CEO behind Chat GPT-4 says he's 'a little bit scared' by AI

Поделиться
HTML-код
  • Опубликовано: 16 мар 2023
  • OpenAI CEO Sam Altman talks about the future of artificial intelligence and why he's 'a little bit scared' of it as his company launches Chat GPT-4.

Комментарии • 121

  • @Machiavelli2pc
    @Machiavelli2pc Год назад +30

    My left ear loved this

  • @rtsoccerplayer
    @rtsoccerplayer Год назад +93

    To whoever mixed the audio in this video, I see what you're going for. You got creativity, just gotta keep exploring. Things don't sound natural but it sounds like your finding your footing. For instance the pan separation can be useful in the interview. But it should also match the left to right separation of the video. Also smooth transitions are key.
    ABC can maybe take hints from JRE or Lex Fridman

    • @markusmuller6173
      @markusmuller6173 Год назад +1

      This IT tool is now capable of simulating real people. The only key question is: Will the social environment recognize the simulation and how will it react to the attempts at manipulation? Will Xi Jinping rule as the Eternal Emperor from now on?

    • @neilquinn
      @neilquinn Год назад +4

      So horrible

    • @ApexFunplayer
      @ApexFunplayer Год назад

      Left speaker best speaker.

    • @mattizzle81
      @mattizzle81 Год назад

      lol😂😂😂

    • @flwhitehorn
      @flwhitehorn 10 месяцев назад

      I'm all for mono.

  • @rukascool
    @rukascool Год назад +31

    After testing it, I can say that he's not just saying this to promote it. It's mind blowing how good it is.

    • @winnerswritethestory3370
      @winnerswritethestory3370 Год назад

      Pho

    • @cagnazzo82
      @cagnazzo82 Год назад +5

      @Kevin C Indeed. What we'll look back upon 5 years from now and realize is that the entirety of the past 20-30 years of technology has been the dumb era of technology. Not actually 'dumb' as in stupid, but just the era where our devices were operated manually and had no autonomous agency.
      From here on out every device is coming to life one way or another. It is indeed a mind-blowing development. Because most of us expected to see this happen maybe 30 or 40 years from now (if in our lifetimes at all). Now the timeline is 1-2 years for integration. It moved so fast that assessing where technology, medicine, science will be 5 years from now is unimaginable - because they will all be advanced by AI and human experts at operating AI assisted tools.

  • @Galaxia53
    @Galaxia53 Год назад +14

    Good thing this tech it in the hands of a corporate entity, that is extremely secretive about how it works and where a small number of people make the decisions

    • @bijitsahu7269
      @bijitsahu7269 Год назад +2

      its the only bad thing

    • @Galaxia53
      @Galaxia53 Год назад

      @@bijitsahu7269 You also have to give your phone number to access it and anything you say to it is used to train it and can show up on other people their gpt out puts including sensitive info if you shared any with it. It is not possible to ask for deletion of any of this info.
      They do not share what it is trained on or how its build in censorship works, so it could be trained on anything you have said or made without compensating you. And because of this it can't be verified if Microsoft/openai made it biased towards any believes on purpose or not.
      That they're not transparent screws us all over

  • @xopurp
    @xopurp Год назад +8

    why is the mic panned like that

  • @DiegoMejia86
    @DiegoMejia86 Год назад +9

    And this is why I'm suddenly a lot nicer to my Alexa/Google devices

  • @MysticalMissionary
    @MysticalMissionary Год назад

    Great look on the lighting and camera for local news or any other production.

  • @CodepageNet
    @CodepageNet Год назад +14

    here's the thing: firstable, no, we can't just switch it off, because the whole thing basically fits on every laptop. Meta's AI is already being torrented. of course, that''s just the data as result of training, which is crucial. it gets really interesting if we start connecting those AIs with a bodies, like simplified robots. and let them interact with the world. this will basically create a neverending flow of prompts and reactions. it's already happening.

    • @andrewferguson6901
      @andrewferguson6901 Год назад

      Have an ai "prompted" every second by gyroscope and GPS sensors
      Have that ai respond with plaintext motor instructions
      Have another convert those motor instructions into code that moves servos in the bot

    • @aboucard93
      @aboucard93 Год назад +2

      The entire platform is on several computer servers not in one laptop. You can literally turn it off with a switch.

