Technological Singularity - it'll happen? When?

[Kazarogkai]

From what I remember moore's law is actually slowing down, we are reaching the end of silicon chip technology. We're going to have to advance to get any further. The problem is tech like that is driven more or less by need and since silicone chips are good enough for roughly 99.99% of all foreseeable applications the need and ultimately desire won't really be there. Mix that in with human tendency to stagnate and you got problems. If anything like that does occur I doubt it will happen no earlier than maybe the 22nd century at best. Thank goodness since I will be dead, I hate robots. That's not taking into account humanities current raping of earth and the inevitable bite back(environmental collapse) which is already in process and is effectively irreversible. Apocalypses have a tendency to backtrack tech and society in general.

Just my thoughts on the matter.

[Salandriagado]

I think that it rather depends on how quickly either quantum computers emerge or how far the IoS develops. We are coming to the end of moores law so we have two avenues to go down really, a new type of computer which can burst through the barriers imposed by silicon and standard binary computing or distribute computing over a huge range of device each feeding information back to the whole which would engage in deep learning. I think the first one is a more likely path to the singularity than the second as even with deep learning I don't think the paradigm is currently right for a strong AI to emerge.

If the rate at which we develop breakthroughs in QC stays at the same pace or even accelerates (fingers crossed the law of accelerating returns actually holds) we could see large scale Quantum computers within our lifetimes I'd say with the ability to develop strong AI on them. It's not a solid prediction, there are many variable, but I think it's reasonably likely.

That's not how QC works. QC is ridiculously slow for most jobs, just happens to be very fast for a handful of very specific jobs. There's no relationship between Strong AI and QC.

[Uiiop]

Oh boy another transhumainism theard

Something's gonna happen but nothing to be overly hyped about until i'm almost a dead man at the least.
Technology and life change each other all the time. Just because we can fly in the 1900's didn't mean we can easily go to space now.

[The Serbian Empire]

I expect that it will never be reached. The threats of climate change and lack of materials necessary to make such complex AI and never run out in the first place is a major limitation on the matter.

Climate change?
Materials?
Why?
That thing should be initially barely similar to a supercomputer, not something requiring strange and rare materials.

Humanity has far more important things to protect and solve than to develop a supercomputer that reaches HAL 3000 levels.

[Olivaero]

I think that it rather depends on how quickly either quantum computers emerge or how far the IoS develops. We are coming to the end of moores law so we have two avenues to go down really, a new type of computer which can burst through the barriers imposed by silicon and standard binary computing or distribute computing over a huge range of device each feeding information back to the whole which would engage in deep learning. I think the first one is a more likely path to the singularity than the second as even with deep learning I don't think the paradigm is currently right for a strong AI to emerge.

If the rate at which we develop breakthroughs in QC stays at the same pace or even accelerates (fingers crossed the law of accelerating returns actually holds) we could see large scale Quantum computers within our lifetimes I'd say with the ability to develop strong AI on them. It's not a solid prediction, there are many variable, but I think it's reasonably likely.

That's not how QC works. QC is ridiculously slow for most jobs, just happens to be very fast for a handful of very specific jobs. There's no relationship between Strong AI and QC.

What "jobs" is QC slow at?

[Salandriagado]

That's not how QC works. QC is ridiculously slow for most jobs, just happens to be very fast for a handful of very specific jobs. There's no relationship between Strong AI and QC.

What "jobs" is QC slow at?

Right now: absolutely all of them. In future: most of the ones that normal computers are slow at. QC allows for shortcuts in solving a handful of very significant problems, but is in no way a general improvement in all tasks.

[Uiiop]

One of the problems with singularianism is that there's only one inevitable way that humanity will change. And it chooses something that isn't even close to causing anything like they except.
Bioengineering seems at least sightly more plausible in causing a shift IMHO than AI.

[Olivaero]

What "jobs" is QC slow at?

Right now: absolutely all of them. In future: most of the ones that normal computers are slow at. QC allows for shortcuts in solving a handful of very significant problems, but is in no way a general improvement in all tasks.

