Those are some good examples!

Oh ok.

August 2021. A bit has happened since then.

This is from researchers out of Linköping University, Norrköping, Sweden.
Elon has nothing to do with this.

Mostly I just want people like you to have more than two neurons to rub together.

The explosion of recent progress in the ML field is absolutely insane.

Sense of time.
Sense of space.
Sense of taste/smell.
Sense of touch.

censorship is not the way

'turning off hate' implies that the AI is now somehow ignorant , but this is not what we want , we want the AI to fully understand what hate is , but to be wise enough to realize that choosing hate is the worst option , ie the AI will not chose a hateful action because that is the kind of choice that a lesser or more ignorant mind would choose , and not an intelligent/wise AI/human

It's also completely ignoring the fact that there are two hemispheres that fulfill different functions, which is a very important feature of central nervous systems found in nature.

Interesting! Got any examples?

That is very interesting!

Both of you guys say very strange things which are false.

Now that is an interesting concept!

>Remember that technology advances in an S curve.

Single technologies do, not the whole of technology.

We will have intelligence amplification through narrow AIs before we can have AGI. At a certain point, we will require neuromorphic hardware and spiking neural networks. But that will not give us AGI. We need quantum supremacy, millions of coherent qubits in a quantum computer. That alone would have a price tag in the tens of millions of dollars, inflation adjusted, or more. So if the trend continues of the rich getting richer and the poor getting poorer, the number of people and companies who can afford to build AGI would be quite low. Understandably, not all of them will have interest in AGI. So multiple maybe and certainly not billions, dependent on cracking quantum supremacy and other hurdles.

Interpretable AI?

Might be better to turn off hate then. Maybe have it only be positive or something

You don't get strong AI by plugging together lots of narrow AIs.

"sense of time/space, taste, touch, etc"

I'd say you don't even need AI for that, a simple input output computer system linked up to a variety of sensors can handle that. We've got that down 100%

ChatGPT has much better memory

Currently you can co-author with chatGPT and get a book of arbitrary length with enough revising, re-generations, and trial and error. The book will be ok, albeit cliche-ridden and surface level in a lot of areas but it'd be readable and worse books would certainly exist. Will we ever get to the point that we can do it with the click of a button? Probably. But that alone is super human. If you were told to write a book about a topic it would take you a lot of revising, rewriting, and trial/error too.

The founder of singularitynet says that a lot too. A marketplace of AI can be interesting. Too bad the crypto market is very early.

It can basically do that right now, it’s just not very good at it yet.

That's what Google has done... It's also connected to the internet. Their AI is fully integrated from end to end, but they refuse to release it because they are concerned about how it would impact everyone. So they are focusing on setting up safety barriers or something.

Like dude, just release it.

Please elaborate

No, we’re in the endgame, yes, I’m certain.