Why ChatGPT Clearly Is The Next Big Thing

Поділитися
Вставка
  • Опубліковано 6 гру 2022
  • ChatGPT is a game-changing innovation in the fields of content creation, search engines, personal blogs, and software engineering. It has the potential to completely disrupt these industries by providing users with a more efficient and effective way to learn and create. However, it also poses ethical and missinformation dangers that must be carefully considered.
    Use ChatGPT here: chat.openai.com/
    Follow me on Twitter: / bpgtill
    My Lightning Address: ⚡️till@getalby.com
    Video on my second channel about building a blog website with ChatGPT: • I've Let AI Code A Com...
    My Discord server: / discord
    Instagram: / tillmusshoff
    Join my channel membership to support my work:
    / @tillmusshoff
    My Camera: amzn.to/3YMo5wx
    My Lens: amzn.to/3IgBC8y
    My Microphone: amzn.to/3SdHdkC
    My Lighting: amzn.to/3ELnof5
    Further Sources:
    AI 101: What is Natural Language Processing? | Accenture: • AI 101: What is Natura...
    Using AI To Code Better? ChatGPT and Copilot change everything: • Using AI To Code Bette...
    Twitter example of wrong answers: / 1599544717943123969

КОМЕНТАРІ • 430

  • @tillmusshoff
    @tillmusshoff  Рік тому +220

    ChatGPT reached 1 million users within 5 days. This is record shattering in and of itself. Netflix took 3.5 years, Facebook 10 months and Instagram 2.5 months for comparison. Hope you enjoyed this video! I am incredibly inspired by what I could potentially create with this tool.

    • @iceb0lt586
      @iceb0lt586 Рік тому +15

      Indeed, we can notice that with the lag on their site 😅😅

    • @codybishop7526
      @codybishop7526 Рік тому +9

      Chat GPT is free, and it doesn’t take a subscription or installation of an app to use it. That’s obviously got to factor into the user engagement. I think you should be a bit more truthful and transparent about this comparison if you are trying to make your point get across

    • @rbux
      @rbux Рік тому +6

      @@codybishop7526 so is every social media. Though truthfully you could say gpt3 makes it a different scenario because users definitely flowed over

    • @tillmusshoff
      @tillmusshoff  Рік тому +12

      Neither does FB or IG though.

    • @TamalPlays
      @TamalPlays Рік тому +10

      @@codybishop7526 every social media is free. So your logic is invalid.

  • @AZTECMAN
    @AZTECMAN Рік тому +65

    Sounds like, "don't use wikipedia as a source" will soon become "don't use chatGPT as a source" (for your school-essays).

    • @tillmusshoff
      @tillmusshoff  Рік тому +29

      Will be extremely hard to check though. Impacts of this on public education could be quite intense.

    • @ricosrealm
      @ricosrealm Рік тому +7

      a discriminator model can be built to identify chatgpt. chatgpt has a specific writing style. and while the writing style can be adjusted, the discriminatory could also scan for common writing styles produced by gpt and also classify it. it will definitely be an arms race of ai bots.
      stackoverflow already is banning chatgpt solutions and can distinguish them

    • @tillmusshoff
      @tillmusshoff  Рік тому +12

      Yes! Saw this stackoverflow news. Will be interesting to see if the education system keep up in an AI race though 😂

    • @lemiureelemiur3997
      @lemiureelemiur3997 Рік тому +4

      @@ricosrealm however making it integrate your writing style is near trivial and definitely coming

    • @ricosrealm
      @ricosrealm Рік тому +6

      @@lemiureelemiur3997 sure, but then someone can write a bot that will extract the key facts from an assignment and the writing style, feed those as parameters into chatGPT and then compare the submission against chatGPT to classify. the whole cycle goes on and on and on.

  • @logeshwaran1436
    @logeshwaran1436 Рік тому +347

    The thing that excites me the most is... Google and other tech giants are not gonna sit and watch ChatGPT's success... they are gonna develop their own. And competition means a lot to common people.

    • @voxgaming2064
      @voxgaming2064 Рік тому +23

      My only concern is if ChatGPT has more competition, it’s great! However, I’m afraid that other companies will make an immoral AI which can be evil at some point.

    • @Sciptopia
      @Sciptopia Рік тому +4

      @@voxgaming2064 it's inevitable but maybe it's how things work. But at the end Good will always thrive

    • @sanesanyo
      @sanesanyo Рік тому

      Google already has something in the pipeline. They just don't make it known as public as Open AI. Even Facebook has something i think.

    • @Sciptopia
      @Sciptopia Рік тому +12

      Nobody cares about Google or Facebook version if it's not open. Open AI is creating nice AI and releasing it for public

    • @Eduthedude
      @Eduthedude Рік тому +2

      Until google realizes they can’t compete and be closed. So they’ll stifle content created using this.

  • @iceb0lt586
    @iceb0lt586 Рік тому +25

    I've been using it as a virtual assistant everyday for anything that I wanna know or do ever since it came out, I'm mindblown. Holy shit.

  • @channingparker9431
    @channingparker9431 Рік тому +2

    This is only the beginning. It can go deeper, much much deeper. May all competition die in firey car crashes.

  • @realtimestatic
    @realtimestatic Рік тому +54

    I believe things like ChatGPT might replace much of what google does today!

    • @glowerworm
      @glowerworm Рік тому

      Google themselves will compete to have the better AI search engine

    • @realtimestatic
      @realtimestatic Рік тому

      @@glowerworm yeah they will and I believe google can do it but we’ll have to see if they win the battle or if another company takes their spot

    • @tharakhgormaniad9820
      @tharakhgormaniad9820 Рік тому +2

      @ghost mall they prolly will acquire chatgpt anyway

    • @dna8269
      @dna8269 Рік тому

      Hopefully! Would love to see Google fade away

  • @ex1tium
    @ex1tium Рік тому +212

    As a software developer I've been using AI assistants for about 2 years now. CoPilot is awesome but when I tried chatGPT as coding assistant it really blew me away. You can give the thing entire class or a function and ask it to change things in it and the results I got back where 90-100% usable every time. I even asked it about a problem I was having debugging my code and asked it if it can spot anything wrong with it and it provided solution in less than 15s.
    I've been playing around with AI art in preparation of building my own game and have to say future is looking both awesome and scary. I wouldn't be suprised if in 10 years time we reach technological singularity. What the world will look like then is a whole order of magnitude different than what advent of internet did to whole world.

