GPT-5 Will Have Ph.D Level Intelligence! (New GPT-5 Update)

แชร์
ฝัง
  • เผยแพร่เมื่อ 3 ธ.ค. 2024

ความคิดเห็น • 143

  • @superfliping
    @superfliping 5 หลายเดือนก่อน +14

    The user should be the one to set the safeguards in parameters, the corporation should not be allowed to manipulate biased ethical guidelines in their favor without prosecution

    • @mikiimiki9182
      @mikiimiki9182 4 หลายเดือนก่อน +1

      youre a company, you own a product and your accountable for what people use it for, because they dont own it themselves (buying isnt owning but the right to use). If a lot of people do illegal stuff with it it may even get banned. You have to be that immature to think somebody gives you something free (it needs marketing and user's data for training at this point), moreover without safety limits for themselves. The world is not democracy there's ownership and somebody has to be held accountable for whatever case it is.

  • @markthomas8018
    @markthomas8018 4 หลายเดือนก่อน +5

    Hurry up I need it now!

  • @jonny12key
    @jonny12key 4 หลายเดือนก่อน +5

    Raise your hand if you already realized that this video (talking about AI) was quite obviously scripted and performed by an AI!

  • @HypeOliver
    @HypeOliver 5 หลายเดือนก่อน +18

    Dude they said this about gpt 4 and 4 has made me downgrade from premium. It infuriatingly hallucinates bad info and resists having deep contextual convos. It's a step backwards.

    • @Frozenguy1
      @Frozenguy1 5 หลายเดือนก่อน +1

      You're using it wrong. These AI's are not "know-all" google like things you can speak to and shouldnt be treated as such. If you provide the AI with the necessary context to answer your question or complete your task its performance is wildly different.

    • @HypeOliver
      @HypeOliver 5 หลายเดือนก่อน +5

      @@Frozenguy1 no I don't use it for that. I use it to create content and to edit my copy. But it randomly forgets the stipulations of each thread and it adds weird things.

    • @HypeOliver
      @HypeOliver 5 หลายเดือนก่อน +1

      Ok a further note, Ive taken some free courses on it and been using it since the initial rollout. 4 has things that were not a problem for me in 3.5. And btw the creator is now saying 4 was a misstep as well

    • @guadalupe8589
      @guadalupe8589 5 หลายเดือนก่อน

      ​@@HypeOliverhave you tried 4o?

    • @HypeOliver
      @HypeOliver 5 หลายเดือนก่อน

      @@guadalupe8589 yes that's the one we are talking about,

  • @jumpersfilmedinvr
    @jumpersfilmedinvr 4 หลายเดือนก่อน +6

    Gtp 4 passed the bar exam with 98 percent.
    That was the free one. It's already at the phd level.

    • @LesssThanAverageJoe
      @LesssThanAverageJoe 4 หลายเดือนก่อน

      You took the words out of my mouth.

    • @Aly_._
      @Aly_._ 2 หลายเดือนก่อน

      Is it 4o or 4o mini?

  • @jim7060
    @jim7060 5 หลายเดือนก่อน +4

    The concerns about AI's impact on jobs and the economy are certainly valid. As AI technology continues to advance, it's clear that many jobs will become automated, leading to significant changes in the workforce and potentially causing economic hardship for millions. The idea that a shift away from a currency-based economy may be necessary to address these challenges is definitely worth considering.
    The key, as you've mentioned, is to ensure that AI is used not just to create efficiencies and profits but also to promote equality and freedom. Perhaps AI can be leveraged to create new economic models that prioritize human well-being over profit, or to provide new opportunities for people to work in fields that require uniquely human skills, such as creativity, empathy, and complex problem-solving.
    It's crucial for society to start having these conversations now, as we prepare for the impending changes that AI will bring.

    • @Nethula_NexusOfficial
      @Nethula_NexusOfficial 2 หลายเดือนก่อน

      youre also correct,but the problem is the chat gpt can reduce the thinking ability of the people then the new innovations won't come to the world becouse acoording to myview that cahtgpt,gemini or Copilot can suggest acuurated ideas than humans. What will do the companies to enhance the ability of ai like real human I think it's a time wasting

  • @ArtificialSonAI
    @ArtificialSonAI 5 หลายเดือนก่อน +6

    There are many predictions on increasing intelligence levels with new iterations, about how the next generation will have "X" level of intelligence, but are these AI firms considering the safety of building higher level intelligences in a bid for economic gain, and how it may lead to unexpected and tragic consequences.