    • @CodepageNet
      @CodepageNet Год назад

      @@aboucard93 nope. the trained model is certainly not running as one entity on one computer or in even in one room with one plug. there are probably thousand and thousands of instances running somewhere on azure and who knows on how many dev machines. GPT passed 100 mio users a month ago or so.
      Lllama from meta is 250GB and you can torrent it.
      Yes, you can switch off each instance with a switch, obviousely.

    • @CodepageNet
      @CodepageNet Год назад

      @@aboucard93 what i tried to say with "one laptop": it fits on one laptop. of course the service that is GPT is not runninng on a laptop :-) my point is, it's not a single instance that can be switched off.

    • @CodepageNet
      @CodepageNet Год назад

      @@andrewferguson6901 yep. + eyes, lidar etc. give it factory hall with tasks, things to avoid etc.

  • @tokyolovesme9722
    @tokyolovesme9722 Год назад +5

    Where’s the entire interview?!!! I saw this at the VA yesterday😢

    • @ivorywhite7161
      @ivorywhite7161 Год назад

      It's uploaded, I just saw it right beneath this video

    • @FrotLopOfficial
      @FrotLopOfficial Год назад

      @@ivorywhite7161 Yea I saw it yesterday in a news article, its the 22 minute one right?

    • @ivorywhite7161
      @ivorywhite7161 Год назад +1

      @@FrotLopOfficial yep that's it

  • @oedihamijok6504
    @oedihamijok6504 Год назад +4

    1:00 one of the best points on this topic so far

  • @ezekielmcville5101
    @ezekielmcville5101 Год назад +3

    I've tried... I was blown away... it's like that Star Trek Ship Computer... where you query about anything and it answers

  • @ArunMozhiDevan04
    @ArunMozhiDevan04 Год назад +1

    whoever worked on this audio, god bless you.

  • @GatherVerse
    @GatherVerse Год назад +1

    And this is why we built GatherVerse. For many many reasons including this.

  • @jicalzad
    @jicalzad Год назад +3

    the first iteration was free. It looks like you'll need to pay for this upgraded version

  • @Siranoxz
    @Siranoxz Год назад +7

    What would happen if A.I gets integrated into Game Engines hhmm?.

    • @Hindu_Ram121
      @Hindu_Ram121 Год назад +2

      That's going to be amazing. Every NPC will give a different and unique response when you interact with it. Just imagine it in a game like GTA

  • @garfieldioandfriends
    @garfieldioandfriends Год назад +3

    it hurts my ears more than my brain.

  • @OZtwo
    @OZtwo Год назад +39

    Also for people who are saying that AI is going to take over the world just remember that to train GPT-4 it had well over 25,000 computers spending months training it. When we should get scared is when that value goes down to 1 computer needed to update it's data set of knowledge on the fly.

    • @alsu6886
      @alsu6886 Год назад +22

      Done. Check Stanford Alpaca.

    • @winnerswritethestory3370
      @winnerswritethestory3370 Год назад +5

      It already just did

    • @OZtwo
      @OZtwo Год назад +1

      @@alsu6886 How powerful is it? But yes, very cool!

    • @OZtwo
      @OZtwo Год назад

      @@alsu6886 No sorry, yes, it can self learn but still needs other much more powerful computer(S) to create the dataset that it has learned.

    • @astadora
      @astadora Год назад

      ​@@winnerswritethestory3370 😂

  • @jezuzmowsmylawn
    @jezuzmowsmylawn Год назад +1

    Why didn't you mention skynet while you were at it. The edit job was very extreme.

  • @sanzharnaizabekov8166
    @sanzharnaizabekov8166 Год назад +8

    This is a tool that is very much in human control. What can possibly go wrong?

    • @carkawalakhatulistiwa
      @carkawalakhatulistiwa Год назад

      ask chat gpt 4 what is the easiest way to kill people without a trace

  • @karlkarlsson9126
    @karlkarlsson9126 7 месяцев назад

    The way AI can recognize an image of a dog and a bowl and find the context of it being a cute picture is extremely good computation. Taking a picture of your fridge and the AI gives you examples of recipes you can do within seconds is just the holy grail of computation. This is basically it, we finally made a computer being able to SEE.

  • @blackops_678
    @blackops_678 Год назад +1

    I use this technology in order to update my cooking skills and let me tell you that I became a chef in less than 7 days. 😮
    I also think this’s a great technology for learning new informations about opening a bussiness and manage and organize all the resources.