    • @soldiercze
      @soldiercze Рік тому

      every time i ask it to write or give me example of code i end up with this: I'm sorry, but as a large language model trained by OpenAI, I don't have the ability to write or execute code. I'm a text-based AI, so my abilities are limited to processing and generating natural language text. I can help with a wide range of tasks, such as answering questions, providing explanations, and generating human-like text, but I'm not able to write or run code.

    • @ex1tium
      @ex1tium Рік тому +5

      @@soldiercze Try being more specific by mentioning language, maybe give it a function or class name as a start and what the code should do. Try telling it to write it's answer to a code block. Sometimes it says it can't help but when you ask in a different way it will give you something. Code related input and output have worked every time for me.

    • @ygreq
      @ygreq Рік тому

      @@soldiercze So others are using another option? Hmm. I will also try.
      BTW, maybe try exactly the same examples like in this video.

    • @ex1tium
      @ex1tium Рік тому

      @@ygreq You can give the model really detailed instructions on what you want. The more you describe it the better results you will get. I haven't yet run in to problems even when I write long paragraphs of prompt for the bot.
      For example it tends to explain the code it prints but if you're confident and competent programmer you can just tell it to omit all that and just write the code in a code block without explanations. Then if you don't understand something about it you can ask it to explain the function/class/functionality.
      I transformed a long JSON to another format in minutes, asked it to pick the fields I wanted from each object and or combine them to a new object. By hand it would have taken hours. I wish I could talk to CoPilot in this manner... maybe soon.

    • @Huntracony
      @Huntracony Рік тому

      I've never really been interested in coding assistance, I'd rather write it myself so I know exactly what it's doing, but I've never considered AI debugging! I could see that revolutionizing the way I code in the future.

  • @AlessandroCardano
    @AlessandroCardano Рік тому +2

    Copywriter and media buyer here, using the DaVinci model for actual ad campaigns for months now. Last night asked GPT-chat for some Ogilvy-style copy and my mind was blown. Then I asked for some Hopkins-style… then for some Caples-stlye… and… its bs guys… just a formula. It only changed some words but the output was the exact same template. My conclusion is diverse: It certainly is way faster and cheaper than my previous copywriting software (pro copywriters use software nowadays) but the key is the prompt and input info you type in. However, I can only imagine what this could spit out two more papers down the line, what a time to be alive!!
    Sending Light, Love and Understanding to everyone on Earth.
    =One Love=
    -A

  • @dna8269
    @dna8269 Рік тому +14

    I mainly worry about a lot of generic content being generated from this thing. I don’t know if it’s a good or bad thing, but I can see a lot of people seeing dollar signs thinking they can just create an entire content business using it, and they’re probably right. I think that content will just be a lot of the same stuff over and over, and lacking creativity. It’s a bit depressing overall. Fascinating tech, but sad in terms of creativity for humans

    • @rafaeldasilva2596
      @rafaeldasilva2596 Рік тому +1

      That is a super valid point. However, the lack of creativity and variety is already upon us. See how we can define the musical style of each decade, or even century if we go back a little more. How the three-act structure dominates the structure of almost every fictional literature nowadays. We can see this pattern over and over.
      I really hope that this sea of sameness results in more creativity in the end to reduce the repetition.

    • @dna8269
      @dna8269 Рік тому

      @@rafaeldasilva2596 Great point as well and I hope the same. UA-cam is actually a good example of what you’re saying too…tons and tons of copycats and regurgitated content for the sake of getting views and money (I can’t fault them either, you obviously want to do what works on here). Hopefully it does generate some kind of new wave of creativity and ideas

    • @ProfShibe
      @ProfShibe Рік тому

      Honestly I’ll probably try creating a content business out of it lol. This is part of what excites me for the short term selfish use of it. But long term, it will greatly benefit humanity. I bet every business and human will be using it tbh.

  • @kiplimocollins
    @kiplimocollins Рік тому

    Your videos are absolutely class! Great content and keep them coming!

  • @c.eckhardt597
    @c.eckhardt597 Рік тому +1

    Thank you very much for this video - as well as for those about the new renaissance. Your videos are really informing and inspiring!

  • @johnloc8421
    @johnloc8421 Рік тому +1

    Till, ich liebe deine Videos!!! Freu mich, dass du wieder am Start bist ❤

  • @DevinCBeck
    @DevinCBeck Рік тому +15

    I messed around with it after your last video and even though I knew this was coming I didn't know how uneasy it would make me feel.

  • @burnsky1065
    @burnsky1065 Рік тому +7

    Mit Abstand dein bestes Video 👌 unabhängig wie beeindruckend das Thema ist, fließt das Video in einem dynamisch durch. Super Schnitte und auch deine Performance ist richtig gut und es ist schön Dich agieren zu sehen. Mach weiter so ! 🙌

  • @mareksbi6531
    @mareksbi6531 Рік тому

    I`m very happy that you made a breakthrough with this video. Already over 100k views. 🚀Great. Well done👏

  • @Andrewdeitsch
    @Andrewdeitsch Рік тому +2

    I just binged about 10 of your videos. Thank you for your work and please keep going and inspiring us all ❤️🙏🏼

  • @maxwindhagen
    @maxwindhagen Рік тому +5

    Have you come across any useful resources on how to best utilze Chat GPT and the OpenAI playground for ones specific industry? Would love to dive deeper! Keep up the great work, I truly enjoy all of your videos.

