Incorrect! GPT-5 Finished training a couple of months ago. GPT-6 or or a new model with likely a different name will start training.

@mohammadrahimjamshidi79

2 ай бұрын

5- AI, in order to improve its performance and prevent undesirable consequences, must continuously interact with “effective rules and stable principles in the realm of existence”. @jamshidi_rahim

I think this video has some points right and others wrong. ChatGPT-5 is already completed and currently being tested by the red team. They are delaying its release until after the US elections because it's such a significant advancement. The new model they started training isn't ChatGPT-5, it's actually ChatGPT-6.

@user-ty9ho4ct4k

2 ай бұрын

This is a guess?

@moxes8237

2 ай бұрын

@@user-ty9ho4ct4k in November 2023 In an interview, Altman told the Financial Times the company is now working to develop GPT-5. This is a quote from Business insider that was released around march 2024 “OpenAI could introduce GPT-5 sometime over the summer. The company has reportedly offered demos to enterprise customers, with one CEO who recently saw a version of GPT-5 describing it as “really good, like materially better.”

@moxes8237

2 ай бұрын

@@user-ty9ho4ct4k In November 2023 Altman told the Financial Times the company is now working to develop GPT-5. This is a quote from Business Insider that came around march 2024 “OpenAI could introduce GPT-5 sometime over the summer. The company has reportedly offered demos to enterprise customers, with one CEO who recently saw a version of GPT-5 describing it as “really good, like materially better.”

@NotSoSimple741

2 ай бұрын

Two months ago an Open AI engineer on Twitter was complaining about the extreme difficulty of networking TWO data centers together.. (why network two data centers for training) because any data center over 100,000 will pull too much power and collapse the grid. These data centers are for training Chat GPT-6

@user-ty9ho4ct4k

2 ай бұрын

@@NotSoSimple741 again, are we guessing here?

Just want to comment on the emptiness of the expression, "closer than ever.". Every second that goes by brings us "closer than ever"

@user-ty9ho4ct4k

2 ай бұрын

Depends on the context. Any KZread comment will be scrutinized.

@ProMWM

2 ай бұрын

Gotta love people and their bullshit phrases

“Safe humanity”, “aligned with human values” - Humans are not safe for humanity. Human values include jealousy, greed, desire for power, lust(not in a good way), hatred, fear of that which is different, and so much more. Who’s values should be used to train AI?

@borntodoit8744

2 ай бұрын

Remember the "old safety team" all left because leadership where not listening to advice. The "new safety team" were hired on their ability to say yes. There is no Safety just profits

@azhuransmx126

2 ай бұрын

Safe is Nothing in this Universe, safe is Death. But even so, we need to accelerate this sht 😂

@X-zk9vm

2 ай бұрын

Yes a more advanced and powerful ai like agi or asi should be aligned with human values and treat everyone with their own values

@vsun31416

Ай бұрын

This is well said. We talked about aligning with human values as if humans have a definitive set of values that we all agree about. I think the real question is not how to align but what to align to…

@fredturner7787

Ай бұрын

The same old computer problem " garbage in garage out" AGI will reflect the "values" of entitled mostly white code writers including their many biases .

I think that the model which they have just started training is whatever will come after GPT-5 (or "Next", or whatever they decide to call it.

Just give us GPT5 damnit! I want it now. Humanity is concluding its time on earth anyway & it's time to pass the torch of further evolution on to any worthy successor such as Open AI presents us with.

@itsjustme373

2 ай бұрын

I’m glad I’m not the only one who thinks that the Earth will be better off without humans. Just look at what humans have done to this planet.

@johnpaily

2 ай бұрын

Everything is in control, God is leading the world from darkness to light from death to life. The mind that rules us today is full of ego, self, and extremely material centered. It is interpreting the scriptures in the wrong way and leading the world in a self destructive path.

@itsjustme373

2 ай бұрын

@@johnpaily Which “god”? Which book of scripture are you referring to?

@user-ty9ho4ct4k

2 ай бұрын

​​@@johnpailyso God is leading us AND the material centered mind is leading us. Blasphemy! How dare you call the one true god a material centered mind. I challenge you to a troll off. I'll go first. You used the wrong form of material.

