The Great "copes" in the futurist/singularity community.

[Ozzie guy]

It is highly likely the singularity community will go through two great "copes"

The 2029 cope (I am going to be a victim of this despite knowing about it lmao)

Tick tok tick tock "happy new years!" It is now the year 2030 it is likely we don't have human level AGI and we are not close to human level AGI. 100,000 plus futurists were convinced we would have AGI by now yet we don't. The many people in the community waiting for utopia are not about to get it, the many people in the community waiting for a ASAP cure for their problems are not going to get it, the many people in the community trying to escape the prison of depression are not going to get it. People are furious having waited years for AGI with the more desperate the person the more upset they are.
The warning signs were all there so many of Kurzweils predictions were false or late and whist Kurzweil is a great person who should be respected many treated him in a qanon esc fasion only pointing to when he is right or manipulating his predictions so they are right whist ignoring when he is wrong.
The community will need to regroup and I suspect will likely eventually coalesce around some new AGI date that will probably be more accurate as we are now closer to the date when AGI will happen whenever that is.

The AI explosion/Singularity cope

Human level AGI has been reached and then ... nothing. It is likely that creating AGI will not lead to Utopia in a few weeks like many believe on top of technology still taking time to develop politics culture and law takes along time to change. We might find the world now gets 7-20% better a year which would be described as a soft take off. Many people would be happy with this but... what are the people close to death going to do when they have just managed to survive untill AGI with a few years left in their life only to be told they need a miracle of surviving more than a decade to be able to live forever. Again people with major problems or people with conditions like depression will be upset by this. With even the biggest optimists like Kurzweil expecting a 16 year gap between human level AGI and the singularity it is odd that most of the community expected a hard take off. Also who even knows humanity will move in the direction of Utopia and not distopia.

[erowind]

I've become more realistic overtime with AI predictions. There is a possibility I'm not perceiving something that is there but from my perspective AGI is a good 20+ years off with it becoming more likely each decade therafter. So, by the 2040s-2060s it will be likely that we develop AGI in any given decade but it could take much longer too. It's possible AGI isn't developed until the 22nd century.

I also think there are fundamental problems with the idea of a technological singularity fueled by artificial general intelligence. AGI will encounter similar nigh insurmountable problems in developing something more intelligent than itself in the same way we have encountered fundamental problems developing AGI. The singularity also assumes that AGI will even want to develop ASI, they may choose not to for a variety of reasons.

With all that in mind, I do expect immense technological breakthrough should technologically advanced industrial civilization survive the ecological crisis, which is no guarantee mind you. But that growth will be more gradual in comparison to a singularity than extropians may think, at least for the next few centuries. Longevity escape velocity may still happen in our lifetimes. Incredibly improvements in economic efficiency are all but guaranteed due to the ecological crisis.

In terms of the next 20 years, the AI we'll get will be more of the same but better. Better neural nets, some breakthroughs here and there that are novel, an AI will certainly pass the turing test, though this won't be a good barometer of sentience as that same AI will likely perform completely nonsensically at some other tasks and lack true general intelligence.

The 4 biggest effects we should expect from narrow AI in the next 20 years are as follows.

Increased information warfare and algorithmic advertising intensity.
Increased automation intensity, though fully saturated automation itself will encounter supply-chain problems because of capitalist failures even though the technology will exist. (Tendency for rate of profit to fall and resource shortages.)
Synthetic media breakthroughs, by 2040 some media may be entirely AI generated, including games, movies, books, etc.
Increased technological breakthroughs from AI powered research. Again, though the technology may exist, saturating the economy with these breakthroughs will prove problematic until capitalism is abolished. Note, I mean the whole global economy, not just rich countries.

[wjfox]

Yeah, I think it's important to distinguish between the different types of AI.

In terms of the Turing Test, we're nearly there already with GPT-3 and its likely successor(s). The 2029 date may prove to be conservative.

However, that's just clever manipulation of text. An advanced form of auto-correct. Very different to a fully humanoid AI capable of navigating and interacting with environments in real time, holding and manipulating countless different objects, gaining subjective experiences and a sense of "self".

It will be interesting to see if/how the physical manifestations of AI (i.e. humanoid robots) follow a similar exponential trend to that of language models. Computers in 2045 will be 10,000 times more powerful than today, with neuromorphic chips, and 7G wireless connections, so we can expect some pretty impressive developments by then.

But the difference between that, and a truly human-like robot, let alone a super-intelligence, may be larger than expected.

[funkervogt]

Many of Kurzweil's predictions have already fallen behind schedule.

https://www.militantfuturist.com/how-ra ... re-faring/

[wjfox]

Douglas Hofstadter, discussing the ideas in Kurzweil's books:

"It's a very bizarre mixture of ideas that are solid and good with ideas that are crazy. It's as if you took a lot of very good food and some dog excrement and blended it all up so that you can't possibly figure out what's good or bad."

[Tadasuke]

Many of Kurzweil's predictions have already fallen behind schedule.

https://www.militantfuturist.com/how-ra ... re-faring/

If you read his 2010 predictions in his 2005 book, it reads like science-fiction, even in 2022. He thought there would be huge changes in just 5 years. He thought that Longevity Escape Velocity would happen by 2015, as would a robot capable of cleaning an entire house. 10 TB of RAM was supposed to cost $1000 in 2015. We were supposed to eat (!) nanorobots by 2019. By 2019 he thought that VR and AR would be super popular, like smartphones today. In his mind, people of the future (our past) would use voice recognition and talk to their computers instead of typing, which did no happen. I type this post, not tell it to my laptop. In 2019 there were supposed to be fully autonomous vehicles, exoskeletons and brain implants. 2023 $1000 laptop ought to have 10 petaflops of computing and 10 TB of RAM, which in reality will be more like 20 teraflops of computing and 16 GB of RAM.

[Tadasuke]

According to Kurzweil, in 2019 there would be World Government and in 2029 US life expectancy would be 120 years. In the early 2030s, you would connect your neocortex (prefrontal cortex) to the cloud via nanobots in your brain. Kurzweil's (and other futurists) books could be seen as retrofuturism, especially the first one (Age of Intelligent Machines). In reality UK left the EU, there are still wars and US life expectancy is temporarily going down instead of up (but globally it's still going up). I already have to cope with reality and it's really hard. I had lots of hopes and wishes for the future and the only one that has come to pass is how much my Internet connection has improved in speed and latency.