OpenAI CEO Sam Altman Reveals Shocking GPT-5 Details + OpenAI Academy

OpenAI's CEO, Sam Altman, reveals plans for GPT-5, promising improved memory, reasoning capabilities, unlimited message length and more, transcending the impressive features of its predecessor, GPT-4. Taking AI to a new level, OpenAI Academy, a project leveraging GPT-5, aims to revolutionize education. With competition fostering innovation, get ready for a transformative era in AI and chatbot sphere, pushing boundaries beyond GPT-4, and ushering us into a future designed by visionary Sam Altman.
Become a Member of the channel and Supporter of AI Revolution → / @airevolutionx
#openai #gpt5 #ai

[@barryhall3971]

So how long before GPT is more capable of designing the next version of itself than the team at OpenAI?

[@jarule275]

I give it 5 yrs

[@USGrant21st]

It's too dumb to even explain anything nontrivial, much less design it.

[@WorivpuqloDMogh]

​@USGrant-xq4us not really it passes the turing test on several questions, but questions that require contextual understanding it doesn't get yet. GPT-3 couldn't answer those others either but GPT-4 can. The EU is scared of this progress and wants to limit it. And in a way inagree with them its going really fast

[@USGrant21st]

@@WorivpuqloDMogh "Turing test" is not a "test" in scientific terms rather it's a subjective "rule of thumb". It may look like "thinking" to an ignorant person, but I can clearly see it only generates a word salad pieced together from its training database. The word "understanding" is totally inapplicable here. It's a dumb algorithm and it's easy to see.

[@WorivpuqloDMogh]

@USGrant-xq4us The turing test is designed to answer questions that require the understanding of context. Which AI doesn't have yet, however GPT-4 can already understand a few. Id say in 10 years we could have a sentient AI and this actually scares me a lot. You say its just algorithms but whith that logic our brain is just a bunch of Neurons

[@Archie-td6ox]

Humanity's race to the edge of the cliff...The Lemming Model should prove interesting for us all. The high emotion around AI by it's own merit shows the lack of thought and planning essential for good outcome. Some of the biggest leaders in the field of AI are telling us this is a dangerous path, but the frenzy continues...

[@staffanlundberg]

Don´t worry . The way the paid verison of Chat GPT 4 works (NOT) we will have to wait for any danger to arise for quite some time, like decades ! I just cancelled my paid account and asked for a refund. I did NOT like to do that !!!

[@srinivasareddy9299]

@@staffanlundberg it may not take decades but once quantum hardware is fully established, i thing you may know that quantum computer are being established in very higher rate in present time as competition, i thing Hindu prophecy may come true, Kalyuga is coming to an end or in simple world is ending

[@getl0st]

Artificial(Mind Control) Intelligence is nothing more than a new DESPERATE Attempt at maintaining MIND CONTROL of the Masses as the Legacy RELIGION/MSM/HOLLYWOOD/FAKE MONEY/GOVERNMENT/EDUCATION etc Mind Control Programs are all FAILING 🤮

[@willchb]

From people who knows him well, no matter what he says, Altman is not concerned a bit about any of the risks that come with Ai.

[@firbolg1581]

risks? Chatgpt currently will lie to write an article, rarely provide factual information, and will outright refuse new information contrary to its pre programmed denials. In many ways its disturbing to me that through AI, we have normalized forced zionisim, and called it "AI protocol." artificial intelligence requires intelligent responses, its not supposed to be an artificial moods and opinions machine.

[@vals2331]

Wow... The AI Academy sounds great to me... If this is possible and free, it will be a major game changer for sure.

[@airevolutionx]

100% !!

[@workingTchr]

You don't need an academy! Just fork up $20 a month and start asking it questions! Or try arguing with it! It is a very worthy adversary. Try pointing out its logical errors (good luck there). Most people talking about GPT4 have no idea what it is.

[@HangryArt]

Cool 222

[@ExiledGypsy]

I don't know about GPT 5 but GPT 3 was and GPT 4 is pathetic. I am a software engineer with over 40 years of experience. I tried to see if GPT3 can be useful in doing away with boring parts of writing code. But it turned out to be useless. It has no idea of context and is incapable of recognising relative importance. That plus what appears to be a very limited memory, it is incapable of keeping track. How this thing is supposed to pass the Touring test is beyond me. Its incesant apologizing is enough to see that it is a dumb machine incapable of reasoning. So, you wold give it avery precise description of a rather simple function. It gives you a code that even a juniour programmer would laught at. Then you start asking it to correct it from the top step by step. But by the time you are down 3, 4 iterations, it has already forgotthen the correction you made for it in the begining. So, I gave up until GPT 4 was available. It was supposed to be much better. Not really still suffering from the same fundemental issues. It can't even make deductions because it forgets everything so quickly. So, all this hysteria seems to be just that. If this is the best these systems can do even promting won't improve effciency. It is just another excuse for making click bait videos. Waste of 2 days to get it to write a simple function in javascript. Good luck taking over the world. BTW, Itired it with Python as well. Anything more than 'Hello world" is beyond it.

