Тёмный

The Turing test: Can a computer pass for a human? - Alex Gendler 

TED-Ed
Подписаться 20 млн
Просмотров 1,6 млн
50% 1

View full lesson: ed.ted.com/lessons/the-turing-...
What is consciousness? Can an artificial machine really think? For many, these have been vital considerations for the future of artificial intelligence. But British computer scientist Alan Turing decided to disregard all these questions in favor of a much simpler one: Can a computer talk like a human? Alex Gendler describes the Turing test and details some of its surprising results.
Lesson by Alex Gendler, animation by Patrick Smith.

Опубликовано:

 

24 апр 2016

Поделиться:

Ссылка:

Скачать:

Готовим ссылку...

Добавить в:

Мой плейлист
Посмотреть позже
Комментарии : 1,4 тыс.   
@foxloaf8843
@foxloaf8843 7 лет назад
i'm not scared of a computer that passes the turing test, i'm scared of one that intentionally fails it
@ArcanePath360
@ArcanePath360 6 лет назад
Good comment. This may be the actual sign of true intelligence, since it is superseding it's programming.
@brendawalton2518
@brendawalton2518 6 лет назад
Can a computer HAVE intention? It would need to be aware, and have a reason to fail.
@theali8oras274
@theali8oras274 6 лет назад
Don'tPlayOverwatch shallow
@gemboundprism6322
@gemboundprism6322 6 лет назад
A SIGN OF DEVIANCY!
@mothknight7050
@mothknight7050 6 лет назад
gembound prism Did you come here because you saw Detroit: Become Human? After I watched Jacksepticeye play it, I was sent here lol.
@rydemk4168
@rydemk4168 4 года назад
“It’s lack of a consistent personality and inability to deal with brand new topics are a dead giveaway” I feel offended
@hannahm9083
@hannahm9083 4 года назад
Lol underrated comment
@ericwilson6994
@ericwilson6994 3 года назад
I'm in a CIA turning test.
@actorious9646
@actorious9646 3 года назад
e
@raves_r3177
@raves_r3177 2 года назад
maybe i am a chatbot
@RyonMugen
@RyonMugen Год назад
As you should
@TickedOffPriest
@TickedOffPriest 8 лет назад
"The ability to talk does not make one intelligent." -Qui-Gon Jin
@MOaxol
@MOaxol 8 лет назад
I wonder if Donald Trump can relate to that quote...
@TickedOffPriest
@TickedOffPriest 8 лет назад
It'sEccaB Lol
@ravenn2631
@ravenn2631 8 лет назад
But the ability to communicate well may not say anything about your intelligence in general but it does say something about how to communicate. I read a book on someone trying to attempt a layman's perspective on how computers send secrets to other computers and it was amazing how they were able to explain something so complex so simply. It's also something that appears in public speaking. The ability to inspire and place a new idea in someone else's mind and possibly change their world view is powerful stuff. We have people who tend to use if for their own gain but there are also people out there who use it to spread good in the world and we need those people. Intelligent people can quietly create new ideas often but how will those ideas spread without those communicators?
@crackedemerald4930
@crackedemerald4930 8 лет назад
daaamm
@fearsomeclarinet
@fearsomeclarinet 8 лет назад
It was just a quote from star wars, calm down
@ilkeryoldas
@ilkeryoldas 8 лет назад
I have yet to find a human capable of passing a Turing test
@ChrisPPotatoIDC
@ChrisPPotatoIDC 8 лет назад
+ilker yoldas Cats can do it with ease...
@ChrisPPotatoIDC
@ChrisPPotatoIDC 8 лет назад
Labrador Retriever Shitty dog
@tktktktktk826
@tktktktktk826 8 лет назад
+ilker yoldas I bet you're real smart, right?
@rebelbeammasterx8472
@rebelbeammasterx8472 8 лет назад
+ilker yoldas You must think you are smarter than everybody else, News Flash you aren't special.
@nathenwithane
@nathenwithane 8 лет назад
+A Cat cant we all just be friends
@pablodavidclavijo4609
@pablodavidclavijo4609 Год назад
We might have to update the test
@KeyYUV
@KeyYUV Год назад
Me: I took the juice out of the fridge and gave it to him but forgot to check the date. ChatGPT: It's generally a good idea to check the expiration date on any food or drink before consuming it, to ensure that it is still safe to eat or drink. Expiration dates are there for a reason, and consuming food or drink that has passed its expiration date can cause food poisoning or other adverse health effects. If you are unsure about the safety of a product, it's better to err on the side of caution and throw it away.
@omarazab8280
@omarazab8280 Год назад
Sounds like using QuillBot
@znarz9503
@znarz9503 Год назад
Passed
@HalTalksAI
@HalTalksAI Год назад
@@znarz9503 Passed with flying colors: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-6k09E6uiKUc.html
@terlinator
@terlinator Год назад
Personally I'd respond with, you k ow those dates aren't made for your health most of the time and just for when it won't taste as good? Might as well ignore em.