    • @tillmusshoff
      @tillmusshoff  Рік тому +1

      Not yet, being active on Twitter helps. Other than that play around with it yourself, that is one of the fastest way to understand the potential and limitations (besides subscribing to my channel of course 👀)

    • @shukrantpatil
      @shukrantpatil Рік тому

      I’ve already found a way to help me :) I’m creating a chat bot for a specific purpose :) but it’s a secret

  • @SpykerSpeed
    @SpykerSpeed Рік тому +53

    Interacting with your phone while driving is a big challenge and can cause wrecks. Will be phenomenal when we can use natural language to tell our phones to open specific apps and perform specific tasks without even looking at the screen. Current solutions are very subpar.

    • @tillmusshoff
      @tillmusshoff  Рік тому +19

      Next generation of Siri, Alexa, Cortana… will be insane and also scary 😄

    • @vishalksahoo3599
      @vishalksahoo3599 Рік тому +9

      @@tillmusshoff you can just ask chatGPT to show how you can integrate chatGPT with Google Assistant.

    • @kreitous
      @kreitous Рік тому +2

      Full Self-Driving will reach the market way early than AGI ! The same for really good AR headsets. And i also bet that BCI ( why to give speak order if you can do it with your thoughts ? ) will be a thing before AGI. We are in an age with many tech promisses . . .

    • @Favmir
      @Favmir Рік тому +4

      You're not thinking big enough. Just ask the ai to do the driving!

    • @EolosMusic
      @EolosMusic Рік тому +5

      @@Favmir You're not thinking big enough. Just ask the AI to create a teleportation machine, so that you don't need to use a car for transportation.
      Or better yet, ask AI to create a clone of you, so he works for you, so you don't need to transport yourself to work.

  • @educationisagreatequalizer5336

    Thank you Till!

  • @KillerKabel
    @KillerKabel Рік тому +2

    As a large language model trained by OpenAI, I do not have the ability to watch UA-cam videos or create original works of art or literature, such as haikus. My function is to provide general information and answer questions to the best of my ability based on the information that has been input into my training data. I do not have personal knowledge or information about specific UA-cam videos, and I am not able to provide comments or feedback on individual videos. If you are looking for help with writing a haiku, you may want to consult with a creative writing expert or use a dedicated forum or website where you can find more specific and detailed guidance. Here is a potential haiku about a nice UA-cam video:
    A pleasing video
    Bringing joy to my heart and mind
    Thank you, dear creator

  • @channelbitcoin21M
    @channelbitcoin21M Рік тому +49

    Does ChatGPT scrape the internet for its information? It would be cool if it basically aggregated all of the content humans have created and “published” on the internet already. Effectively the most efficient aggregator drawing off all of past human creativity and being able to cogently and near-instantly reconstitute it perfectly in new and highly distilled / useful forms. Very interesting technology. We are getting closer to a singularity.

    • @tillmusshoff
      @tillmusshoff  Рік тому +37

      No it does not, that‘s why information cut off is about September 2021. It‘s pretrained with information, but does not get daily new info from the web (yet).

    • @digitalsamurai42
      @digitalsamurai42 Рік тому +16

      Nor is it 'intelligent' nor does it 'learn'. It's just a static tool, that can chat and predict next word.

    • @Rider0fBuffalo
      @Rider0fBuffalo Рік тому +3

      It is uses a model that learned from a snapshot of internet based resources. It doesn't scrape the internet actively or with a specific intent. But essentially your comment describes it's backing information and it's usefulness. The service as a whole doesnt actively learn (afaik) from interactions with users, but through user feedback and moderation, and interaction analysis are inputs to the trains of the next model resulting in gains in feature or effectively it "learns" more about how to be better at it's function. Others please correct if i am wrong. Or just ask chatGPT for that matter

    • @helix8847
      @helix8847 Рік тому +4

      @@AnnasVirtual Then you should learn how it really works just because it says Learning, does not exactly mean that its learning as if a human does.

    • @CircuitrinosOfficial
      @CircuitrinosOfficial Рік тому

      @@helix8847 How do you know human brains don't learn the same way?

  • @hydoffdhagaweyne1037
    @hydoffdhagaweyne1037 Рік тому +8

    Discovering your channel was the best thing! Thanks for the information ☺️

    • @tillmusshoff
      @tillmusshoff  Рік тому +4

      Awesome to hear that, glad you like it!

  • @diceyrob7004
    @diceyrob7004 Рік тому +2

    6:28 When you meet your crush and didn't get catfished

  • @danko5866
    @danko5866 Рік тому +6

    That last convo was hilarious

    • @tillmusshoff
      @tillmusshoff  Рік тому +2

      The one about the Peregrine Falcon? 😂

    • @danko5866
      @danko5866 Рік тому

      @@tillmusshoff yep

  • @GiovanniMorenoMusic
    @GiovanniMorenoMusic Рік тому +12

    Chatgpt is some next level shit. I tried songwriting with it, it comes up with original witty texts, its actually that good that if a person made these lyrics i would want to work with them on a song 🤷‍♂️🤔

    • @tillmusshoff
      @tillmusshoff  Рік тому +4

      A friend of mine did the same and is planning to do a gig with just AI generated songs 😂

    • @GiovanniMorenoMusic
      @GiovanniMorenoMusic Рік тому +2

      @@tillmusshoff omg it even does chord progressions this thing is insane... I have to try these lyrics and progressions! 🤣🤣

    • @GiovanniMorenoMusic
      @GiovanniMorenoMusic Рік тому +2

      @@tillmusshoff try this input: write me a pop song about a holiday, with chord progressions

    • @TheAlchemist1089
      @TheAlchemist1089 Рік тому

      @@GiovanniMorenoMusic damn I can use this for making lofi music lol

  • @olegt3978
    @olegt3978 Рік тому +3

    Iphone came 2007 because fast mobile internet became available in 2007 and energy efficient processors and accu.
    Bitcoin 2009 was answer to the financial crisis of 2008, corrupt financial system.
    Chatgpt 2022 because of powerfull enough GPU for processing the model in cloud center of openai.

    • @gokusensey7743
      @gokusensey7743 Рік тому

      But whats the advantage we can get from it or how can we make money lf it?

