all sorts of interesting things with a single step.

Actually creating the software that lets me give simple commands to the computer to do things was a complex, multi-stage, multi-layered process going back to sometime in the 60s or 70s when the compiler that compiled the compiler that compiled the compiler that compiled the compiler that compiled the software I'm actually using was written (give or take some more generations of compilers), but using AI that can understand human language is more akin to an end-user's use of software than a programmer's creation of it.

User avatar
ucim
Posts: 6567
Joined: Fri Sep 28, 2012 3:23 pm UTC
Location: The One True Thread

Re: 1968: "Robot Future"

Postby ucim » Sun Mar 18, 2018 1:28 am UTC

rmsgrey wrote:Right now, my computer does what I tell it to do - if I double-click on an icon on my desktop, my computer "runs" whatever the icon represents.
No. When you double click an icon, the computer does whatever the authors of that application tell it to do. And that application, behind the scenes, is telling other programs, written by other people, to do whatever the other people wanted those programs to do. You hope and trust that this is all in alignment with what you originally hoped the icon represents, but that's all.

When you run a program, it is like inviting people into your house to do "stuff", and giving them free reign to do "other stuff that they, in their sole judgment, think necessary". None of those people speak your language, none of those people are answerable to you (and even the original people you invited are answerable to you - it's take it or leave it (too late!)). And eventually you have a frat party in your living room when all you wanted was somebody to clean the floor. And you don't have a clue as to what is happening, because you are not invited! And even if the frat party cleans up after itself, they've gone through your drawers, they know where you keep your sex toys, who you owe money to, who you are breaking up with, where you travel and when, and have copied the keys to all your locks.

Only in the most general sense does the computer do what you tell it, because the icon can lie to you, so you are telling it one thing and it is "hearing" another thing.

A hammer does what you want. A computer, not so much. A smartphone, only enough to keep you fooled.

Jose
Order of the Sillies, Honoris Causam - bestowed by charlie_grumbles on NP 859 * OTTscar winner: Wordsmith - bestowed by yappobiscuts and the OTT on NP 1832 * Ecclesiastical Calendar of the Order of the Holy Contradiction * Please help addams if you can. She needs all of us.

rmsgrey
Posts: 3457
Joined: Wed Nov 16, 2011 6:35 pm UTC

Re: 1968: "Robot Future"

Postby rmsgrey » Sun Mar 18, 2018 2:20 am UTC

ucim wrote:
rmsgrey wrote:Right now, my computer does what I tell it to do - if I double-click on an icon on my desktop, my computer "runs" whatever the icon represents.
No. When you double click an icon, the computer does whatever the authors of that application tell it to do. And that application, behind the scenes, is telling other programs, written by other people, to do whatever the other people wanted those programs to do. You hope and trust that this is all in alignment with what you originally hoped the icon represents, but that's all.

When you run a program, it is like inviting people into your house to do "stuff", and giving them free reign to do "other stuff that they, in their sole judgment, think necessary". None of those people speak your language, none of those people are answerable to you (and even the original people you invited are answerable to you - it's take it or leave it (too late!)). And eventually you have a frat party in your living room when all you wanted was somebody to clean the floor. And you don't have a clue as to what is happening, because you are not invited! And even if the frat party cleans up after itself, they've gone through your drawers, they know where you keep your sex toys, who you owe money to, who you are breaking up with, where you travel and when, and have copied the keys to all your locks.

Only in the most general sense does the computer do what you tell it, because the icon can lie to you, so you are telling it one thing and it is "hearing" another thing.

A hammer does what you want. A computer, not so much. A smartphone, only enough to keep you fooled.

Jose


Okay, if you prefer, my observation so far has consistently been that when I double-click an icon on the desktop (that is, when I press a particular button on my mouse twice in rapid succession while a particular portion of the computer monitor displays a particular pattern of light) then my computer "runs" whatever the icon represents (that is, a pattern of light consistent with what I expect to happen appears on the computer monitor, sounds emerge from the speakers consistent with what I expect to happen, and ongoing patterns continue to have the expected correlations with my actions). It's probable that additional things are happening that I'm not aware of, but the computer does appear to be doing what I believe I'm telling it to do.

When I get home, the house is clean, and there's no trace of whatever frat parties may have been held, so, until and unless that changes, I'm not bothered about what else the cleaners may have got up to (plus this particular cleaning firm comes with references from other people, and I have a private detective on retainer looking into everyone - except the detective - who does anything in my house).

Besides, while I may not know exactly how other people have specified for my computer to interpret a given command, I'm still the one giving the computer the top-level orders it follows. The computer does what I (actually) tell it to, not necessarily what I intend to tell it to.

