8 replies to this topic

[Set and Meet Goals]

A futurist in this community made a prediction which I thought was super exciting and should be kept track of here.

I think many people here view him in a trustworthy respectful light so the prediction should be taken seriously.

 

I wont say his name in case he doesn't want to put his name in sand to this prediction obviously he would explain his own prediction way better then me especially as I am not involved in AI.

 

Basically this user believes we will have AGI in 2021 and that the creation of AGI is now possible if the required resources are put into it.

 

It will be a first generation AGI so its not going to be that good by any means when compared to a human.

 

The User believes if it isn't released to the public in 2021 it probably means the creators just don't think the creation of a first gen AGI should be announced immediately or at once.

 

 

Personally I think if the first generation AGI happens it should be taken as a 95% + confirmation that human level AGI is possible.

We can also assume human level AGI is not to far as now you just have to make AGI better.

 

 

I think this thread should be used to discuss news on the creation of AGI this year or signs it was made but not revealed such as a organization pretending to make constant AI breakthroughs (to postpone/obscure saying they have made AGI to keep the public calm ect.)

 

I hope that the User liked that I posted this and clarifies his prediction explaining them to you better then I could.

I felt like I was pushing it a bit sharing another users prediction however I thought it was a really important prediction that is coming up soon so I should do so.

[Tempus]

Honestly, I don't think that super AI (2021) can keep up a smooth conversation yet (I'll be the first to be happy if they do), but the path is laid out. As soon as he can maintain a fluid conversation (2025-2030) and knowledge of Psychology can be incorporated, we will see him performing the functions of a therapist, employment advisor, financial advisor, marriage counselor, teacher, secretary, customer service, receptionist, search engine of specialized / scientific information, and why not, as a virtual friend. And that will only be the beginning !!!

 

As for human-level AGI, I think there will be a lot of debate in the coming decades about whether to develop it. IMHO, you don't need human (or higher) level AGI to change the world. We're only beginning to see what highly evolved narrow AI can do.

 

And yes, an exciting decade is coming.

[andmar74]

 

 

As for human-level AGI, I think there will be a lot of debate in the coming decades about whether to develop it.

It's not going to be like that. Some group will put together the software and hardware, and when the AI runs, it turns out it's an AGI.

[Tempus]

 

 

 

As for human-level AGI, I think there will be a lot of debate in the coming decades about whether to develop it.

It's not going to be like that. Some group will put together the software and hardware, and when the AI runs, it turns out it's an AGI.

 

If so, I am calmer. The result of that experiment will not be a highly evolved AGI. It is human-level AGI (or higher) that should concern us.

[Set and Meet Goals]

 

Honestly, I don't think that super AI (2021) can keep up a smooth conversation yet (I'll be the first to be happy if they do), but the path is laid out. As soon as he can maintain a fluid conversation (2025-2030) and knowledge of Psychology can be incorporated, we will see him performing the functions of a therapist, employment advisor, financial advisor, marriage counselor, teacher, secretary, customer service, receptionist, search engine of specialized / scientific information, and why not, as a virtual friend. And that will only be the beginning !!!

 

As for human-level AGI, I think there will be a lot of debate in the coming decades about whether to develop it. IMHO, you don't need human (or higher) level AGI to change the world. We're only beginning to see what highly evolved narrow AI can do.

 

And yes, an exciting decade is coming.

 

Oh I don't mean the prediction was that it will be super, the AGI might be insanely dumb more so that it shows a ability to learn in a general way. The dumb first generation general AI can be improved over the years at a compounding rate with many new versions or software updates ect.

[andmar74]

 

 

If so, I am calmer. The result of that experiment will not be a highly evolved AGI. It is human-level AGI (or higher) that should concern us.

Usually when saying AGI you mean human level AI. I think you are talking about narrow or weak AI.

I do believe AGI will emerge and then it's realized after the fact.

[Set and Meet Goals]

 

 

 

If so, I am calmer. The result of that experiment will not be a highly evolved AGI. It is human-level AGI (or higher) that should concern us.

Usually when saying AGI you mean human level AI. I think you are talking about narrow or weak AI.

I do believe AGI will emerge and then it's realized after the fact.

 

I hope the user clarifies the prediction I believe he means a super weak AGI of sorts.

 

like when I said first gen (which he has said to) I assumed first gen meant it would be nearly as weak as a AGI could be whist still being AGI

[Kynareth]

I doubt it will be next year. First AGI may be a bit like a human baby. It will be tought by humans.

[Set and Meet Goals]

Seems like a big enough breakthrough to add confidence to the AGI by 2021 prediction

 

MuZero: Mastering Go, chess, shogi and Atari without rules | DeepMind