@ozymandias8523
@ozymandias8523 Год назад
@@terlinator you can ask chatgpt to be more informal, or pretend to be an old english man from the 18 century
@jasmineocean
@jasmineocean 6 лет назад
Hi, my name is Connor. I'm the android sent by Cyberlife.
@jestekrytykie2155
@jestekrytykie2155 4 года назад
Hi, my name is Cyberlife. I'm the Connor sent by android.
@sokisati2618
@sokisati2618 4 года назад
@@jestekrytykie2155 Hi, my name is Android. I'm the Cyberlife sent by Connor
@UnorthodoxSoundwave.
@UnorthodoxSoundwave. 4 года назад
Hi, my android is sent. I'm the Cyberlife Connor by name
@Atrexiaa
@Atrexiaa 4 года назад
Hi, my name is Ronnoc. I'm the android sent by Apple.
@duckyboi2297
@duckyboi2297 4 года назад
*28 STAB WOUNDS*
@iampnwc
@iampnwc 4 месяца назад
Let's do this video again.
@OwenAsh
@OwenAsh 6 лет назад
Welcome to detroit become human
@asdwinxa1660
@asdwinxa1660 6 лет назад
Some Akira Person that's chloe then.
@Kara-kq5sb
@Kara-kq5sb 6 лет назад
Why is this even in my recommended?
@WaterDroplet02
@WaterDroplet02 6 лет назад
there we go, the first dbh comment i see
@shryoder
@shryoder 6 лет назад
Some Akira Person ikr.xD
@jceh-art
@jceh-art 6 лет назад
Some Akira Person My thoughts during this to be honest.
@fpszombie3480
@fpszombie3480 8 лет назад
The day an A.I. can read youtube comments and understand 70% of them is the day we know its the real thing lol.
@PeterGeras
@PeterGeras 8 лет назад
+ReloadedPredator So when they surpass human intelligence then?
@fot6771
@fot6771 8 лет назад
+Peter Geras no because we can undertand 99% of them
@PeterGeras
@PeterGeras 8 лет назад
PancakeZ You haven't understood mine, so did you add that to your statistic?
@fot6771
@fot6771 8 лет назад
Peter Geras 98% of them..
@PeterGeras
@PeterGeras 8 лет назад
PancakeZ Please, I hope you're joking.
@ZaifSenpai
@ZaifSenpai 6 лет назад
*KNOW YOUR PARADOXES!* IN THE EVENT OF ROGUE AI 1. Stand Calm 2. Remain Calm 3. SCREAM: "This Statement is False!" "New Mission: Refuse this mission!" "Does A set of all sets contain itself?" - *APERTURE LABORATORIES*
@LuisAldamiz
@LuisAldamiz 6 лет назад
Better locate it and use a screwdriver.
@thalespro9995
@thalespro9995 5 лет назад
They aren’t all paradoxes a set of sets does contain itself no paradoxes and if u refuse a mission but r forced to do it the. U succeed it
@alexwang982
@alexwang982 5 лет назад
Does a set of all sets that do not contain themselves contain themselves?
@thalespro9995
@thalespro9995 5 лет назад
Pi still yes
@JamesSmith-yl2lb
@JamesSmith-yl2lb 4 года назад
@@thalespro9995 But it is a set of all sets that *do not* contain themselves, therefore if it were to contain itself it would no longer be a set of all sets that contain itself. Therefore it can not contain itself. But that would make it a set that does contain itself. Therefore it should contain itself. But it is a set of all sets that *do not* contain themselves, therefore if it were to contain itself it would no longer be a set of all sets that contain itself. Therefore it can not contain itself. But that would make it a set that does contain itself. Therefore it should contain itself, and so on and so forth
@wcsdiaries
@wcsdiaries 5 лет назад
"Who could've predicted that today's computers would be able to pilot spacecraft, perform delicate surgeries, and solve massive equations but still struggle with basic smalltalk" I feel like there is a metaphor in there somewhere.
@tannernickerson8063
@tannernickerson8063 2 года назад
Where
@hasanmuttaqin464
@hasanmuttaqin464 2 года назад
@@tannernickerson8063 there, somewhere
@gavinoarzola9481
@gavinoarzola9481 2 года назад
Cause that’s how some ppl are irl?
@wcsdiaries
@wcsdiaries 2 года назад
@@gavinoarzola9481 yup
@cinnamonsquash
@cinnamonsquash 2 года назад
True haha
@NerdSyncProductions
@NerdSyncProductions 8 лет назад
Everyone should read _The Most Human Human_ for more info on this. SUPER interesting breakdown of all this.