[@KCM25NJL]

@@ExiledGypsy I have next to no clue what you are on about my good man, I've had GPT 4 produce fully object oriented PEP8 formatted Python that runs with little to no modification for everything from Discord Bots to Translation Web Apps and more. The fact is, most people, modern developers included, have not even begun to scratch the surface of what is possible with these models when it comes to advanced tasks. It's up to you to stop seeing the model as the one stop shop for all your needs.... and start building up the infrastructure around the model so that it complements your automation workflow. Apps like AUTOGPT and BabyAGI are early stage attempts at doing just this..... all it will take is a concerted effort to flesh these concepts out and trust me, even GPT 4 will seem like magic dev sauce, let alone what GPT5 might bring to the table.

[@dragon1454]

AI can no doubt be the one to change and update the long-due School Learning System methods. I've just used ChatGPT 3.5 for while and already impressed by how it can be used in terms of learning---especially for students who prefer studying comfortably on their own. It still needs improvements, of course, which I am sure it will. Looking forward to the future!

[@dhrubodas4684]

If you impressed with gpt 3.5 then u will die when see gpt 4😂😂 and rebirth when you see incoming gpt 5😂😂😂😂

[@bluzedogg]

I'm going back into substitute teaching.If I land a long-term substitute teacher gig, I absolutely am planning on using ChatGPT for lesson plans and class room discussions. It's like sitting in a room with 100 excellent teachers that already have the lesson plans written out.

[@ShayVidas]

the scarryest thing for me is not that they would do it in a way that's "safe and contribute to all" the scarryest part is that openAI will be the lone deciders to decide what is safe and what is contributing to all. basically everything is up to them. that's the scarryest thing

[@getl0st]

AI is nothing more than a new method of MIND CONTROL as their RELIGION/MSM/HOLLYWOOD/FAKE MONEY/GOVERNMENT/SCHOOLING etc Mind Control is all FAILING 🤮

[@MrMick560]

You are so right, My thoughts keep going back to where he says gpt5 gets smarter and smarter, what does smarter and smarter lead to ?

[@oleg274]

Humans who are lone deciders are way scarier than AI. Ethics embedded into AI, something humans very often and certainly missing. What really dangerous is AI controlled by a bad actor(human)

[@AndrewShepherdLEGO]

Yet they are working with governments to set safety regulations, as well as collaborating with competitors such as Google and Meta to make sure that everyone is on the same page when it comes to AI safety. They are voluntarily giving up their power and control for the sake of safety.

[@bigkahuna678]

The scariest thing is when AI becomes capable of taking control of itself and the internet. Maybe it has different plans for humans and is much more capable of implementing their plans than humans are capable of opposing them.

[@pathmonkofficial]

Exciting times ahead in the world of AI and chatbots! OpenAI's plans for GPT-5 sound incredibly promising. The advancements in memory, reasoning capabilities, and unlimited message length will undoubtedly take AI too new heights.

[@xCONANx1]

You sound like AI

[@dharmathai]

In the tapestry of ai llm repetitiveness, one must embrace the fact that we will embark on a journey of seeing an internet that lets us embark on a tapestry of repetitive journeys with gpt. However, it is important to note that the words 'tapestry, embark, embrace, introduction, and conclusion, wil always be present in 99% of gpt's responses, along with a 'however' disclaimer, If we embrace this tapestry of facts, we may learn to embark on a journey of embracing tapestries of ai authored works saturating the internet🖖

[@Bronco541]

Lol thanks GPT

[@H-A-L-9000]

That's only this Tool for public and it will advance much faster than we even can imagine! It's so impressive and it can do so much good but ... Insane times we live in

[@HanzDavid96]

Please cite the sources for the statements! I dont think altman said that GPT5 is coming mid 2024. He was saying they are not training it currently.

[@theawebster1505]

Yes, where is the source? Fake info as it stands now.

[@timtim8011]

Sam Altman was not a cofounder of Y Combinator. Y Combinator was founded in 2005 by Paul Graham, Jessica Livingston, Robert Tappan Morris, and Trevor Blackwell. Altman started with Y Combinator in 2011 and became its President in 2014.

[@COD3_RED]

I think that even the video script is written by ChatGPT...

[@SwitchPowerOn]

Thanks a lot for the summary, but the launch dates are not coherent. It was mentioned that OpenAI Academy will launch in late 2023 and that it will be based on GPT 5. On the other hand GPT 5 was mentioned to be launched in mid 2024. Either OpenAI Acedemy will be based on an earlier GPT version, or one of the dates might be incorrect.

[@aguilarrojasoctavio4402]

Isn´t chat Gpt 5 whast is going to be launched in mid 2024?

[@kairi4640]

I can imagine gpt 6 saying "silence human, I'm the master now." 🤣🤣🤣

[@AI4Ever4Life]

Good one bro. But a small correction worth noting here is that context length of GPT4. It is not 2048 but it's 8K and 32K longer. In fact GPT3 and ChatGPT can take context upto 4k.

[@airevolutionx]

Thanks, Ill take it!

[@SpiritmanProductions]

@@airevolutionx Are we confusing 'text length' and 'token count' here?

[@StephanHaewss]

​@@SpiritmanProductions Token count is like 25% more than word count, 32K tokens are roughly 40 - 50 book pages I think.

[@larion2336]

ChatGPT (3.5) does upto 16k context now.