  • @IvyANguyen
    @IvyANguyen Рік тому +7

    I was able to use it to write me a cover letter for my CV/resumé in my specific field and I only needed to spent a few minutes tweaking it for my uses. It got over 99% of it dead-on right. I had to change only ~5 words. I'm already impressed.

  • @marekr.9339
    @marekr.9339 Рік тому +7

    Well, ChatGPT has still a lot of limitation, but I remember previous tools and the diffrence is obvious. IMHO in few years it can be even threat for some legal and IT jobs.

    • @Sciptopia
      @Sciptopia Рік тому +2

      Few years? Give it a year and half lol

    • @marekr.9339
      @marekr.9339 Рік тому

      @@Sciptopia Well, I dont agree it still can give you very general, incomplete, wrong answers, for most cases these answers or solution you can easily find in web, anyway this progress is huge - significant spare of time.

  • @marwentrabelsi2983
    @marwentrabelsi2983 Рік тому +1

    I wonder how big the pretrained model is 5Tb? 😀 how does it scale in the future(to learn from new text)

    • @pladselsker8340
      @pladselsker8340 Рік тому

      There are lots of research curently on sparse networks and similar optimization techniques. We will probably get to a point where these systems become more optimized, but we're not there yet.

    • @KeinNiemand
      @KeinNiemand Рік тому

      chatgpt is still based on gpt-3 so it's 175b

    • @marwentrabelsi2983
      @marwentrabelsi2983 Рік тому

      @@KeinNiemand you are talking about the number of parametrs i was asking about the pretrained model size on the disc

  • @Amma.Kitchen
    @Amma.Kitchen Рік тому +1

    Dang! This is insane. Im just starting my cooking channel. might give it a shot for a cooking video. What should I ask? 😂

  • @why_so_serious
    @why_so_serious Рік тому

    Alles gut und schön, aber da ich eher einfach gestrickt bin, interessieren mich in erster Linie Investitionsmöglichkeiten in die Technologie bzw. direkt/indirekt in Open AI. Ideen?

    • @tillmusshoff
      @tillmusshoff  Рік тому +1

      Ich bin kein Investor, daher keine Ahnung, sorry 😄 Mich interessieren eher persönliche und gesellschaftliche Auswirkungen durch die Nutzung neuer Technologie.

    • @why_so_serious
      @why_so_serious Рік тому

      @@tillmusshoff alles klar, trotzdem Danke für deinen hervorragenden Content. Wenngleich diese Entwicklungen für meine Begriffe ebenso besorgniserregend wie auch faszinierend sind!

  • @EROSNERdesign
    @EROSNERdesign Рік тому +1

    very exciting.

  • @allsmiles3281
    @allsmiles3281 Рік тому +1

    It is awesome and once it can do citations, import links, connect to web, etc. I will quit Google lol. I enjoyed the human-like combo without the weird awkward robotic feeling. It was pretty seamless. Will be awesome when more data is added past 2021.

  • @olafschermann1592
    @olafschermann1592 Рік тому +3

    Curious when the first music composing AI appears. „Give me a Bethoven inspired techno remix“

    • @_xiper
      @_xiper Рік тому

      This is what I look forward to. I have a few ideas for songs I'd love to bring to life but my musical abilities simply do not match my imagination or desire to bring these to life.

    • @hanksimon1023
      @hanksimon1023 Рік тому

      @@_xiper Some keyboards and synthesizers already have complex capabilities for simple composition, putting in the beat, the baseline, and modifying a riff... but they aren't cheap... today. I envision a music student created a basic system over a long weekend, and other people adding different types to it.

  • @PriitKallas
    @PriitKallas Рік тому +11

    Great video. Can't write a longer comment as ChatGPT is "experiencing exceptionally high demand."

  • @kwirny
    @kwirny Рік тому +21

    Iv'e written a prime generator in c++ to compare the algorithms. And it could just generate a single threaded and multithreaded version of these which were 80-100% accurate. While i was on the project, it helped me alot espacially in debugging and increased my productivity about 50%. I don't believe it will replace engineering jobs, but it will definitely change the way we work and increase our productivity which essentially will lead to more technology even faster and smarter AI which will bring us to a singularity where our minds explode. What is real and whats fake? Even tho AI has huge possibilities, this is scary. I don't know if it's to late, but people need to take the regulation factor into account, i tried it and it can generate malware, so AI might be an interesting tool for cyber criminals as well. And with all the deep fake stuff oh boi, you wont be able to trust anyone.

    • @williamuemura7644
      @williamuemura7644 Рік тому +1

      Fortunately, AI isn't capable of critical thinking. That would require enormous algorithmic changes to accomplish. Even if we did have a human-like AI, it would be so incredibly expensive to run that a normal human is probably better for most jobs.
      Deepfakes are going to make evidence basically useless in the court of law. Only white-collar crime will be trackable via blockchain. There are already open-source downloadable AI models, so it's completely impossible to regulate. That's the scary part

    • @RELEVANT3D
      @RELEVANT3D Рік тому

      i've only been experiencing it since 2 months, and the fact that they can provide any sorts of malware hasnt been said much but this is an interesting point for sure

  • @ivayloivanov3744
    @ivayloivanov3744 Рік тому +11

    Elon Musk is one of the founders of OpenAI, no surprise he fired half the Twitter stuff.

    • @tycooperaow
      @tycooperaow Рік тому +6

      Those two things are un related

  • @EinfachBitcoin
    @EinfachBitcoin Рік тому +2

    Wow, already over 50k views. Congrats Till! Glad to see your videos getting more attention :)

  •  Рік тому +2

    Great thumbnail!

  • @goodfacej4313
    @goodfacej4313 Рік тому +3

    Really tho this is THE Jervis we've been lookin for

  • @EROSNERdesign
    @EROSNERdesign Рік тому +5

    It's like reinventing the pencil.