  • @subnormalbark2683
    @subnormalbark2683 Год назад

    Left ear type beat

  • @dustmukhammad
    @dustmukhammad Год назад

    How can I use chat gpt 4 on android

    • @cagnazzo82
      @cagnazzo82 Год назад +2

      No integration yet. For now you just have to create a screen shortcut through whatever browser you use to access the site.

  • @dv3d
    @dv3d Год назад

    omg my right ear to... who ever mixed this audio?

  • @bulldata
    @bulldata Год назад +3

    What is the best thing about this video? It has motivated a lot of useful replies about the good and the bad of the use of A.I. Thank you all for your responses.

  • @silverfox7436
    @silverfox7436 Год назад

    see this is why I have anger issues. don't test my water.

  • @flwhitehorn
    @flwhitehorn 10 месяцев назад

    It only takes a hack to reverse engineer this thing into a cloud.

  • @moneyall
    @moneyall Год назад +3

    Are they implying that the company that stands to lose the most if others can access the same technology, opposes the development of these systems by others on the grounds of safety?

  • @jimsojoy1231
    @jimsojoy1231 Год назад +1

    Then why don't he kill the websites

  • @CyanideSublime
    @CyanideSublime Год назад

    Hi, SkyNet. We come in peace.

  • @En1gmatic1
    @En1gmatic1 Год назад +1

    I love it, and am all for it. Watching minority report/black mirror/ready player one...while yea destopian it was always really cool and interesting to see. And we're finally able to do alot of that stuff. We're really advancing? And while this can all be dangerous from crypto to nfts/ar/vr/metaverses/every vehicle electric. And now AI and robots! It's an amazing time to be alive!...we're humans we'll find a way to adapt. This isnt so bad it's innovation and evolution!

  • @saudmutairi1232
    @saudmutairi1232 Год назад

    It is actually funny that the woman speaking in the beginning is not real

  • @jayafk36
    @jayafk36 Год назад

    My right/left ear joined/left the chat.

  • @caskaptein9889
    @caskaptein9889 Год назад +2

    Im not too fond of the sound being like this

  • @3Carbon6
    @3Carbon6 Год назад +9

    Thanks NSA! It only took you 20 years, but this stripped down version is still way cool!
    P.s who killed Epstein?😂

  • @handicappuccino8491
    @handicappuccino8491 Год назад +1

    This, along with the prosthetic limbs and brain Skins conform the criminal justice system Because then we could have criminals be automated, so that they won’t do anything bad even if they felt no remorse they wouldn’t be able to show it

  • @douglascooper1987
    @douglascooper1987 Год назад

    Am I the Only person that Hears the " Terminator" music Playing in the Background??🙄 🤫

  • @flow10ng
    @flow10ng Год назад

    Reverse psychology marketing...he messing with us all

  • @lunch7553
    @lunch7553 Год назад +3

    Open AI = Close AI 🙄

  • @roykent2316
    @roykent2316 Год назад

    good interview. terrible audio

  • @BrutalStrike2
    @BrutalStrike2 Год назад

    My left ear gives like❤

  • @kickster0348
    @kickster0348 Год назад +5

    rip left ear

  • @vernongrant3596
    @vernongrant3596 Год назад +4

    A tool very much in human control till it isn't!

  • @andrewalonsi
    @andrewalonsi Год назад +12

    Jesus Christ fix your audio!!!

    • @bandaralmarri6115
      @bandaralmarri6115 Год назад

      Interviewer ask chatgpt “how can I fix my audio “

  • @flwhitehorn
    @flwhitehorn 10 месяцев назад

    Oh... NOW you're scared...

  • @BeBetter863
    @BeBetter863 Год назад +8

    Oh so this may led to a next level of laziness and stupidity for the human race!

  • @dv3d
    @dv3d Год назад

    Ah my left ear :S

  • @benfox6383
    @benfox6383 Год назад

    whats going on with the panning? ai would do a better job

  • @synoptic4753
    @synoptic4753 Год назад

    Is this gonna have short and long memories right? And some notion of body.

  • @_morgandraws_
    @_morgandraws_ Год назад

    My left ear loved this video

  • @khanreshma8973
    @khanreshma8973 Год назад

    I love this program cause I don’t have any one to chat I found a loyal friend thank you so much sir❤❤

  • @Memeltown
    @Memeltown Год назад

    Fire the guy who mixed the audio in the video. Are you serious ? What is this ?