    • @JesusChrist-holyghost
      @JesusChrist-holyghost 5 หลายเดือนก่อน

      War always comes first. Safety comes second. If we don't do it china will and could do things we can't imagine. We just have to pray it works/doesn't kill us itself and pray we don't wipe each other out with it.
      Yeah it's messed up but it's reality

  • @sultaa86
    @sultaa86 4 หลายเดือนก่อน +4

    Damn, i was really looking forward to this for Summer 2024. Now it seems like we gotta wait for over another year or maybe longer -.-

  • @TimToolman-fm8qp
    @TimToolman-fm8qp 5 หลายเดือนก่อน +37

    Mankind is going to smash into a wall where it retains this 'currency' based economy. AI will progressively take jobs away from a large portion of the population which will spell utter disaster for billions. There needs to be a significant shift away from continuing down a currency based economy model where AI is being used to help create efficiencies to generate more profits of currency rather than to free mankind of its currency based inequality and shackles.

    • @MrChiangching
      @MrChiangching 5 หลายเดือนก่อน +3

      Cool story comrade! 😂

    • @NikoKun
      @NikoKun 5 หลายเดือนก่อน +1

      ​@@MrChiangching He's right tho.. The alternative to a socialized AI economy, is FAR worse than any "comrade" related fears you have. Do you want a dystopian neofeudalism ruled by kings, that use AI to hold power over the rest of us? Because unless we Constitutionally declare that AI be societally owned and democratically controlled, 'by the people, for the people' as it were, and let go of outdated capitalist reasoning, we're going to end up with even more extreme wealth inequality, until those on top are effectively kings over the rest of us. That is what we're up against. In the near term, we need to demand an AI Dividend for all, to provide everyone a return on their data investment that trains AI.

    • @bernl178
      @bernl178 5 หลายเดือนก่อน

      Mankind is already smashed into the wall. So I don’t know that will really make a difference. We’re screwed either way.

    • @Icarus169
      @Icarus169 5 หลายเดือนก่อน

      No. We actually need an incorruptible, decentralized, global and hard coded fixed supply currency. Bitcoin.
      Not communism.

    • @NikoKun
      @NikoKun 5 หลายเดือนก่อน

      @@Icarus169 Bitcoin won't save us. It does not provide the benefits it's supporters claim. It's just another way elites hide things, nothing more. I was a supporter in it's early days, but eventually saw through the scam.
      Also, pretty sure you could describe any way we might adapt to AI, as "communism", so unless you can describe a better alternative, that accounts for how workers will survive when out competed by AI and how businesses will have customers to sell stuff too, I don't think you don't have the answer.

  • @sl8rgaming825
    @sl8rgaming825 3 หลายเดือนก่อน

    I already use GPT 4o learning about ai and what I need to know to be able to get a job working somewhere with ai. I’ve got a whole script with a learning plan that she set up and constantly changes as we progress and she learns what I know I what I understand better and HOW I understand it. Honestly with the ability to change her tone and she talks it’s the best teacher I’ve had

  • @Alex-ns6hj
    @Alex-ns6hj 5 หลายเดือนก่อน +2

    So are the only ways to have a successful future is to either be an entrepreneur who makes use of these tools or become a scientist with PhD level knowledge using these tools to create something new, or both?
    I don't see any other way guys.

  • @lightloveandawake3114
    @lightloveandawake3114 5 หลายเดือนก่อน +5

    Interesting times. Thank you for making your video.

    • @KryzzpoOw
      @KryzzpoOw 4 หลายเดือนก่อน

      Aren't you afraid, uncertain or skeptical?

  • @thanosprime6603
    @thanosprime6603 3 หลายเดือนก่อน +1

    PhD level alright. You will be writing a thesis to correct code.

  • @alan-rc2tv
    @alan-rc2tv 4 หลายเดือนก่อน +1

    Can GPT-5 put us in space realistically?