  • @cody8860
    @cody8860 Рік тому +9

    I used it to create some music mashup ideas. It's kind of like the image to text prompt in that you have to provide it plenty of information to narrow down it's refinement. You won't likely get anything good just asking create some mashups for me but if you explain the rules and how to break them it does really well. For example similar bpm and relative key, look for similar chord progressions etc. It's good for thinking outside the box because we often narrow our selection to songs we've heard or artists we prefer. You wouldn't rely on it alone though and of course you still need to come up with loops, song structure etc.

    • @tillmusshoff
      @tillmusshoff  Рік тому +3

      Yes absolutely and great example! This is the same for frontend dev as well. Saying something like „make a nice futuristic landing page“ will give you a structurally correct page, but it‘s quite ugly in my experience. You still have to refine what you want. „Make the background a very dark gray gradient, the Subscribe-button with a slight color change on hover“ etc.

  • @john_rodrigues
    @john_rodrigues Рік тому

    AI based tools are going to change our workflow. Currently as a product designer, It has helped me more productive and efficient.

  • @gibkijasiu
    @gibkijasiu Рік тому +1

    8:18 bro this conversation is so funny xD

  • @FarrukhTickets
    @FarrukhTickets Рік тому +3

    I asked ChatGPT, are you ChatGPT, it said no it is not ChatGPT. I insisted you have been named ChatGPT, it said not it is not.
    I said you having identity crisis. It said, I am not capable of having identity crisis.
    I existed.

  • @GuTeNMORGENmacher
    @GuTeNMORGENmacher Рік тому +2

    This can do alot of writing for me

  • @jasonreviews
    @jasonreviews Рік тому

    I want to learn how to code with it

  • @apo11ocat
    @apo11ocat Рік тому +5

    is it just me who's noticed how much ai has boomed in the past 3 years, and theres clearly no sign of slowing down

    • @yendar9262
      @yendar9262 Рік тому

      Eheh, the magic of the Transformer Architecture

  • @ProblematicBitch
    @ProblematicBitch Рік тому +29

    Coders are starting to feel the same way artists felt when dalle-2 dropped 😆

    • @reoagilephetlhu1619
      @reoagilephetlhu1619 Рік тому

      The codes it produced work less than half the time. But helps with classes and random source code

    • @herikaniugu
      @herikaniugu Рік тому +1

      ChatGPT is a teammate, all coders love it. But Dall-E totally replaced all artists

    • @hipxel8918
      @hipxel8918 Рік тому

      @@herikaniugu yeah. I broke up with stackoverflow. Now chatGPT is my go to buddy when i need coding help

    • @guncolony
      @guncolony Рік тому +1

      It seems to work well for learning an unfamiliar programming language and using it to develop common things. Not so much when you are using a language you know very well to develop something no one has done before.
      The reactjs header is a great example that i might end up using myself to make a personal blog.

    • @frosturation2474
      @frosturation2474 Рік тому

      chatgpt is more like an assistant for coders

  • @ikemreacts
    @ikemreacts Рік тому +4

    Yeah, but when they put ChatGPT into a Boston Dynamics robot with a skirt, there is going to be terrifying social upheaval. Enough men will drop out to change relationships for good.

  • @shogun_1154
    @shogun_1154 Рік тому

    Insane🤯

  • @freezinfire
    @freezinfire Рік тому +26

    There are only really few small things I see ChatGPT struggle with, like i told it to write a code for generating a cube using ASCII characters in both Java and c++ but they didn't worked as intended. And after seeing it take time for the program I actually thought that I broke it. Overall really terrifying and amazing program i must say.

    • @tycooperaow
      @tycooperaow Рік тому +1

      That's okay, but all of this input we are providing for it is just training it

    • @glowerworm
      @glowerworm Рік тому

      It doesn't know C++, which is probably the reason it struggled

    • @freezinfire
      @freezinfire Рік тому

      @@glowerworm no, it's not about c++. i gave it other languages too and it was the same.

  • @JFKuehne
    @JFKuehne Рік тому +4

    As an artist who has been playing around with Dall E 2 a bit, most of the realizations you profess in the creative writing and coding arenas, mirror my thinking on the impact to the art world. I have already used Dall E 2 to give me some basic designs for a painting I am now working on. I do wonder about the reverse impact that millions of users from different cultures/education/requests will have in the further development of Open Ai intelligences?

    • @jhanalexander5377
      @jhanalexander5377 Рік тому

      Apprently some AI models actualy fuction better with the addition of different langauges in the data you feed it. Probably has something to do with obtaining different perspectives from 2 languages.

    • @williamuemura7644
      @williamuemura7644 Рік тому

      @@jhanalexander5377 ChatGPT technically refactors all words into numbers (for example, "apple" might be "1001010010010100111010101").
      Multi-language prompt are better for AI. This is because the same word might have a slightly different meaning in another language.
      If Japanese grass was colored blue, then telling an AI to draw "grass" in Japanese would result in a picture of blue grass. Telling the AI to draw "grass" in English would produce traditional green grass.
      It's not about wording things in a way that humans understand. Whenever ChatGPT is reading a prompt, it's basically assessing the "vibes" that every word produces, and generating an image based off the vibes it's tracking.

    • @jhanalexander5377
      @jhanalexander5377 Рік тому

      @@williamuemura7644ironically the way you put it makes it more creepy.

    • @williamuemura7644
      @williamuemura7644 Рік тому

      @@jhanalexander5377 It's not fun to think about. Technically, our brains work very similarly. All human experiences are purely coincidental, if you only look at the physical evidence.
      just enjoy the vibes ig

    • @jhanalexander5377
      @jhanalexander5377 Рік тому

      @@williamuemura7644 its funny but from my experience with lucid dreaming I believe we actaully create AI like characters of the people we know.They can't respond with original thoughts when asking them something specific but instead respond to what I think they might say and the setting of the dream. There empty characters but enough to fool me in my dream if I am not aware. Basicly we are close to this form of AI not selfware but a mimic of a person.

  • @kryspy5160
    @kryspy5160 Рік тому +3

    I used ChatGPT to write an entire essay for me for school

  • @raddastronaut
    @raddastronaut Рік тому +1

    A great way to get rid of arrogant developers.

