An internet chum wrote to me from California yesterday, saying, "Stephen Hawking warns artificial intelligence could end mankind." Although largely ignorant of the tendrils that artificial intelligence puts out, I find myself agreeing and simultaneously wondering what others thought.
Here's my vaporous rambling on the topic: http://genkaku-again.blogspot.com/2017/05/ai-and-juicy-bits.html
Comments
It is always interesting that people have a tendency to push curiosity to the point of self-destruction when it seems obvious to others of us exactly that they are doing. I wonder what AI would do with the world if it did us all in. It seems AI would be the apex of human technological wonders, and the more we move that direction the more we seem to lose our connection with our emotions and our deeper inner workings. So many people have arrived at a point where science and technology is all they are interested in. Changing the course for the human experience for all of us as they go along.
I do always enjoy your noodling writings
It is a bit of an unknown. It rather depends whether you are thinking about an unlimited, highly intelligent and self-modifying AI, or something a bit more limited, like HAL from the movie 2001: A Space Odyssey.
The first could start rewriting itself and might find out more overnight about the physical universe from for example CERNs particle collision data about the real universe than we humans might discover in another three decades of analysis. While the risks with the second are a lot smaller.
I'll read the blog when my tablet's battery has more oomph.
(You do know, don't you, how HAL was named...? He was one Byte removed from IBM.... )
Elon Musk hopes to get ahead of the problem by merging human and artificial intelligence. By giving humans a neural link to computers, he's trying to prevent humans from becoming irrelevant. But would a cyborg still be human?
http://www.businessinsider.com/elon-musk-neuralink-connect-brains-computer-neural-lace-2017-3
The problem is not AI but BS (biological stupidity). At the moment AI, robotics and the search for wet wear integration for the post augmented/mixed reality generation is a more intense ability to be human. That is increased capacity to know, learn, feel, play, create, explore, love, become enlightened etc. Quite a ways to go ...
https://www.theregister.co.uk/2017/01/02/ai_was_the_fake_news_of_2016/
When I think of AI control, I think of something that's been with us for many years... "traffic lights" and how most (but not all) of us obey their command, in fact we fear the consequences if we disobey them (even the programmers obey without question)...Mind you, in saying this we don't have any traffic lights on the island
I mostly fear the consequence of getting smashed into by a logging truck if I don't stop for the magic red light, lol. Our town has 2 stoplights. But when I lived in a bigger city I emailed for a while with a traffic engineer who told me all about how they do the timing and such for the lights. Pretty fascinating. I'm not sure I'd consider lights AI because they are in completely control of humans and programmed by them. They don't make decisions based on anything, they are simply triggered by sensors.
I love reading about AI and Singularity even though it's terrifying.
I think Exponential Technological Growth is the path we are on right now. Between 2000-2010 we got smartphones, a robust global Internet, social media, etc etc. I think 2010-2017 has been a bit of a lag as we are trying to explore how these developments fit into our every day lives. It has been a huge upheaval in how we communicate with others and ourselves.
I read in Popular Mechanics that the next Big Step is to enter a post-smartphone world, where we are in a mix of virtual reality and AI.
As for the Singularity... I don't know if everyone would voluntarily submit their consciousness to a computer. Unless the Earth is on the brink of death. If that's the case then I'll be right in my assumption we sacrifice our planet for technology. I guess billions of years of biological and geological eras cant hold a candle to a few decades of gadgets and gizmos
What is more important is what is real. What is real is that we are all mortal human beings.
-It seems to me, in a way, AI already has. Look about the next time you go out. Do you see the masses with their heads down peering into their smart phones as the walk, eat, drive and sit besides one another? These folks are present but not present, they have virtual friends but feel alienated and lonely. There are consumers, yet they are the ones being consumed by this soma...
There are different forms of AI, the most important is transformative AI
http://www.openphilanthropy.org/blog/some-background-our-views-regarding-advanced-artificial-intelligence#Sec1
AI has early forms that appear to indicate 'something wonderful'. In fact they are nowhere near singularity or free of very rudimetary rule based programming and data processing. Getting better though ...
The problem is that in order to get anywhere you need to satisfy a number of different conditions, such as being able to learn, think logically, plan towards a goal, and so on. Each of these problems have been addressed to some extent individually in different coding techniques, but no-one has been able to combine them all into a single structure and upgrade them so that it makes sense.
I don't know if any of you have been following AlphaGo's exploits? It is an attempt to create a Go-playing AI which was just recently involved in a series against the current world number 1 player, Ke Jie. It's a fine example of deep learning AI which is a key step towards a more integrated AI system, but even there it's learning has been carefully managed by human hands. Still it's interesting that in a few years Deepmind, the company that built it, was able to absorb and surpass the sum total of human knowledge in this area.
Next to AlphaGo another example of current AI research is Wolfram Alpha, which uses a database of curated information to answer questions, make calculations and do a variety of other things. It's an example of the kind of thing these systems are already good at today.
My friend who is a tv/film editor posted this link on Facebook and posted 'oh shit' (because it's about current 'good enough' progress making AI that can edit movies and tv). http://alex4d.com/notes/item/automated-video-editing-is-now-good-enough
I definitely think AI is coming and it will bring changes.
Thanks for video @Shoshin, early days. I have been interacting with bots for years, quickly bored by their simulated attempts at engagement ... The one in the video needs an axe through the skull protection option
https://chatbotsmagazine.com/which-are-the-best-intelligent-chatbots-or-ai-chatbots-available-online-cc49c0f3569d
When a bot get's past @federica, then the simulation is impressive enough to take seriously ...
I've been mistake for a bot on more that one occasion
I would like to take this opportunity to assure other members that I'm bot oopss not