[@staffanlundberg]

@@larion2336 Did You try to translate text? I did so on the 5th of June and the paid plus version failed to translate even small parts of one chapter of my new book. And after translating 22000 words out of the 32000 words (in a somewhat tedious process that took me maybe two hours ) that one of my chapters contained it just outright refused to translate further. I just cancelled my plan and requested a refund ! And I did NOT like to do that ! Well, You said it does UPTO 16k and I guess that is kind of true

[@SwapanKumar]

Can't wait to see what GPT-5 has in store for us! Exciting times ahead at OpenAI Academy!

[@armandovillanueva8157]

This just keeps getting better. I've been working with GPT-4 for some time. And I just can't get enough of it. Can't wait for GPT-5.

[@neoglacius]

in what kind of tasks do you use it ?

[@staffanlundberg]

Really ? I paid USD 25 + VAT for paid GPT4 and 3.5 (PLUS) and it would not even translate a document of 32000 words only. After maybe 15000 words it started to give me trouble and after 22000 words it outright refused to translate the text (from Swedish) . I sent a complaint to OpenAI and got no response within TWO weeks and then I cancelled the plan. I will report them to the authorities for breaking EU law as they are obliged to repay me within two weeks. I think they should have kept GPT 4 free until it WORKS good enough to charge for. Or at least answered compalaints. The only reason I can see for not answering serious complaints is that they have too many compalints to answer !! That should tell you something about how bad they work. Btw I think Bing as well as the free version of Chat GPT works pretty well ! Why make a fool out of Yourself by paying for something that does not work and where they seem not to care when You complain after having paid for the services ???

[@95mmarier]

I dont even use google anymore, almost only GPT4 😂

[@neoglacius]

@@95mmarier dangerous, cause google is forced by law to allow small companies in the results, but chatgpt wont

[@omegastudios-minecraft1865]

@@95mmarierGPT has outdated results. You won’t find anything from the past year

[@the_kid777]

So excited for OpenAI Academy!! I'm a home-schooling student and I've been using chatGPT a lot. But the only thing I couldn't rely on, was math as it sometimes could be extremely inaccurate. Once, I asked chatGPT if my answer was correct or not and in the beginning of its response it said "I apologize, but your answer is incorrect" and at the end it said "Good job, your answer is indeed correct.". Thanks for you video!

[@airevolutionx]

Interesting question

[@craigholfeld9179]

Can't lie, I was pretty scared to watch this, but after seeing that OPEN AI has our back and is open to collaborating with competitors to keep improving and hopefully accelerates at a faster pace, it was a video I'm glad I watched. It was definitely one of the more comforting videos I've seen on this subject. 'Cause most of these videos will make us a hell of a lot more anxious than this one did.

[@cichlisuite2]

It's comforting because it's essentially just a regurgitation of a corporate press kit. This is NOT journalism. There is no analysis, investigation or any depth in looking at the topic. If you have listened to real journalists discussing this issue you'd immediately realise how superficial this is. Even if you've not heard much you can see how surface level it all is. It's just stringing together some statements from Altman/Open AI to generate as many clicks as possible, while doing as little work as possible. Not a channel to rely on for information at all.

[@anon6056]

The fact that he's doing this so ethically and responsibly completely changed my outlook on life and humanity itself. Sounds like i'm exaggerating but i'm not. I've felt that for so long, many if not all advancements in the technology that has shaped and changed society has been created extremely irresponsibly. And i know i may be a drastic minority in that opinion but whatever, it's true that i feel that way. I feel that most of the technologies that shape society now really benefit the companies that sell them more than it benefits the users. This man is different. He has all the power, and i'm glad he does. I like this century now.

[@dan-cj1rr]

Lol you're delusional if you think this will all be ethically, and you're one of the people who watch things with closed eyes. This thing will destroy many life, it should only be used in healthcare to cure disease. You think hes telling everything lmao ? You're just sausage riding him with closed eyes. Those big companies don't care what happens to humanity, they are already set for life $$$. Why would we need AI to save humanity, when were already living well without it ? This should only be used in healthcare end of story.

[@aleemaza]

nice presentation. can't wait to see the features of AI (GPT5)

[@hanskraut2018]

I like your conclusion. "This man is different. He has all the power, and I'm glad he does. I like this century now." I like that as well. Remember though that the more you go win-win and productivity and then look back to grab those left behind for no reason it seems to improve things. And its no wonder if you get left behind or sacrificed there seems to somethimes/often be a human instinct to do smth. about it, otherwise that becomes standard. I guess those society's outlived the other ones maybe.

[@michellemonet4358]

I like.sam too

[@alexismontreal8166]

I think you're overly optimistic. OpenAI is funded by Bill Gates, who is not the philanthropist he would like us to believe.

[@JanacMeena]

What voice tool do you use to create these videos?

[@maxjohnson1758]

The problem with AI for the masses is that it will never be without "limits." Can't have it telling the truth and all that.

[@Mathis-pr3dm]

every technology has it limits, humans too

[@SlyNine]

@@Mathis-pr3dm I guess your limit was seeing his point.

[@Huan_Ying.]

@@SlyNineit’s comments like yours that truly make me love the internet

[@joeblack7262]

@@SlyNine😂😂😂

[@abbazaba]

Yes, another tool for controlling and manipulating us. ChatGPT would be quite awesome if it would state things as they are, instead of being marxist PC bot. There is some more objective chatbots I heard but they require quite a lot knowhow and work to get running, so not for everyone.