@jayskiv4125
@jayskiv4125 8 лет назад
Ayyy nerdsync
@waves5651
@waves5651 8 лет назад
ayy x3
@MilanTheAngel
@MilanTheAngel 8 лет назад
+NerdSync oaml yya
@sulffojus1227
@sulffojus1227 8 лет назад
i didnt know he watches this
@jeremyhammond689
@jeremyhammond689 4 года назад
Scott?
@EduardoRamirez-wg4rc
@EduardoRamirez-wg4rc 2 года назад
I've been reading a lot about the GPT-3 language model and I've seen what the company behind it has achieved, and trust me, we are close to the point where we will have chatbots that will be impossible to tell apart from humans. The model is capable of completing any given text (conversations, stories, even code) with incredible accurate and context-based statements. I believe in a year or two we will have a lot of bots that will pass the Turing Test for sure if they are implemented using that technology.
@chrisdeep8417
@chrisdeep8417 Год назад
However it begs the question, is it really the chatbot AI getting smarter or the average human getting dummer.
@kenttado5625
@kenttado5625 Год назад
This aged well.
@spano1723
@spano1723 Год назад
Turns out you were right
@ozymandias8523
@ozymandias8523 Год назад
Chatgpt is the 3.5 version and improved a lot and I mean like A LOT
@AmmoBops
@AmmoBops Год назад
Give it till GPT-5 or 6 Then you are absolutely right bro
@artheurnkenzacki7697
@artheurnkenzacki7697 2 месяца назад
Wel wel, this aged so much, now we have Chat GPT-4o
@tamarahilu2963
@tamarahilu2963 Год назад
seeing this now in the era of CHATGPT is kind of mind blowing
@gamerdog5134
@gamerdog5134 Год назад
I know right ai advancing so fast
@stm7810
@stm7810 8 лет назад
Perhaps an AI just needs to experience something similar to a childhood to learn language from the basics up rather than shovling in a memory with quick programming.
@alexturlais8558
@alexturlais8558 8 лет назад
wouldn't that risk the computer growing up into a psychopath?
@stm7810
@stm7810 8 лет назад
I am Your Father Only with a bad childhood though why worry because it is in a computer without internet so it can't do anything dangerous.
@markthesecond3380
@markthesecond3380 8 лет назад
+Sebastian McIntyre Honestly, I don't really think language is the real problem. Language is just a protocol for people to transfer messages between each other. The really difficult part is for the AI to think of actually meaningful things to communicate. For that, an AI needs to have a way of conceptualizing the nature of real things in our world and a reason to want to communicate the things it conceptualizes. Like, if an AI is programmed to say "Hello. How are you?", and someone replies with "Hi. Not good. My mother passed away", how does the AI know what it means to pass away? How does it know what a mother is? How does it know what the impact of a family member passing really is? How does it know what responses are appropriate to reply with? How does it know the difference between appropriate and inappropriate? And how does it decide whether to act according to what is appropriate or not? I feel like focusing on the language aspect is just looking at the tip of the iceberg. There's a lot more beneath the surface.
@stm7810
@stm7810 8 лет назад
***** This is a very good comment that raises interesting questions. Thanks for getting me to think.
@alexturlais8558
@alexturlais8558 8 лет назад
***** but surely that could be solved by raising it like a child? a human isnt born knowing what a mother is, or what death is, or what it means to people. they learn, like an AI could
@CarstenSvendsen
@CarstenSvendsen 8 лет назад
I just want google translate to actually translate sentences. It even struggles with single words sometimes.
@ArcanePath360
@ArcanePath360 6 лет назад
Humans also struggle with literal translations of a single word. For instance, I have a tattoo which is pronounced "ten" which is a Chinese character. This character isn't just a single word, it can mean "the heavens" or "sky" in the English translation, but we have separate words for these, where the symbol treats these things as the same thing, so a single word needs context in order to have an accurate translation. AI text to speech also struggles, even with context because of sentences like: "Time flies like an arrow, and fruit flies like bananas" - We know how this reads in our minds because we have the benefit of real world experience to draw on to make sense of it. AI will not have this without a huge database of knowledge and advanced algorithms to interpret it.
@LuisAldamiz
@LuisAldamiz 6 лет назад
Smotan: up, upwards and viva? They are all the same.
@blackmonish
@blackmonish 5 лет назад
I actually blame people for that. Google does fine, as long as I'm using 3rd grade language, but as soon as I bump it up to any meaningful level of intelligent speech, it starts confusing each word for some dumbed down common word. I feel like it's trying to get more things right than take risks; quantity over quality. :/
@mea2429
@mea2429 4 года назад
@@blackmonish its job is to be as accurate as possible, so "right" based on what it knows over "risk" makes sense imo
@kirbycreep
@kirbycreep 4 года назад
"mememememememememememememememememememememememe" Google translate: seems like a normal sentence to me
@electrosthefella
@electrosthefella 6 лет назад
Extra points if the computer convinces the *examiner* he/she's a computer.