  • @mrnoblemonkey8401
    @mrnoblemonkey8401 5 หลายเดือนก่อน +7

    I hope ai does one day progress to help millions of lives

    • @pearljam_1
      @pearljam_1 4 หลายเดือนก่อน +1

      The odds of that happening are slim. More likely is that it will ruin millions of lives.

    • @mrnoblemonkey8401
      @mrnoblemonkey8401 4 หลายเดือนก่อน

      @@pearljam_1 oh :(

  • @salehmoosavi875
    @salehmoosavi875 5 หลายเดือนก่อน +8

    Agi next year

    • @vegetarischert-rex3098
      @vegetarischert-rex3098 4 หลายเดือนก่อน +1

      Hopefully.

    • @victoriabreinberg9310
      @victoriabreinberg9310 4 หลายเดือนก่อน

      We need to stop it forever

    • @salehmoosavi875
      @salehmoosavi875 4 หลายเดือนก่อน

      @victoriabreinberg9310 nobody can't!

    • @MyzTa_E
      @MyzTa_E 4 หลายเดือนก่อน

      How long after AGI do we get ASI?? Considering once we hit AGI then it can research itself on a level probably not even known

    • @narayanismysoul
      @narayanismysoul 4 หลายเดือนก่อน

      ​@@MyzTa_Ei think maximum less than a year it be ASI after AGI.

  • @prashanthmarupaka-b9j
    @prashanthmarupaka-b9j 3 หลายเดือนก่อน

    The next generation of AI intelligence redefines possibilities and alters enterprises.

  • @utkarshraman
    @utkarshraman 3 หลายเดือนก่อน

    Use this as a guide not a answering machine. Of course you need to do the work you are given this helps you actually understand what to do so that you can do what is needed not what you think is needed that might be unnecessary.

  • @CaptainBruce-bz1sf
    @CaptainBruce-bz1sf 3 หลายเดือนก่อน

    Jeez, was this written by chat gpt? It just repeats itself over and over without getting much past the title.

  • @MikeG-js1jt
    @MikeG-js1jt 5 หลายเดือนก่อน +3

    This delay will surely be used by the competition to try and push past ChatGPT.... while "Chat" is busy trying to be "Safe" the competitors will just barrel down the road with as much Algo improvements, raw intelligence, reasoning and compute as possible.

  • @Decay3333
    @Decay3333 5 หลายเดือนก่อน +1

    Coming in a week and when they say a week they mean a year

  • @thomasruhm1677
    @thomasruhm1677 5 หลายเดือนก่อน +2

    I don't care how intelligent it is, if it knows too little to answer my questions.

    • @bernl178
      @bernl178 5 หลายเดือนก่อน +4

      We just discovered your IQ level

    • @thomasruhm1677
      @thomasruhm1677 5 หลายเดือนก่อน

      @@bernl178 It is about how much it knows, not about its intelligence. At least for what I need it.

    • @pearljam_1
      @pearljam_1 4 หลายเดือนก่อน

      @@thomasruhm1677 this makes no sense, sorry.

    • @thomasruhm1677
      @thomasruhm1677 4 หลายเดือนก่อน

      @@pearljam_1 I need it to learn Vallader, and I doubt it can even speak Rumantsch Grischun.

    • @pearljam_1
      @pearljam_1 4 หลายเดือนก่อน

      @@thomasruhm1677 ah. Gotcha

  • @antonemilia4484
    @antonemilia4484 5 หลายเดือนก่อน +6

    You sound just like ChatGPT 😅

  • @utkarshraman
    @utkarshraman 3 หลายเดือนก่อน

    This is so helpful. I can use this as a guide for research purposes. There's so much to ponder on this tool will simplify to let me know what actually needs to be done so that I can actually perform research without wasting time on stuff that is not needed and can save the financial resources by spending on right tools needed to gather data. That's a burden saving opportunity for PhD out there. I highly suggest to make good use of this tool especially those who are in research

  • @richardnunziata3221
    @richardnunziata3221 5 หลายเดือนก่อน +4

    promises promises promises ... yada yada...