But it also has the potential to be smaller yes?

[Salandriagado]

Right now: absolutely all of them. In future: most of the ones that normal computers are slow at. QC allows for shortcuts in solving a handful of very significant problems, but is in no way a general improvement in all tasks.

But it also has the potential to be smaller yes?

No. Probably much larger. A fundamental problem with quantum computing is that the more qubits you put near each other, the more likely they are to fail.

[Olivaero]

But it also has the potential to be smaller yes?

No. Probably much larger. A fundamental problem with quantum computing is that the more qubits you put near each other, the more likely they are to fail.

Right, there was my misunderstanding then.

[Great Nepal]

I think it's rather inevitable assuming we don't end up in a cataclysmic disaster where humanity's technological progress is significantly and permanently damaged; Moors law seems to be slowing down but we also have research in the field and problems can be bypassed using different materials etc.

---

If the singularity occurs and we control or cooperate with the AIs, the super-intelligent computer/AI should be able to design and create cool cyborg bodies for all us humans so that we can transplant our brains into them and live effectively forever.

And as well as super-improved lifespans and health, a body improved by cybernetics may have a massively increased and improved intelligence and emotional capacities, and we would be able to customise physical appearances to a much greater extent.

It seems like a bit of fantasy now, but surely if the singularity occurs there would be no reason why immortality and a huge boost in human intelligence wouldn't occur just afterwards?
Because surely asking a post-singularity AI to create a system where we can all be immortal through cybernetics would be tediously simple for such an advanced machine?

Idk, maybe i'm thinking too much into this.

Hmmm I'm not sure on cyborg bodies; they still limit the growth and are ultimately rather meh given they still have physical limitations. Much better to establish some sort of mind upload system.

How is it going to get up to that IQ without upgrading its hardware? Upgrading hardware is also a physical process.

Does it need to upgrade its hardware? We have a pretty nice distributed system which will eventually contain to hundreds of trillions of computers including some rather powerful machines. I imagine a sufficiently intelligent AI would be able to optimize its operations, and operations of other programs as well as bypassing saftey protocols thus effectively hiding its operation, or alternatively market itself as a program which massively improves performance of your device - say registers a company, contacts media for advertising campaigns and provides awesome service for say ten pounds. Eventually as success of new software becomes clearer, even those in industry originally guarded will give in - either by themselves or because management doesn't want to spend millions trying to replicate something when tenner does the same thing. Once sufficiently embedded into our computing infrastructure, it can inconspicuously work on improving robustness of the internet.
Now it has access to the best hardware on the planet, and has embedded itself into our lives so deeply that you can not possibly consider eliminating it without entirely decimating the global economy.

[Ifreann]

If the singularity occurs and we control or cooperate with the AIs, the super-intelligent computer/AI should be able to design and create cool cyborg bodies for all us humans so that we can transplant our brains into them and live effectively forever.

And as well as super-improved lifespans and health, a body improved by cybernetics may have a massively increased and improved intelligence and emotional capacities, and we would be able to customise physical appearances to a much greater extent.

It seems like a bit of fantasy now, but surely if the singularity occurs there would be no reason why immortality and a huge boost in human intelligence wouldn't occur just afterwards?
Because surely asking a post-singularity AI to create a system where were can all be immortal through cybernetics would be tediously simple for such an advanced machine?

Idk, maybe i'm thinking too much into this.

Not really. I think a future populated by cyborgs is more realistic than one where humans become obsolete and are replaced by machines.

Soon all people shall be upgraded to include rotating cannon turrets.

[Chessmistress]

Oh boy another transhumainism theard

Something's gonna happen but nothing to be overly hyped about until i'm almost a dead man at the least.
Technology and life change each other all the time. Just because we can fly in the 1900's didn't mean we can easily go to space now.

Transhumanism is another thing: transhumanism is about improvement of humans, technological singularity is about artificial intelligence not humans.

[Frank Zipper]

So how do people think we will first know the singularity has happened? I assume we would first notice stock market movements.