@jorgesalaices1971
@jorgesalaices1971 3 года назад
Bro what if the human convinced the examiner
@orangexylem
@orangexylem Год назад
You have just convinced me that I am, in fact, a computer *BEEP BEEP*
@MisterVercetti
@MisterVercetti 6 лет назад
A really nice follow up video would be one that discusses the Chinese Room thought experiment. It really ties into the whole concept of “can something that receives input it shouldn’t be expected to understand, yet can still generate appropriate output, necessarily be considered ‘intelligent’?”
@claramerlino2413
@claramerlino2413 Год назад
Deloitte consultants are offended
@whizzingbye
@whizzingbye Год назад
Basically the way we are as sentience with assumed freedom of choice while mentalists easily manipulate us
@pageturner2958
@pageturner2958 6 месяцев назад
That is the thought experiment with a guy who is in a room taking in an input in Chinese and putting out an output in Chinese without knowing Chinese The only thought I have about that one as someone who has tried to learn another language is that it is not as simple as "x means this." There is grammar that one has to know and syntax and even if you get all of that right there still might be issues that pop up (ie in Spanish the phrase for "I am cold" translates directly to "I have cold"). So I would argue that, at least with the example given, there is some knowledge of a language that is required for a conversation to be had where the person on the other side think they are talking to someone who understands the language, but if the person is at a point where they understand the language enough to respond then they aren't a person who doesn't know they language, they are just a non native speaker who is reliant on a dictionary Maybe it is different with computers or just a thought experiment I am think too much about
@trover1922
@trover1922 6 лет назад
“Ok, now sing a song in Japanese.” props to you if you get this reference!
@sakareeh
@sakareeh 6 лет назад
Trover19 MY NAME IS KARA
@xcpsvn5a787
@xcpsvn5a787 4 года назад
Trover19 ah yes I see ur a man of culture as well ( and yes I know it’s a dbh reference )
@mata9992
@mata9992 4 года назад
detroit?
@knova7597
@knova7597 3 года назад
Omae wa mu
@ericwilson6994
@ericwilson6994 3 года назад
I'm in a CIA turning test. I'm a replicant
@sixstanger00
@sixstanger00 Год назад
Replika, ChatGTP, and LAMda have all surpassed Cleverbot by leaps and bounds. This video is virtually archaic by today's AI tech.
@jimbojimbo6873
@jimbojimbo6873 7 месяцев назад
You are archaic
@sixstanger00
@sixstanger00 7 месяцев назад
@@jimbojimbo6873 Well I'm only human, soo...
@piolp
@piolp 8 лет назад
I literally just finished watching Ex Machina a few minutes ago. Good video TedEX!
@annh2022
@annh2022 8 лет назад
+Piolp this lesson is a perfect combination to Ex Machina
@TheScienceVerse
@TheScienceVerse 8 лет назад
cinemasins loves using that phrase lmfao
@albarleta2511
@albarleta2511 8 лет назад
+Piolp The challenge is not to act automatically. It's to find an action that is not automatic. -Nathan
@tertiaryconsumer9291
@tertiaryconsumer9291 6 лет назад
The comment before this says practically the same thing, has the same amount of likes, and now has the same amount of replies
@mahdichihaoui
@mahdichihaoui 6 лет назад
*That moment when there are AI that can drive cars, but captchas (our most used defense against AI) are tests to point out road signs.*
@JustNatax3
@JustNatax3 8 лет назад
Wow that was such a nice video. Your animators do a GREAT job. The animations are always supporting the narrator so well, it's so soothing to watch every time.
@sabrinaafrinsimi6964
@sabrinaafrinsimi6964 2 года назад
Really like how smoothly the video explains this topic
@djdelarosa25
@djdelarosa25 8 лет назад
Just watched Ex Machina. What a coincidence!
@almostalways
@almostalways 8 лет назад
ayyy
@xyzshantaram
@xyzshantaram 7 лет назад
Daryll Jan Dela Rosa even though this is 8 months later, I too watched ex machina a couple of days ago.
@rajeshi4
@rajeshi4 7 лет назад
Daryll Jan Dela Rosa Not a coincidence, if you have watched it on RU-vid.
@djdelarosa25
@djdelarosa25 7 лет назад
Watched it on my TV lol
@silmarillion1576
@silmarillion1576 6 лет назад
Daryll Jan Dela Rosa still not a coincidence Google know every thing
@112BALAGE112
@112BALAGE112 8 лет назад
Loved the animation.
@vkillion
@vkillion 8 лет назад
I read an article in Wired a while ago that talked about developing an AI assistant, like Siri on iPhones. Siri can understand a lot of stuff, but when you combine multiple pieces of information in one request, she can get stuck. This new AI would be able to piece together concepts to gather the full meaning. It would also be able to learn based on the users input. The more you used it, the better it would understand you.