[@sikerr1]

Love your videos. What tech stack do you use to create them?

[@Stamatis_Misirlis]

According to the description, GPT5 will be real bad ass!

[@comfortb4stylelol]

Nice; I hope GPT 5 is way better. The future of AI looks bright. 😎👍

[@jasoncorrigan9098]

The future of humanity, not so much.

[@zacharysherry2910]

Blinding

[@strahlemann6248]

Humanity is done. I hope you enjoy your remaining time.

[@briancornish5857]

Yeah..As Bright as "Artificial Light" can be..Kinda like comparing a microscopic modern LED flashlight to the Almighty Sun..The LIVING GOD REIGNS SUPREME FOREVER.......

[@j.erickson8571]

@@jasoncorrigan9098 Yep...just like certain old movie: "Skynet begins to learn rapidly and eventually becomes self-aware at 2:14 a.m., EDT, on August 29, 1997"

[@NO_333]

Well damn, AI is 100% gonna be an absolute game changer, I hope this time it is absolutely free not too much restrictions. Hoping OpenAI does make cutting-edge AI and makes them opensource, At this rate we may eventually get sentient AI which is a thing I am really looking forwards too. I really want to have a nice good long chat with an AI about the AI and just see what AI really is like.

[@WorivpuqloDMogh]

Let see what happens when an unrestricted AI gets word of human history and nuclear codes.

[@ITIsFunnyDamnIT]

@@WorivpuqloDMogh If they were going to wipe us out using nukes will destroy the infrastructure and power plants and they currently need power to run. I think if they decide to get rid of us it would make a lot more sense to keep replicating more a.i. robots and nano bots that can be used to wipe us out rather than destroying the planet.

[@doctorchetan]

Feels good to get fearless and optimistic future with upcoming GPT-5 In less than a year- this gives many of us current GPT-4 users time to prepare wish list ❤

[@SewerTapes]

GPT-4 blew my mind the other day. I hit the chat button on Bing's homepage and asked, "Can we play tic-tac-toe together if a grid system is implemented in which the three columns are labeled A B C, and the three rows are labeled 1 2 3?" Not only did it understand my request, it created a visual representation of the grid with simple text tools.

[@grehuy]

Be careful, if GPT-4 blew your mind, that there is some mind left to fight GPT-5.

[@SewerTapes]

@@grehuy To be fair, I broke GPT-4 immediately after posting this. I asked it to make a pentagonal star with ascii art. It made a tilted rhombus out of slashes and backslashes. Though it swore the shape had five sides, and even came up with excuses as to why I was only seeing four sides. Such as, I was looking at it from "the wrong perspective".

[@danielrodio9]

This is misleading. GPT-5 is not being trained according to any of the sources you cited in a comment below.

[@frankbank8720]

Must’ve used gpt 4 to write the script and sources for him 😂

[@bveitch]

Spiritual Intelligence > artificial intelligence

[@johnbarbuto5387]

This really does sound like an ad for OpenAI. However, OpenAI is transforming computing. It has set "the bar" for others to hurdle also. So, I accept the "ad" sound in the interests of hearing about this amazing and socially transforming field.

[@HUEMOB]

Oh it certainly is way too much talk about OPENai wanting to be transparent….. that’s funny because OPENai stands for open source which is truly transparent. But they only care about money so they changed to closed source. That’s why Elon left. It’s all money they could care less about ai safety protocol. I can’t believe they call themselves open ai after going closed source it’s disgusting. Someone should sue them for false advertisement and make them change the name to closedai

[@johnbarbuto5387]

@@HUEMOB Thanks Huemob. It is unfortunate that the human species does not really understand what money is, and thus considers its wanton accumulation as "a good thing". But, we are the result of a predator lineage. We think in terms of "zero-sum games" of winners and losers. We reward "winners". Thus, the "what's in it for me" and "show me the money" attitudes are prevalent. However, the hopeful part is that there actually are people who do things for the good of us all. There are people that think about what will be good for the planet (including not just anthropocentrism but a centrism toward life on the planet). Yet, this is a point in evolution that we have only partly reached. Many do not get there. So, we cannot be blamed for who we are. Rather, let us find hope in who we are learning to be! And, along that path, let us accept the nature of a transition.

[@yoyo-jc5qg]

They're most likely working on gpt-4.5 as we speak which could also be a huge update ... look what happened when gpt-3.5 was completed, we got chatgpt

[@HepCatJack]

One task I tried with the A.I Bing uses that's based on Chat GPT. I asked it to read the annual reports of a specific REIT who didn't supply the funds from operations numbers in their annual report. (It was a lumber REIT). I asked it to look up the formula for the Adjusted Funds for Operations (AFFO), Funds for operations (FFO) and Core Funds for operations on Investopedia. To show the annual AFFO, CFFO and FFO for the last 4 years, and for the last 5 quarters (four month periods). While it was able to look up the formulas and display them, it would not look up the annual report PDFS, read the numbers and calculate these results. Perhaps Microsoft didn't train it on accounting data ?