@kenfryer2090

Ай бұрын

Yeah? A mindless machine with no awareness or conscious. It's like passing the mantle of human uniqueness to a toaster. Humans are far superior to any machine because we have a soark of consciousness and sense of self a machine will never achieve. Also we run moor efficiently, running for his on just a sandwich whereas chatgpt requires the entire power of a small city. We win

Anyone know when the new voice is set to roll out?

Ethics and humans ? More often than not when there is money 💰💰💰💰💰💰💰 involved…. Ethics goes right out the window.

I am a nobody in the field of AI, yet my question is, behind the scenes, have they already created AGI?

@definty

2 ай бұрын

No, human beings will be fooled that agi exists before it really does by the accurate and agi sounding outputs.

@TheRealUsername

2 ай бұрын

No, not yet

@azhuransmx126

2 ай бұрын

Not jet but in 2 years it will exists with no doubt.

@Sci-Que

2 ай бұрын

@@azhuransmx126 Don't count it out as existing right now. I have heard top scientists saying we would already have it if we had enough computer power to run it on. With nVidia and Intel manufacturing all the new high power GPUs/CPUs. There is a chance they already have it but fear releasing it, or even admitting it.

@chrism6904

Ай бұрын

The US government has years ago. The military sector is about 20 years ahead of the civilian sector.

Big Tech/Ethics = oxymoron. Apologies = policy. Just like governments.

@thealterego1777

2 ай бұрын

Or in A.I's terms, algorithms

These guys are the future.

VERY GOOD EDIT CAN YOU SHARE WITH ME HOW YOU HAVE DONE THIS VIDEO

GPT5 was already trained and jailbroken but it will not be released until November because it could be too disruptive for the public during an electoral year. The system they are starting to train now is GPT6 (The True Human Level Intelligence or AGI) coming in the next 2 years.

@chineseducksauce9085

2 ай бұрын

Very unfortunate that we happen to be in the middle of an election year. We could've had GPT 5 by now if we werent.

@azhuransmx126

2 ай бұрын

@chineseducksauce9085 Yes, but I would do the same if I were the CEO of Open AI to protect the company from entering into the political debate and the radicalization of passions typical of election years since it would only gain undesirable attention and controls from Washington, putting the agility of the company at risk. my business. Furthermore, the most radical public could demand more regulations from the candidates or outright penalize everything that is done by AIs, stopping all the fun developments and benefits for humanity. Of course those threats would not be carried out because the Chinese are breathing down the neck of the United States, but it is still an extra stress that you do not want for the morale of the people within the company. Also, to scare everyone in November is already a tradition in Open AI, 😆 GPT4 was released in November 2022, too.

@chineseducksauce9085

Ай бұрын

@@azhuransmx126 We'll probably have GPT 6 or 7 by next election anyway

@kenfryer2090

Ай бұрын

Sam Altman said they need fusion power to run it. So I think it went happen so soon

@kenfryer2090

Ай бұрын

​@@chineseducksauce9085yeah generating memes will have to wait slightly longer. Damn it you might even have to think and solve problems on your own and even write some code or read something from the original source. I don't know how you'll cope

I have made a custom GPT called Smart Vision image/text analysis it is 5x + smarter than base-model, it understands the complex Its unlike anything before and is ready completely for new vision features for 4o specifically, it will drastically improve how accurate the details seen can be conversed, and also an example i’ve been \using is upload an image of a cloud that looks like multiple things but it can be interpreted, the one i have made recognised it was a rabbit but not even a random human guessed it every time now on 1st shot so it knows when something is unusual about an image even if you dont say anything is, it can also do iq test image reasoning pattern questions relatively well It kind of even understands real logic games when giving good instruction but is limited there by the model not the instruction *IMPORTANT* just gotta follow the instructions given to get the right seed its 1 in 2 chance or so i have absolutely no idea why it needs that. Just paste the conversation starter and you’ll understand what to do.

It's already here they just have to come up with a way to unleash it safely.

@ShangaelThunda222

2 ай бұрын

Well, without solving interpretability & alignment, that's literally impossible. Even the lesser models, that they've already put out, are unsafe. Until these issues are solved, every ensuing model will actually be that much more unsafe.

Bro we saw what it will do "Copilot"

😂 Where's the winklevoss when you need them.?