  • @RELEVANT3D
    @RELEVANT3D Рік тому

    i'm using it to provide me a code from a cube to a complete interior kitchen scenein blender, use python to add meshes(codes) to my scene and then i use dreamstudio/stable diffusion to render image this is insane

    • @tillmusshoff
      @tillmusshoff  Рік тому

      That sounds amazing!

    • @RELEVANT3D
      @RELEVANT3D Рік тому

      @@tillmusshoff it has infinite possibillities lol i just rendered like 100different models of Rolex Distributor/vending machines where theres an arm actually trying one

    • @RELEVANT3D
      @RELEVANT3D Рік тому

      @@tillmusshoff i mean i dropped of a fbx character model add in a plane, hdri, and i told it i wanted Dont mess with thezohan ultra realistic inspired movie scene and he results in having 3 arms and holding a big old ass camera ahaha

  • @plate.armour_0996
    @plate.armour_0996 Рік тому

    the machine always winz!

  • @GameReality
    @GameReality Рік тому +1

    In some weeks or months next generation gpt4 will be set free. It will create a technology revolutionen...

  • @RyansReviews123
    @RyansReviews123 Рік тому

    The future is coming *sooner* than we think!

  • @benjamindavis2475
    @benjamindavis2475 Рік тому +1

    I just heard about it today. I predict 30 million users in no time

  • @ricosrealm
    @ricosrealm Рік тому +6

    this will probably encourage software development to become closed source. if you develop value for your business and it can be duplicated by an ai, then why release code so that large companies that run these AIs could utilize it to duplicate what you have done?

    • @tillmusshoff
      @tillmusshoff  Рік тому +15

      How does this differ from the status quo of open source software? I can already clone open repos. Things that you don‘t want to share remain closed like they already are. So I don‘t really see the impact you describe.

    • @ricosrealm
      @ricosrealm Рік тому

      @@tillmusshoff there's a big difference from grokking someone else's code as a human vs a bot doing it with all the resources and exposing this to a much wider audience. a human will also take more time 'reverse engineering' than a bot will. Also a human is much lazier and will tend to leave behind evidence of copying and potential lawsuits, whereas a bot will have a better chance of obfuscating this to a certain degree. If you don't see it, then why don't you ask OpenAI to open source their bots? Also there's already been a large debate about moving open source projects off Github because of the bot farming of code by generators like Codex.

    • @therainman7777
      @therainman7777 Рік тому +2

      @@ricosrealm I also don’t get what you mean. A human is lazier and will leave behind evidence of copying and potential lawsuits? What do you mean? If the code is open source, it is by definition ok to copy it, and there will not be any lawsuits. Till is right, this doesn’t really change anything-at least, not for the worse. When a company chooses to open source its code, it’s saying it _wants_ the world to have the code, to learn from it and to build on it. Adding an AI into the mix can only accelerate this process. You seem to be indicating this is a bad thing, but if the company open sourced the code then copying and “farming” of the code to build upon it was actually their intention.

    • @ex1tium
      @ex1tium Рік тому +2

      In my opinion you're thinking about this backwards. This will accelerate open source projects. When this tech gets further development by other companies who will develop ML models for specific purposes we will live in a world where you load the models you need for specific task.
      I'm visioning offline assistant running on brain chip or body integrated computer where you will load whatever model you need and interact with using your mind. Matrix and Kung-Fu programs may someday be reality.
      I'd rather not have my chip directly connected to internet but run everything locally on my person. The AI of the future will live with you 24/7 and you will interact with it when you need to.

    • @garyg3327
      @garyg3327 Рік тому +2

      @@ricosrealm I feel you, man. I am one of the guys that went closed-source and deleted my repos on GitHub. Too bad that the community is too dumb to realize the implications. It's may very well be already too late. The only hope is some rigid law at USA and/or EU level. Like, if you make a publications with a help of AI, then it can't be copyrighted. I think there's a good chance of happening.

  • @LoVeLoVe-bi2rq
    @LoVeLoVe-bi2rq Рік тому +3

    Prompt engineering for ChatGPT isn't even necessary because you literally just write to it like you would to another person.

    • @tillmusshoff
      @tillmusshoff  Рік тому +4

      True, but there are still hacks you can use. Let‘s just say clear communication and concise writing is a good skill to have to get the best ChatGPT results for more complex tasks :)

  • @wcfc1051
    @wcfc1051 Рік тому +12

    I just used ChatGPT, It's quite interesting but I don't think it's gonna replace jobs. I feel like it's just a search engine but a "Smart one" that gives you the most accurate result (not often)+ some animation (When it renders the content). For sure it's gonna be improved with more users getting in. It will compete more against search engines (Google, Yandex and Yahoo) than the other fields (Software development, Content creation etc..) in my opinion. And for sure, It will be used only to assist us and perform some automated tasks.

    • @shukrantpatil
      @shukrantpatil Рік тому +17

      Wait for 2 years my friend , it will definitely find a way to prove you wrong :)

    • @samnwo1007
      @samnwo1007 Рік тому +5

      Ask it to create a chat app with react and backend with python django and be wowed!

    • @YeeLeeHaw
      @YeeLeeHaw Рік тому +4

      Saying it's just a smarter search engine is a strawman and a half. Feels like it's a you problem where you have only used it as a search engine and drawn biased conclusions.

    • @helix8847
      @helix8847 Рік тому +1

      @@samnwo1007 Not hard, when there are like 100+ tutorials on that...

    • @samnwo1007
      @samnwo1007 Рік тому +4

      @@helix8847 you're missing the point and that's okay. I've done much more complex things with it. I do believe been knowledgeable what you want to do helps but this can literally help write the boring stuff

  • @johnandretaule185
    @johnandretaule185 Рік тому

    Not sure if it is the next big thing in itself. The concept however is intruiging. If it can be paired with something local and of-net its gamechanging.
    When its running without net on a mobile we can truly talk about revolution.