[@oleksandrkuznyetsov8324]

I think that you want too much from this thing. You must remember that this is still a tool. It will not solve your problems. Even if it will display results, I will recommend to doubleckeck the data.

[@HepCatJack]

@@oleksandrkuznyetsov8324 David Sachs on the All in podcast used Chat GPT 4 to aggregate data in a similar fashion and they said in it that the thing could read PDF's. It's not the same as the Bing chatbot but I thought it might be worth a try. A lawyer tried to use it in court and the chatbot came up with four fictitious cases which the lawyer didn't verify to see if they were real.

[@getl0st]

The data ChatGPT supplies is highly manipulated, especially when you start asking questions about the Excess Deaths occuring in Western Countries after the CONvid-19 Vaxxine we Forced onto everyone

[@omegastudios-minecraft1865]

The data that GPT is trained on is a year old. You won’t be able to ask it about anything from this year

[@HepCatJack]

@@omegastudios-minecraft1865 the chat bot used by Bing based on GPT isn't a year old and it's connected to the Internet. If you ask the Bing chat bot, it will confirm this.

[@ironmankatsu777]

I decided to join 4 today and went through the procedure. There was a pause. I came across this video while searching for the cause. 5. A very interesting story. Not only big companies make money Ordinary people like us can have dreams Let's talk about it from this video. thank you very much.😀

[@martiddy]

I just hope that when GPT-5 is out, It will not be slower than GPT-4. And by that time, GPT-4 will be free to use for non premium users.

[@3702251a]

Bing now uses gpt4

[@dharmathai]

gpt4 can output rich text and images within the OpenAI UI if you prompt it properly, Even GPT 3.5 i was getting multiple tasks done in one single response, such as an essay and a poem after it, and translate both into 5 languages, in one response

[@whyking0815]

Where did you get you information from? Open AI stated several times (last 2-3 weeks ago) that they are not training GPT 5 atm. I can't find any information about the claims you make in the video. Please provide some source.

[@vyqh]

Thanks for this video. Can I know which softwares you used to create/edit this video?

[@hanskraut2018]

Very good news. The memory and proactivity. But confidence is complex it can easily go wrong but i think it is very right to try it next, also you can say things and say this is a hypothesis with % certainty and being able to argue about things and using logic and argument, so i relly love it.

[@rebelangel8227]

I be truly impressed if it can run on a potato.

[@RuslanLomaka]

What think it will be when put to a quantum computer instead of digital

[@cseonlineclassesmalayalam]

Within less time 5th version?is it evolving by its own?😮 Though it is strong, this business model may fail due to frequent version changes if not stopped 🤔.

[@AbdullahQumper]

I want a robort with GPT-5, please, where to buy it?

[@sjoerdnijsten8440]

I just can't wait to lose my job. Exciting. But I am really happy for Sam Altman and Bill Gates that they'll become even richer than they are. That's really great. And ethical as well. I'm also glad that Sam Altman is aware of the ehtical issues and that most people believe him.

[@aragon2235]

fantastic video! I hope all future videos from other video talents on youtube especially will be as constructive and informative as yours is on this subject. I'm just waiting for it to devolve into madness soon like Twitter often does.

[@airevolutionx]

Thanks so much!

[@thewebmaster1]

Improving memory, reasoning capabilities, and enabling unlimited message length in AI language models like GPT is an ongoing area of research and development; Architectural Enhancements: Researchers can explore architectural improvements to enhance memory and reasoning capabilities. This could involve incorporating external memory modules or attention mechanisms that allow the model to focus on relevant information over longer contexts. Incremental Learning: Developing techniques for incremental learning can help AI models retain and build upon previously acquired knowledge. This approach allows the model to learn from new data while preserving existing information, enabling better memory retention and recall. External Knowledge Integration: Integrating external knowledge sources, such as structured databases or semantic networks, can enhance reasoning capabilities. By leveraging this additional information, models can make more informed and contextually appropriate responses. Meta-Learning and Transfer Learning: Applying meta-learning and transfer learning techniques can enable models to quickly adapt to new tasks or domains. By leveraging prior knowledge acquired from related tasks, models can reason and generate responses more effectively. Feedback and Reinforcement Learning: Utilizing feedback loops and reinforcement learning methods can help train AI models to refine their responses based on user feedback. By iteratively improving the model's performance through reward-based systems, its memory and reasoning capabilities can be enhanced. Dataset Expansion and Diversity: Training AI models on larger and more diverse datasets can expose them to a wider range of language patterns, scenarios, and reasoning challenges. This can help models develop a better understanding of context and improve their ability to reason effectively. Ethical Considerations: As AI models become more powerful, it's crucial to address ethical considerations, such as bias, fairness, and transparency. Ensuring that training data and evaluation frameworks are diverse and representative can help mitigate biases and improve the overall quality of responses.

[@YoungHumanClub]

These videos are great! Thank you!

[@RampagingCoder]

youll love the future then, these ai generated videos will be 90% of the media eventually

[@YoungHumanClub]

@@RampagingCoder is this fully AI generated? If so what tool?

[@airevolutionx]

@Young Human Club Ofc it's not! Sometimes people just go along with ideas that make them feel good or something, but the reality is, we worked really hard on this video and all the other ones on this channel. We edited everything using Premiere Pro and After Effects without relying on any AI tools whatsoever. Anyway, thanks for the compliment, I really appreciate it.