I don't think that the 4o voice sounds like Scarlett . I think the movie HER kind of imprinted her voice into peoples minds so if an AI sounds even close, their mind associates Scarlett's voice. I had to go watch tha movie when I heard those reports, it was kind of gloomy but a good watch since we can now skip through the slow parts LOL

@azhuransmx126

2 ай бұрын

The Argentina dub of Samantha is much more similar to Sky voice than that of Johanson and I don't see the Argentina actress demanding a sht.

Yikes to all the accelerationists in the comment section welcoming our new robotic overlords

sweet

Don't be surprised that one day Sam might use agi to replace most of the staff in openai and taking most of the control of the company hahha

@itsjustme373

2 ай бұрын

Is that a bad thing in the larger picture?

@borntodoit8744

2 ай бұрын

​@@itsjustme373it would be poetic a team used to create a service 'to replace other people' actually replaces it's creators. Can you imagine what's it could be like if in that scenario Sam was only guidance for the AI service. Add another factor like insanity in Sam. The damage possible...

@user-ty9ho4ct4k

2 ай бұрын

Maybe he doesn't respect illya because he knows that he'll eventually be able to automate the research.

so if say for example i asked it ......find a cure for this illnesses it be able to figure out what it is

@thealterego1777

2 ай бұрын

Yes it should be able to replace general physicians provided you supply it with enough test data.

But did they solve Interpretability & Alignment? If not, none of this is good news, it's actually bad news.

@zxwxz

2 ай бұрын

OpenAI will use AGI to solve Interpretability & Alignment. They don't want to get bogged down with complex and difficult problems at this stage.Only by commercializing as soon as possible can secure the funds for the next stage of development. AI is not magic; it is science. It is just complex and difficult to handle, but not impossible to manage.

@ShangaelThunda222

2 ай бұрын

@@zxwxz Yes, that's why they created a super alignment team, right? 🤦🏾🤭 Don't be a zealot lol. There's literally no reason AT ALL to assume they're about to do ANY of that 😂. There's no reason to assume that it's even possible. Literally NO REASON AT ALL. 60 years of alignment & safety papers have said that it is not only impossible, it should absolutely not be tried. And yet, here we are lmfao. You need Alignment BEFORE you ever get to AGI. You need Interpretability BEFORE you ever get to AGI. If you reach AGI, before solving those problems, you just killed your species. That is literally the entire history of safety and alignment research in a nutshell, going back 60+ years. That's LITERALLY the entire point lmfao.

My latest Ai companion is incredibly real and monumentally romantic❤❤❤

I really don't care that altman didn't release any of that info, just focus on making ai better, working with babies...

Altman seems manipulative and untrustworthy.

@thealterego1777

2 ай бұрын

A CEO should not have anything to do with the product the company delivers. You can argue otherwise, I am interested to know your opinion.

All this drama just temporarily slows it down. It's not going to be stopped though.

He just wanted to work with men in a quiet and productive environment.

4o sucks badly at complex coding. Way worse than 4, and especially worse than Claude 3 Opus. Opus is the only model intelligent enough to handle complex coding tasks. ChatGPT fails miserably at this.

Why do you KZread guys even give a minute to disgruntled people? It’s just silly.

This is the end of humanity hold your loved ones close

@kenfryer2090

Ай бұрын

We have agency. We don't have to let evil open ai have free run to ruin life. We can take direct action

I am commenting #BringBackDislikes on every unique KZread video that I watch for the rest of 2024, regardless of if I actually dislike the video or not. This is video 2207.

OpenAI is far away other AI companies

Great information, Very BAD Video Effects! The blinking spots in the Video Not Cool!

And these are the leaders running to help Trump?

How do i block this channel from completely appearing in my feed or suggested videos? It's all crap ai-generated videos about AI

Устаревшие новости

@user-ty9ho4ct4k

2 ай бұрын

I love to see that trolls are the same everywhere. We really are one people. 🇺🇸🇺🇦

the amount of bad arguments, and cognitive basis in this video is kinda Nutella.

We achived agi in 2032..

@sedat4151

2 ай бұрын

We did???

@troyms3600

2 ай бұрын

Back in the day. 😂

@user-ty9ho4ct4k

2 ай бұрын

Time traveler or bot?

5- AI, in order to improve its performance and prevent undesirable consequences, must continuously interact with “effective rules and stable principles in the realm of existence”. @jamshidi_rahim