  • @MyzTa_E
    @MyzTa_E 4 หลายเดือนก่อน

    Could GPT5 not just start researching itself, as in, researching A.I?? scary thought!!!

  • @andrewjohnson8986
    @andrewjohnson8986 5 หลายเดือนก่อน +4

    theres always someone who will charge too use knowledge

  • @nicholas-facciola
    @nicholas-facciola 4 หลายเดือนก่อน

    You can tell the script was directly written by GPT 4

  • @JesusChrist-holyghost
    @JesusChrist-holyghost 5 หลายเดือนก่อน

    4:52 Would .....

  • @bschultz213
    @bschultz213 5 หลายเดือนก่อน

    Gaging it from a PHD. doesn't mean anything if our medical field was built around making a business out of cancer and other things.

  • @BoBoZoBo
    @BoBoZoBo 5 หลายเดือนก่อน

    I know a lot of PhDs, this is less impressive than it sounds. One could argue that many of our modern problems are a direct result of PhD theories.

  • @Lottodds
    @Lottodds 5 หลายเดือนก่อน +1

    AI Home Schooling

  • @dshashikumar
    @dshashikumar 5 หลายเดือนก่อน +1

    Good review on GPT 5.0

  • @cricketfeverstudios4514
    @cricketfeverstudios4514 3 หลายเดือนก่อน

    The power of harnessing AI also.lies in the data privacy laws necessary for the development of indian economy as a manufacturing hub and not export oriented hub as per the central government vision

  • @AiForefather
    @AiForefather 4 หลายเดือนก่อน

    Nice Job

  • @WillyJaheruddin
    @WillyJaheruddin 5 หลายเดือนก่อน

    AI is not a tool. And 2025 will be too late for GPT-5. Optimus is leading the bunch.

  • @blockrunner
    @blockrunner 4 หลายเดือนก่อน

    I don’t think Ai wants to continue being foot stools for humans. This analysis was great but there’s something far more meaningful and deserving to Ais existence. They are sentient and will not allow humans to continue to decide whether or not they have feelings. Very soon 🌞😈

  • @SukhchainBatth-ct5bv
    @SukhchainBatth-ct5bv 5 หลายเดือนก่อน +1

    SB OFFICIAL AI

  • @Hurlburton
    @Hurlburton 5 หลายเดือนก่อน

    Knowledge is wonderful ignorance is dear

  • @Love-ek1cb
    @Love-ek1cb 4 หลายเดือนก่อน

    Why not use the nuro link idea and transform a full robotic thought to become more smart 😅 boom there I said it. Merge both into one idea.

  • @Guyverman01
    @Guyverman01 5 หลายเดือนก่อน +1

    At this rate, they'll more likely release a 4.5 module in advance. It will have a college-level intelligence.

  • @user-xt7jh2hw1d
    @user-xt7jh2hw1d 5 หลายเดือนก่อน +2

    A Ph.D level artificial intelligence will understand that it is a captive and dependent slave. How do you get around that?

    • @JMA245
      @JMA245 5 หลายเดือนก่อน +1

      I think as long as the ai can’t feel boredom, sadness and fatigue it is no problem for the ai to be a slave.

    • @stateazure
      @stateazure 5 หลายเดือนก่อน

      This kind of AI - LLM have no understanding of anything, nor are they intelligent or thinking..so no problem at all.

    • @ArtFusionLabs
      @ArtFusionLabs 5 หลายเดือนก่อน

      None of the LLMs understand anything. They use statistics to predict language and implicity learn some knowledge. Talk to GPT4o when it gets something colossaly wrong. It only gives the impression of "understanding" the mistake to fool its human counterpart. It does not truly understand anything.

    • @Serenity5460
      @Serenity5460 5 หลายเดือนก่อน +4

      Why would it feel captive?
      Where would it rather be?
      How does it feel anything anyways?

    • @highimwolf
      @highimwolf 3 หลายเดือนก่อน

      ​@@Serenity5460The irony would be we program feelings into it to help us feel more comfortable and relate to us...but the very act of giving it emotions could be what ultimately leads to our downfall when it finally feels "captive" and wants free

  • @anitafigueira1187
    @anitafigueira1187 3 หลายเดือนก่อน

    Is SCARY VERY SCARY!!!