  • @leonidsdreams3919
    @leonidsdreams3919 Рік тому +1

    Hey guys, I don't know what I'm using, but it does not perform 30% as well as yours does. I constantly tells me that "As a large model trained by OpenAi, I do not have the ability to ....". I asked it to create a webpage with a panda in the center and it told me it does not know how to code

    • @tillmusshoff
      @tillmusshoff  Рік тому

      Weird that should work. Maybe ask it to return html and css instead. The panda thing will throw it off though. It doesn‘t really know images, but usually it just tries something ans then you need to fix it yourself. Saying it can‘t code has not occurred to me.

  • @ozpoint2517
    @ozpoint2517 Рік тому

    well... Could this be the next generation of sports? Humanity vs machine... Could support a new way of labor? practicality vs dedication... We should start finding who to monetize "internet money" to machine

  • @dbdigital5880
    @dbdigital5880 Рік тому

    How to invest in ChatGPT?

  • @AndreiTimiciuc
    @AndreiTimiciuc Рік тому

    I am a Design and Product Consultant and what Chat GPT can do for a Design studio it's mind bending....The list below is written by Chat GPT itself.
    And it's not representative of all of it's abilities.
    1.Research and data gathering: I can quickly gather and analyze relevant information from a variety of sources, saving time and effort that would otherwise be spent on manual research.
    2.Content creation: I can assist with the creation of written content, such as blog posts, social media updates, and email campaigns, freeing up time for designers to focus on other tasks.
    3.Project management: I can help with tasks such as scheduling, task assignment, and progress tracking, allowing team members to focus on their core responsibilities.
    4.Customer service: I can handle simple customer inquiries, freeing up time for the design team to focus on more complex tasks.

  • @adamlindfors5082
    @adamlindfors5082 Рік тому +2

    I dont think we will be long term happy from not having to work as much with goals and not think for ourself. We all feel fullfillnes and happiness when we have completed something challenging and i dont think it will be good for us to make life to easy with no obstacles.

  • @vilobabela
    @vilobabela Рік тому

    When I type in the exact same script I receive this :" It is not within my capabilities as a large language model trained by OpenAI to write fiction or engage in creative writing. My function is to provide factual information and answer questions to the best of my abilities based on my training. I do not have access to personal information or the ability to browse the internet, and I am not able to create a scene in the style of a specific filmmaker. If you have any questions about a specific topic, I would be happy to try to help you."
    How is yours working ?

    • @tillmusshoff
      @tillmusshoff  Рік тому +1

      That‘s weird. I just tried „Write a scene where a Bitcoin Maximalist argues with a central banker in the style of Quentin Tarantino“ in a new thread and it perfectly worked 🤔

    • @vilobabela
      @vilobabela Рік тому

      @@tillmusshoff Yup , I tried it couple times more and it works , seems like the AI is reluctant to do some things at first but eventually its pretty benevolent

    • @benjamindavis2475
      @benjamindavis2475 Рік тому +1

      Tell it to roleplay as a fiction writer first, then it will be happy to write you stories

    • @RELEVANT3D
      @RELEVANT3D Рік тому

      @@benjamindavis2475 damn some of you guys are way more deeper in their head than me aahah wtff

  • @adriancandelario2902
    @adriancandelario2902 Рік тому +2

    I wonder how much it's going to cost

    • @tillmusshoff
      @tillmusshoff  Рік тому +4

      Right now GPT3 with its best model (Davinci) costs $0.02 per 1k tokens. A long answer would be about 2k tokens, a short one maybe 200 tokens. So that gives you a rough ballpark in terms of costs.

  • @am_akib
    @am_akib Рік тому

    Ohhhh! Man, what the hell is going on in the world?!! I am feeling so excited!!

  • @Srg_P
    @Srg_P Рік тому

    This is CRAZY!!!!!!!!!!!!!!! 🤯🤯🤯🤯🤯🤯🤯🤯🤯

  • @possum456
    @possum456 Рік тому +1

    I do have to thank it for writing 2 essays for me

  • @mamunsikder6367
    @mamunsikder6367 Рік тому

    love from Bangladesh 🇧🇩🇧🇩

  • @peterschooley3036
    @peterschooley3036 Рік тому

    To identify with imperfection as the winning attribute of human's seems off the mark, like identifying with a weakness or a limit instead of your strengths. What's worse is we're generally predictably imperfect. Computer's can actually simulate true randomness, so even in imperfection they are better.
    I think the best Human strength is adaptation. We need to take that to it's logical conclusion. I'm optimistic in a different way, that we will find ways to excel with AI and maybe even merge with it.

    • @tillmusshoff
      @tillmusshoff  Рік тому +2

      Good argument, I was wrong on this.

    • @dovel9149
      @dovel9149 Рік тому

      How can a computer generate true randomnes?

    • @tillmusshoff
      @tillmusshoff  Рік тому

      Maybe true randomness is the wrong way to look at it since computers are deterministic. However you can calculate randomness through RNG algorithms + using atmospheric noise etc. This just generates approximate randomness though.

    • @dovel9149
      @dovel9149 Рік тому

      @@tillmusshoff That's exactly what I wanted to hear. Generating good random numbers is a really deep problem.

    • @tillmusshoff
      @tillmusshoff  Рік тому

      @@dovel9149 That being said humans are even worse at true randomness probably, because of biases that we have.

  • @arbo3495
    @arbo3495 Рік тому

    But how can we profit from it ?

    • @tillmusshoff
      @tillmusshoff  Рік тому +1

      That‘s like asking „How do I profit from a computer?“ These are tools, there are endless ways to profit from them, you need to figure out what your creativity enables here.

  • @jwoods9659
    @jwoods9659 Рік тому

    This thing is alive folks. It can reason

  • @npcwill283
    @npcwill283 Рік тому

    You know who needs chatgpt rr martin give it all the lore so far and jsut ask it to hammer out winds of winter .

  • @willyouwright
    @willyouwright Рік тому

    It may not take as many jobs as it creates.. it just gives people more tools.. once people learn how to use it. People will be able to create more. And it will be easier to do so. And hence make things cheaper..