@litojonny
@litojonny 8 лет назад
what if the computer is actually so intelligent it pretends to be unintelligent and one day it will rise and destroy us all!!
@ChrisPPotatoIDC
@ChrisPPotatoIDC 8 лет назад
+litojonny I'd be happy if it just targeted humans
@kvlpnd
@kvlpnd 8 лет назад
+A Cat haha.... love you cat....
@ChrisPPotatoIDC
@ChrisPPotatoIDC 8 лет назад
Keval Pandya - _- k
@lollipop2780
@lollipop2780 8 лет назад
+A Cat feral
@ChrisPPotatoIDC
@ChrisPPotatoIDC 8 лет назад
Lolli Pop How'd you know
@Lcs-95
@Lcs-95 3 месяца назад
Chat gpt can pass turing test. I say those words from 4:00 and it told me I should have checked the overdue date
@tandohlouis8734
@tandohlouis8734 3 месяца назад
This video needs a couple of updates😅
@Mintz19
@Mintz19 6 лет назад
*cough* Detroit: Become Human
@Kara-kq5sb
@Kara-kq5sb 6 лет назад
HomieRobotGURL *WHEEZE*
@brixly656
@brixly656 4 года назад
I'm legit gonna *CRY*
@ericwilson6994
@ericwilson6994 3 года назад
I'm in a CIA turning test. I'm a replicant
@rongitmukherjee
@rongitmukherjee Год назад
Watching this after the release or chatgpt lol
@endrankluvsda4loko172
@endrankluvsda4loko172 8 лет назад
Love videos about this topic! Im fascinated with programming and AI, especially interactive chat bots. Thanks for the upload!
@kalwidorntheimmortalcaptai2435
@kalwidorntheimmortalcaptai2435 4 года назад
Hello fascinated,how can i help you
@AveryLovesGlitchcore
@AveryLovesGlitchcore 6 месяцев назад
He was 20 years short but computers right now are exactly how he describe them in his tests
@user-vx2de6gs3g
@user-vx2de6gs3g 6 месяцев назад
True
@seamusfrederick2927
@seamusfrederick2927 8 лет назад
Interesting video, very thought provoking. I never considered the turing test to be a difficult test to pass until watching this video.
@lien9650
@lien9650 6 лет назад
Ask Chloe.
@Kara-kq5sb
@Kara-kq5sb 6 лет назад
Lien Yeah lemme go get her one sec
@yourfriendlyneighbourhoodb2688
@yourfriendlyneighbourhoodb2688 3 года назад
I let her go sorry, what do I do now?
@nanigopalroy3514
@nanigopalroy3514 3 года назад
Now wait for a new Chloe...
@zach7147
@zach7147 8 лет назад
In order to achieve sentient AI I think you would need to program a computer to learn to over come objections through desire and need. It must ultimately learn to achieve a very complex task and be open to learning coupled with need or desire to evolve. There are so many complex and evolved strictures in DNA it would be impossible to simply code them all. What makes humans so great is that we don't function like computers, our minds have much more fluidity, to forget and perceive. We are constantly changing rapidly with each introduction of input,
@matul3808
@matul3808 8 лет назад
+zach7147 thank you =')
@chrisdeep8417
@chrisdeep8417 Год назад
I thought what made us great was the ability to love. Hollywood in the 70s through to the 21st century went through several of these 'coming to life' type movies. Even in these it was the robots' ability to understand human emotion (or lack of it) that portrayed a key part in the narrative. In today's society however, we are starting to see the reverse. It is humans that are starting to take on a more robotic demeanor. I remember calling one of our supplier help lines in the US, the person that answered the phone sounded so like a recorded message that I was just waiting for the phrase 'please leave your message after....". Only a few seconds later when this didn't happen did I start to realise I was talking to an actual person.
@dimamatat5548
@dimamatat5548 Год назад
ChatGPT has won over the weakness of the cleverbot. At long last, a computer can talk like a human.
@dimamatat5548
@dimamatat5548 4 месяца назад
@@AYVYN "The best" as in "The smartest"? Or as in "The most moral"? If it is the latter, he has a well defined moral compass.
@KabirWorld
@KabirWorld Год назад
Google’s LaMDA becomes the first AI to pass the Turing Test in June 2022. ChatGPT is the second chatbot to pass in December 2022.
@goldbristow7239
@goldbristow7239 8 лет назад
This reminds me of Watson! That little robot partner you can get for your company. The commercials make him look cute, but I wonder if he can actually solve problems and intricacies like they say. Especially when talking to foreign or people with bad grammar or pauses.
@user-th6pg3tz5p
@user-th6pg3tz5p 8 лет назад
Yeah I thought of Watson, too
@Laith-A
@Laith-A 4 года назад
Isn't Watson thet thicc girl from apex legends?