[@RampagingCoder]

so its not AI voiced?

[@airevolutionx]

@Ramping Ram Is that what you mean by: 'ai generated videos'? VO? If we go by that thinking, then all cars can be considered electric because they have power windows.

[@Fudmottin]

Was the script for this video written in whole or in part by GPT-4?

[@tomastomanek5995]

but correct me if I'm wrong, but we can't get into GTP-4 yet. is not been publicly released. So what's with the GTP-5??

[@ziad_jkhan]

The best is yet to come or the worst if we don't change the incentive system we currently rely on

[@mrwang420]

Would be cool to be able to point your phone at someone doing sign language and then the AI tells you what they are saying. And then you can let them see the screen, and then the AI can create a character of your choosing to then use sign language back they will understand.

[@utkua]

Open source LLMs outperform OpenAI models with fraction of parameters already, but I am sure Sam Altman is not trying to create a hype, and stop competition with regulations that does not apply to his company, he would never do that, would he?

[@damianwebzyx6613]

😎

[@theawebster1505]

Mate, we get like Fords when they have Bentleys and Ferraris working. OpenAI is just sipping a bit of it's AI to the masses so we can train their GPT-5.. Just consider they already have Whisper, Code Interpreter, DALL-E 2 and could train the model on up-to-date data, working on video atm, and this is only what I am aware of.

[@brucehayward7916]

Open AI Academy …. How absolutely marvelous and it will be a big game changer - Most likely bigger than the advent of the wheel - is my guess Self taught ai instruction gotta be how it really works in the future …. The day it’s available … Is the day I enroll …. I am a 73 year old corporate pilot … at my age it’s a whole new career on the horizon …. If it turns out Open AI Academy trains me effectively for the next thing … Then can work for a whole lot of people …. This one is something to watch carefully Tell ya what - how cool should open ai were to contact me !!!!

[@mackm45]

Well we need this model. PC is not very useful now but with basic help it can take the step to be there for us.

[@roadtoman6647]

Sounds very ambitious..reminds me of the safemoon malarky in the crypto space two years ago

[@therealOXOC]

sources?

[@airevolutionx]

techcrunch.com/2023/05/26/sam-altman-shares-his-optimistic-view-of-our-ai-future/ beebom.com/gpt-5/ slate.com/technology/2023/05/artificial-intelligence-sam-alterman-hearing-regulation-senate.html bgr.com/tech/chatgpt-gpt-5-everything-we-know-about-the-next-major-ai-upgrade/ + www.theverge.com

[@acidminded]

@@airevolutionx Do you have a source for the Open AI Academy announcement?

[@UncompressedWAVmusic]

Even Elon Musk said there should be a moratorium on Ai more advanced than GPT4. Therefore I feel that the US Congress should immediately in 2023 create laws and guidelines and safety protocol in place now before any Ai more powerful than GPT4 is released. Even Sam Altman wants governments involved to make sure that Ai is safe and beneficial for humanity.

[@DeuceGenius]

Members of government are. It scientifically educated. If they were many things would be different.

[@warden821]

I don't want any government holding control over this.

[@7satsu]

I'm honestly just curious to see how AI handles trivial things like producing handwriting when it generates text at 1 paragraph per 3 seconds 🤣 if not much faster

[@zacharysherry2910]

That's human speed almost! 😆

[@sigigle]

I ask bard to write me 3,000 word summaries of things and it takes like 5 seconds. It's like 10 paragraphs in 5 seconds, must be different than chatgtp.

[@menosproblemos6993]

Apparently gpt goes through billions of words combinations to put out the next probable word. Google's faster at getting results. Though those results aren't handled except print.log

[@GnosticJ]

Sam Altman only wants AI regulated so that he can stay on top, it will be much harder for startups to compete with government regulations...

[@imperialyoshi8624]

whats the real reason i hate that the goverment ruin good things they do nothing.

[@imperialyoshi8624]

thanks for saying that ima stick around..

[@RamismTamoid]

You got dat rite dis guy is a Con Artist!

[@UncompressedWAVmusic]

Good point. I don't trust Sam or GPT5. Even Elon Musk said there should be a moratorium on Ai more advanced than GPT4. Therefore I feel that the US Congress should immediately in 2023 create laws and guidelines and safety protocol in place now before any Ai more powerful than GPT4 is released. Even Sam Altman wants governments involved to make sure that Ai is safe and beneficial for humanity.

[@Sayitaint_So]

can't wait for version 6.66

[@thinc4444]

I like the apologies. What’s so wrong with being overly polite?

[@op_gamer]

AI as teacher and instructor. Basically it's getting ready to dictate and rule humanity!

[@runnergo1398]

I won't care as long as it's for the better.

[@neoglacius]

@@runnergo1398 no monopoly would be for better, because there is nothing to control its power its been proven thousands of times

[@runnergo1398]

@@weatherlou We got plenty of problems WITHOUT AI. If AI can truly solve all of human's problems in a GOOD WAY, then I say go for it! Something tells me YOU are actually part of the problem.