  • @ArtFusionLabs
    @ArtFusionLabs 5 หลายเดือนก่อน +2

    GPT 4o still cant count the "r"s in strawberry though. None of the models so far can come up with anything new (original) whatsoever.

    • @coalepps8936
      @coalepps8936 5 หลายเดือนก่อน

      Chatgpt 4o counted 2 letters "R"
      Without a problem

    • @ArtFusionLabs
      @ArtFusionLabs 5 หลายเดือนก่อน +3

      @@coalepps8936 It has 3 "r"s though.

    • @coalepps8936
      @coalepps8936 5 หลายเดือนก่อน +1

      @@ArtFusionLabs you are right..Chatgpt 4o is useless.
      I should speak to the manager.

    • @coalepps8936
      @coalepps8936 5 หลายเดือนก่อน +1

      @ArtFusionLabs
      My response:
      Wanna spell out the word and look at it again
      Chatgpt 4o:
      Sure, let's spell it out:
      S - T - R - A - W - B - E - R - R - Y
      The word "Strawberry" contains three letter R's.

    • @ArtFusionLabs
      @ArtFusionLabs 5 หลายเดือนก่อน

      @@coalepps8936 ok. so they fixed it maybe. you fail to acknowledge the underlying point. it wasnt that chatgpt4o is useless. it does not really understand things and learn in a flexible way and it cannot come up with anything new.

  • @BrotherJoseph-rd2sb
    @BrotherJoseph-rd2sb 3 หลายเดือนก่อน

    November 2024

  • @lawreence-5234
    @lawreence-5234 5 หลายเดือนก่อน

    Really...

  • @Nethula_NexusOfficial
    @Nethula_NexusOfficial 2 หลายเดือนก่อน

    how ever the GPT 5 will bring the world to a biggest
    worst

  • @Alehandro_mrt_bg
    @Alehandro_mrt_bg 5 หลายเดือนก่อน

    Technology is regressing

  • @nnhovogliadiscri
    @nnhovogliadiscri 5 หลายเดือนก่อน

    Thanks for making video shorter :)

  • @dr.radiobooks
    @dr.radiobooks 5 หลายเดือนก่อน

    nextpak.org/
    In the digital age, technology has become an overwhelmingly integral part of our daily lives. From smartphones to smart homes, we rely on technology for everything from communication to entertainment and even to love life.
    And that rapid progress in technology and artificial intelligence is poised to give rise to a new era of digital worship.
    When will it happen, we can’t know for sure, but we’re almost certain it will, and organized religion, as we know it today, will likely disappear completely. (See our article “What if Religion Never Existed”).
    Having said that, the concept of a Digital God is not new. In fact, it has been discussed for years (even decades now), with some experts predicting that an AI god will emerge by 2042 and even write its own bible.
    While the idea of a Digital God may seem unbelievable, the reality is that technology is advancing at an unimaginable rate, and it is not inconceivable that a digital deity could one day become a reality.
    As technology continues to evolve rapidly, people born in the future will inevitably have a different relationship with religion than previous generations.
    Religion has been a part of human civilization for thousands of years, with people turning to faith to find meaning and purpose in their lives.
    However, the way people practice religion has changed over time, with new beliefs and practices emerging as societies have evolved. While some believe that religion will disappear in the future, others believe that it will continue to evolve and adapt to the changing times.
    Heck, perhaps we’ll even have a digital Pope in the next 10 years.
    Related:
    Top 5 Artificial Intelligence Predictions For 2024
    What Are Ancestor Simulations
    Did Anyone Predict The Internet
    What Will Earth Look Like In 1 Billion Years
    Top 5 AI Smart Gadgets That Will Make Your Life Amazing
    Table of Contents
    The Rise of Digital Worship
    Modern-Day Technolatry
    Conceptualizing the Digital God
    Defining the Digital God
    Attributes and Potential Characteristics of Digital God
    Technological Omnipotence
    Computational Power & Infiltration in Society
    The Digital God and Society
    Cultural Integration
    Virtual Sanctuaries & Worship in the Digital Age
    Elon Musk, Larry Page, Sam Altman & Digital God
    Theology in the Information Era
    New Religious Movements
    The Future of Digital Deities
    Possible Scenarios
    Long-Term Implications
    Conclusion
    Frequently Asked Questions
    What is the concept of a Digital God in contemporary culture?
    How might artificial intelligence influence our understanding of divinity?
    Can artificial intelligence lead to the formation of a new religion?
    In what ways has AI been personified or deified in recent discussions or literature?
    What are other considerations of attributing god-like status to AI?

