The Singularity - Official Thread

[Ascanius]

I don't really believe in The Singularity anymore. I don't really believe that extreme things will happen. I think that's implausible, unlikely and unrealistic. Maybe good for science-fiction books, movies or games.

I really like Tadasuke's opinion on transhuman technologies, because I also think nanotechnology and brain-computer interfaces aren't coming in the next 20 years. The strong AI won't come until 2030 at the earliest. It's like Ray Kurzweil, just not as optimistic. Research in nanotechnology is not very advanced, the latest is that drugs can be delivered to cancer cells. And I also imagine the brain-computer interface to be complicated.

But with the rejuvenation technique, one could become immortal maybe in 10 years. To do this, one can further develop the cellular reprogramming discovered by Yamanaka. I think this point should also be considered.

[DerrickOlley]

I don't really believe in The Singularity anymore. I don't really believe that extreme things will happen. I think that's implausible, unlikely and unrealistic. Maybe good for science-fiction books, movies or games.

I think in 20 years life will be recognizable and similar to today, but not entirely. More people on the planet will enjoy a higher standard of living. There will be more of everything. No WW3, no asteroid impact, no supervolcanoe eruption. Global GDP will be about double today's. There will be a plethora of smaller and bigger improvements in many things and many places. Our lives will be visibly improved, but not by a huge amount. There will be less pain and suffering among humans, but not gone completely. Fusion will be practical. Large-scale energy storage will be practical. SpaceX Starship will routinely go somewhere. Lots and lots of electric (and sometimes even fully autonomous, even without a driving wheel) vehicles. Airplanes and ships will still run on oil. AR glasses will finally become practical, useful and comfortable, but serious work will still be done on a server, workstation, desktop PC or a laptop (or manually by hand). There will be laptops lasting 85-100 hours on a single charge. AI will be extremely knowledgeable, it will be much wiser and smarter than an average human. AI will present no ill-intent by itself, because it won't be as narrow-minded as some humans. AI and researchers will be still limited by computing resources. More illnesses will be curable or at least manageable.Check it out. An average person will earn more and stuff will get somewhat cheaper because of automation. Working conditions will improve. There will be no widespread UBI or "technological unemployment". Unemployment level will be similar to 2023. There will be more people between 90 and 100 years old. 3D printing organs for transplantation will be a serious thing, that will be gradually getting widespread. Exoskeletons will get practical for various use cases (they will be even replacing wheelchairs). Biped robots will be hired to work in factories, warehouses, supermarkets and other places. There will be no nanobots and neural implants will be used only by people with severe problems + some tech enthusiasts. Normal people will use AR glasses, smartwatches, TVs and laptops. More things will have computers, Internet connection and even displays built into them. ChecThere will be a lot more sensors around. VR and gaming as a whole will be more popular. People will talk, (role)play and learn in virtual realities, but they won't be any more advanced than what was shown in the Ready Player One movie. Designer-babies will be slowly starting to become a thing. There will be artificial wombs available for mothers or fathers to grow their children there, but most people won't accept them yet. Climate will be a bit different, because that's what climate always does - it cyclically changes. Fearmongerers of "misaligned AI" and "climate change" will seem silly by then. But people will find something to be scared about, because that's what some people do. There will still be horror movies and dystopian novels. There will be still people who talk and write about "The Singularity" (and still people who ridicule it). Life will continue.

You talk about a beautiful future! This is quite possible, given that literally 15-20 years ago it was impossible to imagine that you could charge your car through a power outlet using Tesla as an example! That is, life and the environment are improving and the future is already here!

[lechwall]

Is anyone else a "doomer" here that the Singularity will end badly? Doesn't have to be full Yudkowsky we're all going to die but just a general feeling it won't end well?

[Ascanius]

My view on the future:

[Tadasuke]

screenshoted from https://twitter.com/MikePFrank















Against AI Doomerism, For AI Progress: https://maxmore.substack.com/p/against- ... i-progress

PETITION FOR KEEPING UP THE PROGRESS TEMPO ON AI RESEARCH WHILE SECURING ITS TRANSPARENCY AND SAFETY: https://laion.ai/blog/petition/

[Tadasuke]

Some people think that A(S)I will cause progress to go hyper-exponential, causing a technological singularity. But it may be, that what A(S)I will do, is just countermeasure stagnation instead (remember Yuli Ban's Y2K trap?). So without A(S)I, we would see very low rates of progress: 0.5% YoY GDP growth and 3% YoY computing/price growth. With A(S)I, we will see moderate rates of progress: ~4% GDP growth and ~20-25% computing/price growth. But we won't see Singularity-like crazy rate of progress, that some people expect: 20% GDP growth and 120% computing/price growth. So we do need A(S)I, but it won't cause everything to have almost vertical rate of growth.

Keep in mind that 4% YoY GDP growth will accumulate to 2550x in 200 years and 20% computing/price growth will accumulate to 6.86*10^15x (almost 7 quadrillion [in the short scale] times more) in 200 years. So we still can become Type 2 Civilization on the Kardashev scale. But we won't become immortal in 10 years or something like that.

[Tadasuke]

I suggest/propose simplifying rates of progress to three speeds for making things uncomplicated:

For [global] GDP growth:
<2% - slow growth
≥2% & <8% - moderate growth
≥8% - high growth

For [general purpose] computing [for a constant price] growth:
<10% - slow growth
≥10% & <50% - moderate growth
≤50% - high growth

Prior to the 20th century, GDP (and computing) growth was slow. In the 20th century it was usually moderate and I predict that in the 21st century it will be (and it already is) moderate as well. It may happen, that slow or high rate of growth happens for a single year, but it won't be a typical, routine or frequent occurrence.

[Cyber_Rebel]

Personally, I prefer the term "proto-singularity" just like we have a term for proto-AGI or strong narrow AI. Starting to think we should also start seeing the singularity as existing on a spectrum, where you have a slow-middling-hard takeoff depending on the strength of the A.I. in question, and the speed at which it impacts society.

[weatheriscool]

I believe the singularity will occur when either
1. A.i and robotics takes over and becomes the superior being on this planet.
or 2. We merge with it and become that super being.

We are not going to be the weaker creature on this planet. AGI can of course be that superior being and start the unbelievable advancement of tech + science, but the question is will find a humanity that doesn't wish to change useful or will it wipe it out? Humanity would be far better off becoming genetically enhanced in every way manageable along with merging with the interface that is the AGI. This would make us far less likely to be found to be useless as part of the the future system. Of course, I believe we're probably 100-150 years from that system! AGI of some sort could be hear later this decade but as long as we limit it, it will not become superior over us. I could see a skynet if we're dumb enough to build armies for it and grant it power over it so the system could be a lot sooner.

[Powers]

Maybe the singularity as an event may never occur.