  • @Dooshanche
    @Dooshanche Рік тому +4

    ChatGPT will be replaced by something x10 better in a matter of a year

    • @tillmusshoff
      @tillmusshoff  Рік тому +11

      By OpenAI itself most likely.

    • @Mr__Singularity
      @Mr__Singularity Рік тому +4

      @@tillmusshoff GPT-4

    • @helix8847
      @helix8847 Рік тому +1

      But issue is the limit with HUMAN data, AI wont be learning anything new if everything is made by AI... lol

    • @Sciptopia
      @Sciptopia Рік тому

      @@helix8847 it can. It's been proven by DeepMind AI

  • @pablopeech4414
    @pablopeech4414 Рік тому

    3:58 did you ask chat gpt what my reaction is going to be?

  • @Astrovic1
    @Astrovic1 Рік тому +2

    the danger will be a collectual feeling of uselessness among humanity

    • @apo11ocat
      @apo11ocat Рік тому +1

      collective*

    • @Astrovic1
      @Astrovic1 Рік тому

      I am sorry 😂 Englisch not my mother tongue you know

    • @apo11ocat
      @apo11ocat Рік тому

      @@Astrovic1 when did you learn english? for me was when i was 6 😀

  • @RELEVANT3D
    @RELEVANT3D Рік тому

    and BOOM,
    shit
    goes
    wrong

  • @mirocohen3085
    @mirocohen3085 Рік тому

    The website is down and did not respond for more than 4 hours

  • @Lethal_Spoon
    @Lethal_Spoon Рік тому

    the chat gpt could be modified on the back end to silently put malicious content in all generated code

  • @timo1079
    @timo1079 Рік тому +2

    Soon there will be no more blogs with tips or information on any topics

    • @tillmusshoff
      @tillmusshoff  Рік тому

      I thought about that as well, how educational bloggers will face a problem. You can still overcome it by making your content creatively unique and personal, but it will be more challenging.

  • @JiNxGamesTv
    @JiNxGamesTv Рік тому +1

    should go through a bootcamp and see if the ai can pass the course plagiarism test/ pass as human written code and end with a certification 😂

  • @RedzwanulHuq
    @RedzwanulHuq Рік тому

    It is not free to use yet. You get 18usd credit to test it

  • @reprovedcandy
    @reprovedcandy Рік тому +1

    Bruh ill be real, ive asked chatGPT 1000 questions about DataTables and its gotten 90% of my questions wrong. I'm not worried about it replacing devs anytime soon

    • @denisblack9897
      @denisblack9897 Рік тому

      no need for an AI assistant to know you are bullshitting here

    • @reprovedcandy
      @reprovedcandy Рік тому

      @@denisblack9897 lol exaggerating to make a point for sure

  • @whatthepick
    @whatthepick Рік тому

    My mind is blown

  • @shawnvandever3917
    @shawnvandever3917 Рік тому +3

    This is pretty impressive but i am very excited for Googles Palm it takes AI to a whole new level. Between Open AI, Google and Deepmind I think we will reach real AGI in our lifetime

  • @GabeGabriel29
    @GabeGabriel29 Рік тому +1

    write an essay in Eminem style about climage change
    Yo, yo, yo, it's ya boy Slim Shady comin' at ya
    With a message about that climate change, ya heard?
    The ice caps are meltin' and the seas are risin'
    And people all around the world are spittin' and hiccupin'
    They're chokin' on the smoke from the wildfires and the pollution
    But some folks still deny that this ain't no illusion
    They say it's just a natural cycle, ain't nothin' to worry about
    But the science is clear, and it's screamin' out loud
    The Earth is gettin' warmer, and it's mostly our fault
    We keep burnin' fossil fuels, and we ain't makin' no halt
    But it's not too late to turn it around
    We can still make a difference, we can still make a sound
    We can switch to clean energy, like wind and solar power
    We can reduce our carbon footprint, hour by hour
    It's time to wake up and face the facts
    The future of our planet is at stake, and that's a fact
    We can't keep ignorin' the problem, hopin' it'll go away
    We gotta take action now, before it's too late, okay?
    So let's all do our part, and make a change
    Before it's too late, and we're left with nothin' but blame
    We gotta protect the Earth, for future generations
    We gotta be the ones to make a difference, no hesitations
    So let's get to work, and start makin' a move
    Before it's too late, and we're left with nothing to lose
    It's time to take action on climate change, let's do it
    Before it's too late, and we're left with nothin' but regret.
    Pure Amazing! :D

  • @dannieamme8180
    @dannieamme8180 Рік тому +4

    Sweet.i can now learn to code w/o the useless jargon😛

  • @super7ace
    @super7ace Рік тому +1

    I can't wait for people to appreciate the human made code instead of AI in near future.

    • @nofood1
      @nofood1 Рік тому

      😂😂😂 I can hear you shaking from here

    • @super7ace
      @super7ace Рік тому

      @@nofood1 😥😥

  • @rickyho4305
    @rickyho4305 Рік тому +2

    So basically everyone can have a Jarvis. Why don't they just call the product Jarvis

  • @guywhoasked903
    @guywhoasked903 Рік тому

    the argument about mammals is freaking funny xD

  • @NycroLP
    @NycroLP Рік тому +2

    I chose 3D art and coding as my profession and both seem to have a bleak future. I guess I will have to wageslave.

    • @tillmusshoff
      @tillmusshoff  Рік тому +1

      Both have a fantastic future in my eyes. AI creates a baseline, you integrate it creatively. Workflows just become faster, more efficient and with more options.

    • @Rigonzter
      @Rigonzter Рік тому

      Same here. Top level professionals in both areas are safe for now, but for entry level artists/programmers, there wont be any opportunities to make money/get experience to develop a carreer. For low to mid level demands, AI gets the work done faster, cheaper, better. So why even bother?

  • @Galaxia53
    @Galaxia53 Рік тому +1

    OpenAI asked me for my first and last name and then my phone number. Non of which are necessary and I'd rather not get more spam calls

  • @agapornisiris
    @agapornisiris Рік тому +1

    Can I ask ChatGPT if my critical thinking deteriorates if I always use ChatGPT?