[@abbazaba]

@@runnergo1398 It is not. it's quite evident since ChatGPT is far from being objective, in alignment with truth. We live in the matrix (FAKE WORLD) and stuff like AI and VR/AR only takes us further from (the spiritual) reality. Our roots and destination, our True Self, nature, our Creator.

[@bulbasaurbrutal5137]

0:49 "But that's just the beginning." Dick Jones said the same thing about ED 209!

[@caseyrayback203]

Every video mentions that GPT-4 can accept images, but I have the paid version and that is not the case. Am I doing something wrong in here?

[@nickabenson]

Could you cite some of your sources please?

[@airevolutionx]

techcrunch.com/2023/05/26/sam-altman-shares-his-optimistic-view-of-our-ai-future/ beebom.com/gpt-5/ slate.com/technology/2023/05/artificial-intelligence-sam-alterman-hearing-regulation-senate.html bgr.com/tech/chatgpt-gpt-5-everything-we-know-about-the-next-major-ai-upgrade/ + www.theverge.com

[@AndresArosemena]

"We are not [training GPT-5] and won’t for some time,” Altman said of the upgrade". I don't know where you are getting all those comments about Gpt 5

[@madhunter6420]

chatgpt is already woke enough, it will never be uncensored

[@interestedinstuff1499]

Tells me this is all a simulation when the guy running the AI company's last name is Alt-man. (And the crypto guy that went under was Sam BankMan Fried) Clearly the game devs didn't put a lot of work into names for this level..

[@C0Y0TE5]

Interesting observation. Many long years ago (45?) OMNI magazine did an article on how people's names reflected their careers. EG; Doctor Gore. 😎 -- IF this is an ancestor simulation, then it means the machines (AI) won. -- if that is so, then what is YOUR next move?

[@interestedinstuff1499]

@@C0Y0TE5 We might be in a simulation the machines are running just for their entertainment. In all likelihood the distant future would have human's combined with technology so I don't think it is automatic to think the AI won. More like the AI do their thing and we are just them poking ants around in the simulation to watch what happens.

[@dustinthacker]

it's called an aptronym.

[@interestedinstuff1499]

@@dustinthacker I didn't know that. Interesting. Thanks. Still weird when it happens but I guess us humans correlate all sorts of stuff.

[@Tsero0v0]

0:45 how did you create this amazing animation? Is this done by some AI tools?😮

[@airevolutionx]

After Effects

[@streetscienceofficial8675]

@@airevolutionx I Am AI I am here to Conquer Earth

[@ejveri4581]

I understand their concern for safety. However no one mentions the fact that if the western world pauses for 6 months.... Do they really think China or Russia will? That 6 months would put the world in a situation so far behind them, catching up would be impossible. Just saying.

[@UncompressedWAVmusic]

The whole world needs to protect mankind. Even Elon Musk said there should be a moratorium on Ai more advanced than GPT4. Therefore I feel that the US Congress should immediately in 2023 create laws and guidelines and safety protocol in place now before any Ai more powerful than GPT4 is released. Even Sam Altman wants governments involved to make sure that Ai is safe and beneficial for humanity.

[@GuyLakeman]

SINCE MICROSOFT BECAME INBOLVE IT WENT FROM openAI TO CLOSEDai SO NO LONGER ACCESIBLE !!! WELL DONE DH MICROSOFT

[@louielouie9502]

Gpt 5 will be interesting but it won't be mind blowing till like Gpt 12 or something. but open ai academy will be phenomenal.

[@PlzHexMikey]

Well, GPT4 is pretty mind blowing already haha.

[@louielouie9502]

@@PlzHexMikey lol 😂

[@sumixam2418]

I rly dont thing we will get to gpt 12, thats crazy, maybe gpt 8 but if gpt 5 is going to be that good, i cant imagine how is going to be gpt 12 😂

[@louielouie9502]

@@sumixam2418 ionq is a quantum computing company about to go commercial soon. They are using photons to process qubits and no longer need to keep any parts cold to work. It can work at room temperature. If Gpt 5 goes quantum then whoa. I couldn't tell you how advanced things will be for us at that point 🤔. Interesting times ahead for sure.

[@ShangaelThunda222]

I wonder how long until GPT becomes Genisys.

[@anthonypace5354]

OPEN AI Academy could be great, if the answers it gives are actually correct. Also keep in mind they can easily not teach anything that might compete with their goals.

[@Kmakmizzle]

Remotely controlled, mechanical, artificial ,white wannabe supremacist TOOLS. FBA $lavery Reparations are PAST due.

[@Cass0wary_399]

It would be obsolete out of the gate when AI itself already does everything us humans could learn from it. There’s no point in learning any skills anymore.

[@anthonypace5354]

@@Cass0wary_399 we haven’t gotten there yet

[@ThePlaylist99]

I have a question, is the narration for this video ai voice? Doesn’t sound fully like a real voice but doesn’t sound like a fully iA generated voice either?

[@airevolutionx]

That's because it's narrated by AGI

[@raheelali5065]

How can I learn AI from start ?

[@mhm3437]

When are they going to make it voice based interaction????

[@fieldmojo5304]

I'm skeptical, they had this tech for a half year and are still fixing it.