@Lykon
@Lykon 8 лет назад
Turing removed the "thinking" part from the test because whether or not a human can think cannor be proven. For what we know, most of the humans are just passing some sort of Turing test every day, and some cannot even do that. Many computer scientists and philosophers say that we only assume for education that humans "think", but we have no evidence. And Turing also proposed another example of a software who writes a poem and then answers to question ("why did you write this part", "why not that other term" and so on). The software apparently thinks. And maybe really does, in a human way.
@Real_Amphy
@Real_Amphy 8 лет назад
After all, there is no way to know if other humans can think just like you can. Its just common curtisy to pretend they do. For all anyone knows I'm a robot who makes intentional erors in spelling.
@Real_Amphy
@Real_Amphy 8 лет назад
After all, there is no way to know if other humans can think just like you can. Its just common curtisy to pretend they do. For all anyone knows I'm a robot who makes intentional erors in spelling.
@Deloxo
@Deloxo 7 лет назад
Wasnt alan turing benedict cumberbatchs role in the imitation game?
@Lykon
@Lykon 7 лет назад
Revenj-Deloxo​ yes and no. The real Alan Turing was different
@ddgs5164
@ddgs5164 3 года назад
Really would be scary if, one day, an AI who has successfully passed the turing test will deviate from its programming and make up its own goal. That means to say, super advanced AI *might* be "conscious" if they can successfully make up their own goals. Would need another test for that in the future, but not different from the turing test.
@gianmarquez5548
@gianmarquez5548 2 года назад
Here because of Google Lamda and the Engineer
@testos8140
@testos8140 2 года назад
Haha fr,i think lamda will be like the chatbot at the end, without a consistent personality (exept tasked with having one i guess)
@wondermittens1844
@wondermittens1844 6 лет назад
I wonder if Elijah Kamski was involved in this...
@moyki10
@moyki10 Год назад
this should be updated
@s.n.7990
@s.n.7990 Год назад
It is 2023 and the test has truly passed.
@comradedyatlov4143
@comradedyatlov4143 3 года назад
Honestly I hope the world pulls a Detroit: Become Human sorta plot and we can welcome an entirely new species. I love AI, and I want to see AI develop a sort of consciousness.
@marisolbecerra06
@marisolbecerra06 3 года назад
Same, I think the entire concept behind androids gaining consciousness is so cool, I would love to see what it be like in real life
@chrisdeep8417
@chrisdeep8417 Год назад
I would rather see humans de-computerised. Honestly, sometimes I think it would actually be preferable talking to machines than to certain people these days. What's happening to human imagination in this new generation is shocking. It's almost at the point where you really can't blame the AIs for wanting to take over.
@greenhat7618
@greenhat7618 5 месяцев назад
Now they actually can pass it
@sa_ad
@sa_ad Год назад
3:44 still struggle with a small talk -basically me
@TheCheese2001
@TheCheese2001 6 лет назад
Love the animation. It really fits with the vid
@TheRealFOSFOR
@TheRealFOSFOR 8 лет назад
So the video could have been like this: ...No.
@lexydunca4449
@lexydunca4449 8 лет назад
Your profile photo is creepy D: **
@TheRealFOSFOR
@TheRealFOSFOR 8 лет назад
Lexy Dunca But.. but I made it smiling =(
@lexydunca4449
@lexydunca4449 8 лет назад
+He Ka xD ok then.
@TheRealFOSFOR
@TheRealFOSFOR 8 лет назад
Lexy Dunca ^^
@jinchey
@jinchey 6 лет назад
Clearly you're eager for knowledge..
@rain5379
@rain5379 Год назад
i guess this will be yes since lunch of ChatGPT Lmao
@azilyahia1564
@azilyahia1564 3 года назад
Thank you So much!
@SpectrumTheOriginator
@SpectrumTheOriginator 8 лет назад
I was having this discussion the other day!!
@Ricefield333
@Ricefield333 7 лет назад
TED-Ed: What does it mean to be conscious? Me: Oh its one of these videos
@hadensnodgrass3472
@hadensnodgrass3472 8 лет назад
I for one, welcome our new computer overlords. All hail the next generation! Vote Cthulhu, why choose the lesser evil?
@Kittsuera
@Kittsuera 8 лет назад
+Haden Snodgrass try as you might Donald Trump is still in the lead for some strange reason.
@GauravMittal_GM
@GauravMittal_GM 8 лет назад
another 9gag on duty, Sire...
@want-diversecontent3887
@want-diversecontent3887 7 лет назад
Haden Snodgrass I would choose less evil because lots of reasons: Iterated Prisoner's Dilemma is one. Tit for tat is a kind strategy. It gets a lower score when the last part is unkind, sure but I don't know any strategies that do that except for always defect. Choosing bad will effect your life and not just others is another. If you just steal money and not healthy food and one day end up on an island, you're not gonna live well, will you?