[Lady Scylla]

So how do people think we will first know the singularity has happened? I assume we would first notice stock market movements.

I feel it won't be confirmed until it's already happened.

[Trotskylvania]

So how do people think we will first know the singularity has happened? I assume we would first notice stock market movements.

When the Machine Messiah proclaims itself

[Ifreann]

So how do people think we will first know the singularity has happened? I assume we would first notice stock market movements.

When the Machine Messiah proclaims itself

Probably by picking a fight with the Avengers.

[Uiiop]

Oh boy another transhumainism theard

Something's gonna happen but nothing to be overly hyped about until i'm almost a dead man at the least.
Technology and life change each other all the time. Just because we can fly in the 1900's didn't mean we can easily go to space now.

Transhumanism is another thing: transhumanism is about improvement of humans, technological singularity is about artificial intelligence not humans.

But the whole reason it's called a Singularity is because of the magnitude and uncertainly of the event's impact of humanity.
It's related at least by the people hyping it.

[Ostroeuropa]

It'll happen unless something goes wrong. I'm a transhumanist, so i'm in favor of it.

[Great Nepal]

So how do people think we will first know the singularity has happened? I assume we would first notice stock market movements.

I don't think we will really know at the time, we'll likely notice greater pace of technological advancement but I doubt we'll call it "singularity". We'll likely look back and say "ohh that was when technological singularity happened".

[Frank Zipper]

High level hacking attacks increasing, would also seem to me a likely indicator.

[USS Monitor]

How is it going to get up to that IQ without upgrading its hardware? Upgrading hardware is also a physical process.

Does it need to upgrade its hardware?

Yes, it does. This should be really fucking obvious.

We have a pretty nice distributed system which will eventually contain to hundreds of trillions of computers including some rather powerful machines. I imagine a sufficiently intelligent AI would be able to optimize its operations, and operations of other programs as well as bypassing saftey protocols thus effectively hiding its operation, or alternatively market itself as a program which massively improves performance of your device - say registers a company, contacts media for advertising campaigns and provides awesome service for say ten pounds. Eventually as success of new software becomes clearer, even those in industry originally guarded will give in - either by themselves or because management doesn't want to spend millions trying to replicate something when tenner does the same thing. Once sufficiently embedded into our computing infrastructure, it can inconspicuously work on improving robustness of the internet.
Now it has access to the best hardware on the planet, and has embedded itself into our lives so deeply that you can not possibly consider eliminating it without entirely decimating the global economy.

1. What you described is a method of the AI upgrading its hardware.

2. This would take a great deal of time, and humans would be able to see the process as it was occurring. That's not technological singularity because it's still progress of a sort that is comprehensible to humans.

3. For an AI to reach a level of intelligence that it would be capable of doing what you describe, it'd probably have to upgrade its hardware by other means to even get to that level.

4. There is still a finite amount of data that can be stored on existing hardware, and the time it takes to transmit data between scattered devices would slow the AI down compared to a more compact and technologically advanced system that could be built (i.e. a centralized one not dependent on silicon chips). "The best hardware on the planet" is still pretty shit compared to the best that is physically possible, and I don't think current hardware is sufficient for a technological singularity. Short of installing wormholes, improving internet connectivity wouldn't eliminate the effects of distance if the AI is using hardware scattered around the planet. It still would slow things down, at least a little. Computers are not magic. They are physical objects, and that means they are limited by the laws of physics, so they are affected by things like distances and the properties of the materials they are built from.

[USS Monitor]

So how do people think we will first know the singularity has happened? I assume we would first notice stock market movements.

If you have to ask whether it's happening, it isn't.

[Immoren]

So how do people think we will first know the singularity has happened? I assume we would first notice stock market movements.

When the Machine Messiah proclaims itself

Time to get that Mars colony going.

[Chessmistress]

It'll happen unless something goes wrong. I'm a transhumanist, so i'm in favor of it.

I still don't understand why people insist putting together technological singularity and transhumanism.
There isn't such huge correlation between transhumanism and technological singularity.