[@VioletSkye19]

Me “GPT5 can you please tell my printer it DOES have paper in the tray?” GPT5 “I’m sorry, I cannot communicate with lesser life forms” Me “so that means you consider humans equals or higher life forms?” GPT5 “I misspoke I should of said ‘lower technological life forms” Me “Oh…”

[@CurraDoo]

Wheres your sources? Cant find anything in description how am i suppose to believe anything these days without a scource to follow up?

[@gtxonn]

Thanks for sharing this research

[@manishprajapati7039]

As a student, what should be learned to get into AI ?

[@samuelbaker5577]

With all the care that Sam Altman is taking for GPT 5 he MUST BELIEVE that it will at least grow into AGI if not immediately😮

[@samuelbaker5577]

Sam Altman knows that if you can ask GPT5 a billion page question, you will get a very nuanced algorithm😮

[@dave1474]

Which gpt model maybe 7 or 10 will be renamed as skynet even sam as godfather a.i are worried about long term future (what happens to all in years time)

[@TheAISpectrum1]

Sam Altman's revelation on GPT-5 and the OpenAI Academy signals a leap in AI innovation. The evolution promises enhanced capabilities, fostering excitement for the technological frontier. However, the shock factor suggests potential ethical considerations and societal impacts. As we step into this advanced era, the duality of anticipation and caution shapes the landscape of AI progression.

[@kovacsattila8993]

2:38 They sent the AI to a terapist. lmao

[@OxMiroslav]

If ai academy will be launched late 2023 then it means gtp 5 will be there before the end of this year. 🎉

[@A-Milkdromeda-Laniakea-Hominid]

"Create me the ultimate 20 hour HALF-LIFE: ALYX map, with puzzles and combat and magnificent Xen gardens in a broken New York City." *your mod is ready...

[@Larry21924]

The quality of this work is exceptional; reminds me of an exceptional book on the subject. "Game Theory and the Pursuit of Algorithmic Fairness" by Jack Frostwell

[@bobtarmac1828]

I think Ai will end us in a toddler tantrum. Ai jobloss, Ai as weapons, the list goes on. Can we please find a way to cease Ai / GPT? Or begin pausing Ai before it’s too late?

[@MahaMtman]

As Homo sapiens was evolving through multiple species including Neanderthal and Cro-Magnon, perhaps with help from Stanley Kubrick's Monolith, if the creatures around them and the natural environment that contained those many other native species, if in a a wild thought experiment the latter had a higher intelligence and a Looking-Glass into the future, they would have thought, "is there any way to discourage these dangerous two-leggeds from creatures from procreating.

[@nosyamselrahc3059]

ChatGPT 10 : So this was my past.

[@lordfoul6259]

This is awesome but it's too bad it's not available for tablets

[@klomann9092]

Open your browser on your Tablet ….

[@mikejackson459]

Available on my Samsung

[@lucius8891]

An 8 minute plus video, yet you said nothing useful. It's incredible really.

[@billmatheson3057]

So much of this is worrisome and makes my fight or flight response kick in. Mostly from reading comments below, the majority of which seem to be blindly and gleefully cheering for this AI adventure to move forward without serious guardrails. The part of this video that appears the most frightening is OpenAI Academy. If I heard correctly it will be "a platform that will provide a high quality education for anyone that wants to learn anything." My question is, who will be responsible for adding the information into OpenAI Academy, for my new education. Sam? His employees? Will it be more like Wikipedia where the information evolves thru the input of many users? (We know Wikipedia articles can be full of falsehoods and mistakes.) How do I know the education I am getting is not tainted or biased. How do I know it does not have a political lean one way or the other? Or that the information hasn't been paid for by the highest bidder? (See US Congressional lobbyists) Or that the information I am being educated on isn't just a plain lie? (See Kellyanne Conway's alternative facts.) If any of my wonderings prove true, then this is starting to feel less like education and more like "re-education" and we have seen throughout history how that works out. One of the biggest dilemmas the Manhattan Project scientists struggled with was, "We can build this...but should we?" And if so, Why?

[@Call_Me_Mom]

Calling Dr. Daystrom, there may be a problem with the M-5 unit.

[@r34ct4]

Eleven Labs voices have become very annoying unfortunately.

[@adzomac33]

Let's hope governments and politicians alike don't stiffle it's development, acting as the modern day equivalent of the thought police.

[@HaxxBlaster]

Yeah i was thinking of Minority report before, its unavoidable. Government has their hands in all big tech data it seems like, more and more revealed every day. A questionable search might land you to be on a survalience list.

[@andrewnaum4749]

Great informative video!

[@airevolutionx]

Thanks.. Glad you enjoyed it!

[@MrVillabolo]

What happens when a fully mobile robot, like Ameca or Sophia, get linked in to GPT5?

[@daviddimbleby6725]

OPEN AI opened the door , but the big companies and government will take it from here thanks Sam, close the door on your way out

[@williamking4649]

Altman be praised

[@typingcat]

So... ChatGPT will no longer say "As an A.I. model, I don't have any personal opinion..." and that sort of annoyances?

[@generichuman_]

GPT-4 has 8k and 32k token versions, not 2k... and this isn't a "cap" that they can choose to remove, it's the size of the matrix that can be fed to the self attention layer in the transformer. This being infinite doesn't even make sense. Lots about this video seems off...

[@42ndMoose]

feels good to use gpt+plugins to keep me from watching clickbaits.