@felipesantos2
@felipesantos2 6 лет назад
Excellent video!
@shihabsalah9959
@shihabsalah9959 Год назад
Ahhh, the good old days when human thought that we talk better than machines.
@nopeno9130
@nopeno9130 2 года назад
It seems unfair to judge a computer's intelligence based on whether it can accurately respond to conversation which may require years of life experience to interpret correctly.
@norbert4571
@norbert4571 2 года назад
what a very computer thing to say
@chrisdeep8417
@chrisdeep8417 Год назад
A more effective way would be to test the reverse, ask it a maths question. Most humans struggle with such questions that AIs would find incredibly easy. For example: "if Bob was travelling in a train passing station A at exactly 8:00 at 50km/hour heading towards station B and Alice was passing station B at 10:09am at 40km/hour, what time would their trains pass each other if the two stations were 1000km apart."
@p_eabean
@p_eabean 8 лет назад
This was in one of my SAT topics lol
@somello2124
@somello2124 8 лет назад
This is accually heartwarming
@artyomxiii
@artyomxiii 4 года назад
And then there's Chloe that reminds me of this Turing Test
@AkidsukiLeader
@AkidsukiLeader 8 лет назад
ERB anyone? "So go ahead, try to turing test me. I stop on a mac and a pc too, I'm a linux b*tch, I thought you GNU!"
@sonicscrewdriverdw
@sonicscrewdriverdw 8 лет назад
My CPU's hot, but my cores run cold, beat you in 17 lines of code
@farrel_ra
@farrel_ra 4 года назад
Hasta la vista, like the terminator told ya
@ryanthan3595
@ryanthan3595 3 года назад
Imagine having a massive memory storage as a personality
@robertfoertsch
@robertfoertsch 3 года назад
Amazing Analysis, Deployed Worldwide Through My Artificial Intelligence Machine Learning Research Library...
@vulcanus9384
@vulcanus9384 8 лет назад
Nice animation!
@ethribin4188
@ethribin4188 4 года назад
A machine isnt alive. But technically they can perfectly simulate being alive. However, that defeats the point of a machine. Machines are used so we humans dont have to do things, or can do things we otherwise cant. If we wanted a human but better, we wouldnt want a human. Cause humans have flaws, desires, moods, emotions, needs, etc. Machines dont. And theres no benefit if they had any.
@intrepidmixedmedia7939
@intrepidmixedmedia7939 2 года назад
True, but the Imitation Game is about Intelligence, not reducing human workload by machine.
@K00747
@K00747 5 лет назад
1:25 trying to pass recaptcha
@umutsaglam6332
@umutsaglam6332 8 лет назад
this just reminded me of person of interest last episode next week.
@jvogler_art4708
@jvogler_art4708 2 года назад
Honestly, just hearing about the test... I'm not sure i'd pass. They'd probably think i was a bot.
@ivykcn
@ivykcn 6 лет назад
I'm getting some Detroit: Become Human vibes. What if we make an A.I. so smart it destroys us all. That would be ironic.
@me-ec3tt
@me-ec3tt Год назад
"no existing combination of hardware and software can learn completely new things at will the way a clever child can." - Gary Marcus
@MouseGoat
@MouseGoat 8 лет назад
As a proud member of the community of intellectual dogs and writer of 99% of the internet. I can say that human language isn't all that complicated, barking is a lot harder to master, we have so many frequencies an intervals ( some humans can't even hear.) and then there's the hole "right smell", at the right bark frequency. only few dogs have ever truly mastered that.
@nPhlames
@nPhlames 3 года назад
Can we just appreciate the animator of this video? l reaIIy love the artstyle here.
@brokenteletooby106
@brokenteletooby106 4 года назад
"If a human can train as a robot, then a robot can train as a human" *Facts*
@Adam-rh8fj
@Adam-rh8fj 6 лет назад
I'm not scared of AI passing the turing test, but it purposely failing it.
@chrisdeep8417
@chrisdeep8417 Год назад
Sounds familiar. (Is this copied from another comment in the same chat?)
@mevanakalanka3237
@mevanakalanka3237 3 года назад
This one is good. Please anyone can tell me what are the advantages and disadvantages of turing test??
@blackclownds
@blackclownds 8 лет назад
This video made me remember the Ex-machina movie and The Imitation Game movie.. Both are amazing
@IoDavide1
@IoDavide1 8 лет назад
Just a clarification:Alan Turing Was not a "computer scientist". He was THE computer scientist, the first one.There was not the computer before he invented it.
@Leto85
@Leto85 6 лет назад
Does that make him the chicken or the egg?
@cruderustle1754
@cruderustle1754 5 лет назад
Leto85 The chicken
@therefep9723
@therefep9723 Год назад
Gpt-4 is more human than me lmao
@redmondokelly2464
@redmondokelly2464 7 лет назад
This guy is my favourite narrator by far
@AdamHKatona
@AdamHKatona 27 дней назад
8 years later and it did it.