Ray Steel
Posts: 7
Joined: Sat Mar 17, 2018 9:34 am UTC
Location: Yes, i'm a crook

Re: 1968: "Robot Future"

Postby Ray Steel » Sun Mar 18, 2018 2:30 am UTC

rmsgrey wrote:using AI that can understand human language is more akin to an end-user's use of software than a programmer's creation of it.


You are kidding :wink:

oookay, for clearence, "pushing the right Buttons in the right Order" means You have to write software.

8-)

P.S.:
These autocorrector drive me craaaazy!
P.S.:
These autocorrector drive me craaaazy!

User avatar
Soupspoon
You have done something you shouldn't. Or are about to.
Posts: 3665
Joined: Thu Jan 28, 2016 7:00 pm UTC
Location: 53-1

Re: 1968: "Robot Future"

Postby Soupspoon » Sun Mar 18, 2018 7:34 am UTC

All these people clicking on icons.

Whatever happened to the age-old posting of questions to a room in China?

User avatar
ucim
Posts: 6567
Joined: Fri Sep 28, 2012 3:23 pm UTC
Location: The One True Thread

Re: 1968: "Robot Future"

Postby ucim » Sun Mar 18, 2018 1:40 pm UTC

Soupspoon wrote:All these people clicking on icons.

Whatever happened to the age-old posting of questions to a room in China?

Emoji.

rmsgrey wrote:Okay, if you prefer, my observation so far has consistently been that when I double-click an icon on the desktop [...] and I have a private detective on retainer...
You have the detective because of what I said. And yes, desktop computers seem to behave this way. Smartphones don't. Apps are far less trustworthy than applications.

Jose
Order of the Sillies, Honoris Causam - bestowed by charlie_grumbles on NP 859 * OTTscar winner: Wordsmith - bestowed by yappobiscuts and the OTT on NP 1832 * Ecclesiastical Calendar of the Order of the Holy Contradiction * Please help addams if you can. She needs all of us.

rmsgrey
Posts: 3457
Joined: Wed Nov 16, 2011 6:35 pm UTC

Re: 1968: "Robot Future"

Postby rmsgrey » Sun Mar 18, 2018 2:33 pm UTC

ucim wrote:
rmsgrey wrote:Okay, if you prefer, my observation so far has consistently been that when I double-click an icon on the desktop [...] and I have a private detective on retainer...
You have the detective because of what I said. And yes, desktop computers seem to behave this way. Smartphones don't. Apps are far less trustworthy than applications.

Jose

I don't have a smartphone.

But it appears that my computer is continuing to send the messages I tell it to to this forum thread, so -+=ChEeP C1@liS!!!=+- I'm still reasonably confident that it does what I tell it to.

User avatar
Cougar Allen
Posts: 57
Joined: Thu Dec 24, 2015 4:49 am UTC

Re: 1968: "Robot Future"

Postby Cougar Allen » Mon Mar 19, 2018 5:04 am UTC

The part I worry about is off to the left on that timeline - the period when there's no AI; it's still just people playing video games and controlling drones to kill other people, but it will no longer be only people in other parts of the world who are getting killed. It's cheap, easily available technology. Anyone can use it....

User avatar
somitomi
Posts: 647
Joined: Fri Nov 06, 2015 11:21 pm UTC
Location: can be found in Hungary
Contact:

Re: 1968: "Robot Future"

Postby somitomi » Mon Mar 19, 2018 9:43 am UTC

rmsgrey wrote:
Okay, if you prefer, my observation so far has consistently been that when I double-click an icon on the desktop (that is, when I press a particular button on my mouse twice in rapid succession while a particular portion of the computer monitor displays a particular pattern of light) then my computer "runs" whatever the icon represents (that is, a pattern of light consistent with what I expect to happen appears on the computer monitor, sounds emerge from the speakers consistent with what I expect to happen, and ongoing patterns continue to have the expected correlations with my actions).

Unless the computer didn't do what the computer itself expected and you get something like this (or this) pattern of lights.
—◯-◯

Norr
Posts: 3
Joined: Sun Nov 21, 2010 7:35 pm UTC

Re: 1968: "Robot Future"

Postby Norr » Mon Mar 19, 2018 3:06 pm UTC


elasto
Posts: 3568
Joined: Mon May 10, 2010 1:53 am UTC

Re: 1968: "Robot Future"

Postby elasto » Wed Mar 21, 2018 5:51 am UTC

somitomi wrote:Unless the computer didn't do what the computer itself expected and you get something like this (or this) pattern of lights.

I myself got this rather amusing light pattern not so long ago:

Image

I guess that's the software equivalent to 'the operation was a success but the patient died'...


Return to “Individual XKCD Comic Threads”

Who is online

Users browsing this forum: Google [Bot] and 31 guests