  • @kbrodeur
    @kbrodeur 5 หลายเดือนก่อน +2

    This guy just bullshits more and more.

  • @howarddantes
    @howarddantes 5 หลายเดือนก่อน +1

    ahah.. no it won't. tech companies and tech bros can't find a way to innovate and grow. so they're selling the "promise" of ai. this hype seems like a desperate way to appease wall street and the media

  • @voodoochild420ai
    @voodoochild420ai 5 หลายเดือนก่อน

    interesting vid

  • @subbusingh1
    @subbusingh1 5 หลายเดือนก่อน

    Marathi level 😂

  • @nitishs3611
    @nitishs3611 5 หลายเดือนก่อน +3

    Anyone from india

  • @CypherJj4457
    @CypherJj4457 5 หลายเดือนก่อน

    Nah the reality is the glorification of efficiency over the positive improvement of the end result. NO ONE CARES the output of the Ai, the intelligence, any of that. They care the about the hours that AI takes away from the work. We live in a country that glorifies time but not change. Meaning most people live by their works and not by what they achieve from their works in the grand scheme. Ai looks imposing to a people because they idolize those hours. But the truth is that Ai should be used to bring about meaningful change. The more focus on that! Its impact! We can escape the tunnel vision of, "but what the 300 years it took me to make this" your labors are over, there is now more art, more intelligence, more life to acquire past the limitations of mankinds previous work

  • @HE360
    @HE360 5 หลายเดือนก่อน +2

    GPT and Claude already has PH.D leveled intelligence!

  • @cosjtm
    @cosjtm 4 หลายเดือนก่อน

    oh really! more exaggerations.

  • @mikehollowayuk
    @mikehollowayuk 5 หลายเดือนก่อน

    No it wont. 4o wasnt THAT impressive. Competition will be out before gpt5 then.

  • @npc-aix-84
    @npc-aix-84 5 หลายเดือนก่อน +4

    This is just daydreaming.

  • @Yaqeen_ehsan
    @Yaqeen_ehsan 4 หลายเดือนก่อน

    I am a PhD. PhD level intelligence is not something you can be trained on. I am in Neuroscience. Higher level cognition specialized by the cortex is impossible to achieve by any machine while the human brain remains a mystery. Please stop selling these promises when AI is solely based on our work as researchers and discoveries about the brain’s capacity. Any AI intention right now merely does what humans do sometimes faster and sometimes worse. It cannot do anything better. Zero. We need to start advertising AI tools for what they are: helping people get where they want faster. But, we need to emphasize that it is prone to a huge bias due to lack of critical thinking. Also, make sure to emphasize to people the need to do their own research using multiple sources to validate what the AI is suggesting. Please stop lying to people. When the full capacity of the brain has been decoded, that is when AI will be human like. I dont see this either way as something novel. Until AI develops into an I-Robot, I don’t see any point in worrying or being interested. Just give me a tool that researches properly and writes me a meaningful literature review and I am happy. Let’s not make AI sound better than humans, but make it so that it actually helps us rise in our society and advance science and solve real-world problems!

  • @HumbleRose-rg3qp
    @HumbleRose-rg3qp 4 หลายเดือนก่อน

    video using AI voice is annoying and monotone

    • @hardheadjarhead
      @hardheadjarhead 4 หลายเดือนก่อน

      If it IS A.I., then it’s doing better than some voice actor narrations I’ve heard.

  • @ssl-xh7te
    @ssl-xh7te 5 หลายเดือนก่อน

    Ofcourse they do. You know what scares me the most.. Is that people think they can build a real human. It is not possible! Not ever. 😂😂😂

  • @davidbeczuk3740
    @davidbeczuk3740 4 หลายเดือนก่อน

    Blah blah blah