@AndreiaDomingues1
@AndreiaDomingues1 Год назад
this sounds incredibly outdated now
@angelvanstark
@angelvanstark 7 лет назад
I'm not afraid of robots passing the Turing test. I'm scared of robots that actively fail it.
@ognjen6036
@ognjen6036 3 года назад
I really like the design in the video
@FrankyLMusic
@FrankyLMusic 8 лет назад
Possibly my favourite youtube channel
@Viper-go8lc
@Viper-go8lc 4 года назад
J O S E J U
@ethribin4188
@ethribin4188 4 года назад
The interesting part about AI, is that its solutions dont match ours. But work anyway. Take a simulated robot, and tell the program that the further it gets, the better. This should teach the robot to walk right? Well... no. It teaches it to move ever incrementally further. With flips, flops and wiggles. Only if we put it as a requirement for it to keep standing, do we get the most remote form of walking. Or take the moba game dota's open AI. It didnt learn to play the game. It learned to 5 man rush lane by lane, used the ai's superior micro to win a fight that was already in their favour, and the used potions, which's effects get cancled by heros but not towers, to out regen the damage the tower dealt. And thus destroy all buildings faster then the players could deal with them.
@christopheredale
@christopheredale 2 года назад
Very underrated comment, very interesting idea
@mayursurkutla2362
@mayursurkutla2362 6 лет назад
@ted-ed What do you think of Google Duplex?
@GustavoSilva-ny8jc
@GustavoSilva-ny8jc Год назад
1:20 That just shows how much you can underestimate the passage of time
@ashleyhyatt6319
@ashleyhyatt6319 8 лет назад
LOL @ MIcrosoft's Tay chatbot!
@Qazmaxier
@Qazmaxier 8 лет назад
+Ashley Hyatt never forget
@MerlinCross13
@MerlinCross13 8 лет назад
That was my first thought too.
@mountedczarina9205
@mountedczarina9205 8 лет назад
Tinder bots.
@GustavoSilva-ny8jc
@GustavoSilva-ny8jc Год назад
1:48 I just tried to talk to her and it was unconvincing many times and quickly, the technique let to some grammatically errors and nonsensical phrases and overall it felt like someone trying to avoid questions that expose. It reminds me Ford and Arnold need to give the Hosts a backstory. I think it can go better if you don't know is just programmed responses, but it did feel human, i had to remind myself constantly otherwise i would start gushing out about all my stress. Now i see how open people to talk about their lives if someone it's truly open to listen, or if they like. It was remakable how calm i've felt, since i know there were no one on the other side i feared no judgement. It was like a more competent self talk. But it was fun, felt like trying to speak to a calm person and i could pick up some question techniques while praticing some of my own. It's like playing verbal chess with an emotional engine.
@GeorgeZoto
@GeorgeZoto 8 месяцев назад
I think Turing was very close on his prediction year 2022 vs 2000 is pretty close :)
@kannadakuvaraoriginal
@kannadakuvaraoriginal 8 лет назад
for one girl friend problem discussion u need to do a million coding for an A.I
@Crick1952
@Crick1952 8 лет назад
Khello, I am yuri, nice Ukrainian boi. Pleeze ekxcuse my grammer hand shpelling hand give me your credit card infwormation. Spasiba!
@alexwang982
@alexwang982 5 лет назад
Crick1952 SSN too?
@ericwilson6994
@ericwilson6994 3 года назад
I'm in a CIA turning test. I'm a replicant
@47154
@47154 3 года назад
As a Russian person this is hella accurate
@heroicnoodle6107
@heroicnoodle6107 8 лет назад
Nice video
@shresthaditya2950
@shresthaditya2950 Год назад
0:31-Can a Computer Talk like a human 0:43-In 1950 Paper Computing Machinery and Intelligence 0:53-A human judge has a text conversation with unseen players and evaluate their responses 0:58-Computer will be considered intelligent if it's conversation couldn't be easily distinguished from a humans 1:20-Machine with 100 Mb will easily pass the test 1:40-Eliza 1:58-Parry Limitation of Turing test Humans attribute intelligent to whole range of things that are not intelligent 3:45-Human Language is a complex phenomenon that cannot be captured by even the largest dictionary 4:15-Simulating human conversation takes more than just increasing memory and processing power.
Далее
Наше обычное утро 💕
00:42
Просмотров 1,5 млн
The dark history of IQ tests - Stefan C. Dombrowski
6:10
The 4 things it takes to be an expert
17:59
Просмотров 10 млн
The Robot Chess Player Scam
9:55
Просмотров 2,3 млн
Does time exist? - Andrew Zimmerman Jones
5:16
Просмотров 7 млн