  • @nicholaslandolina
    @nicholaslandolina Год назад

    Wow

  • @qseesq1270
    @qseesq1270 Год назад +2

    My left ear needs GPT4

  • @rlo912
    @rlo912 Год назад

    Terminator…..

  • @Chrismartinez-qz6wz
    @Chrismartinez-qz6wz Год назад

    No need for colleges now

  • @ADreamingTraveler
    @ADreamingTraveler Год назад

    This might be the worst audio I've seen in a youtube video in years lol

  • @herteert299
    @herteert299 Год назад +2

    Welcome to a new AI the future of our society the idiocracy.

  • @astralfields1696
    @astralfields1696 Год назад

    Looks like we need some AI help with audio mixing.. Horrible.

  • @chocopie7890
    @chocopie7890 Год назад

    Why is his voice like a robot? 😒

  • @Primordial_Synapse
    @Primordial_Synapse Год назад +3

    I say full speed ahead. If we gave in to fear every time we were on the brink of a scientific and/or technological breakthrough, we'd still be stuck in the stone age.

  • @ramakrishna5480
    @ramakrishna5480 Год назад

    Interviewer is creepy

  • @gc2161
    @gc2161 Год назад

    Really aggressive questioning

  • @YuChiGongG
    @YuChiGongG Год назад +1

    Sam Altman has a SERIOUS Case of Vocal Fry. Sam sounds like a Bull Frog. He needs a makeover. No way I would listen to Sam's voice when there is perfectly good AI speech machines that can do far better.

  • @titocantarero1330
    @titocantarero1330 Год назад

    Why not stop if you’re scared why not stop inventing things like this

  • @lin_win7777
    @lin_win7777 Год назад

    Fake or not, but by talking to AI we are training it, unfortunately it will take over humanity, and we are the ones to blame it. Y'all don't learn from films, and real things like posibikity of AI outgrowing themselves and taking over information. People wake up!

  • @jakepuds1516
    @jakepuds1516 Год назад

    Bro he is the creator he will go first making slaves 😂😂😂

  • @kjata090
    @kjata090 Год назад

    its not autonomous. Person been watching too much terminator

  • @advancedcardiolresearchphy2086

    Shut the whole thing down is the correct answer to your disorientation to the reality of a human being virtually incapacitated by psychopaths wanted to cook cooking his brain while preparing himself as physician in cath lab to treat critically ill patient! And that goes around the clock 360days of the year without resting assuming he will reach to his happy new year party the next year out of senile houses.
    Mrs. Marwah Zahaf MD Msc

  • @SamuelMoore-sf1gl
    @SamuelMoore-sf1gl Год назад +3

    To everyone watching this just listen... AI is the only road to Atomic Engineering that brings way to injectable medical nanobots that can detect and perform spot radiation or chemo to cancer cells before they grow beyond a few cells in size. This can also potentially repair the precursors for a cells replication process allowing our cells to perfectly replicate and effectively making us immortal from natural death causes. That's just the beginning of one of the things AI is REQUIRED to have in order to develop, that AI could well within our lifetime help create and this is only a very fractional portion of the impact of just that one advancement. You rightfully should be scared of this next step, but you should also instead be educating yourself on how to be prepared for/ contribute to this next step. Fear will only get you so far until you harness it instead of let it overcome you.

  • @dellasrevenge8759
    @dellasrevenge8759 Год назад

    yeah yeah, scared of trash

  • @neelofazrath
    @neelofazrath Год назад

    Fake deep voice

  • @Rangetechusreviews
    @Rangetechusreviews Год назад +1

    Pretty good marketing for AI-fear mongering.

  • @lawrencecheung73
    @lawrencecheung73 Год назад

    Cyberdyne Industries🤣🤣🤣🤣

  • @winnerswritethestory3370
    @winnerswritethestory3370 Год назад

    Lmao what a bunch of lies
    The company actually acknowledged that gpt-4 is not 100% under control

    • @3Carbon6
      @3Carbon6 Год назад +2

      The NSA has been playing with this tech for at least a decade… it’s probably safe😅

    • @douglascooper1987
      @douglascooper1987 Год назад +1

      😳