The death of AI (yet again)

#81
Indeed.

Cross-posted to the "AI program beats human 3-0 at ancient Chinese game GO - thoughts?" thread:
"Godlike" Artificial Intelligence Just Officially Beat The World's #1 Go Player


For his part, Ke was unintimidated by AlphaGo's rising dominance in his chosen game, boasting last year that he would never lose to an AI.

By the end of Tuesday's contest - which achieved the closest possible result, with AlphaGo winning by just half a point - it's fair to say Ke had adopted a new attitude.

The human champion, who appeared visibly baffled at times through the match, described the AI as peerless, even divine.

"I feel like [its] game is more and more like the 'Go god'. Really, it is brilliant," he said at a press conference afterwards, wishing to never have to go again through such a "horrible experience".

"In the past it had some weaknesses," he added, "but now I feel that its understanding of the Go game and its judgments are beyond our ability."
 
#82
Question (open-ended): are "understanding" and "judgements" (referenced in the above) predicated on consciousness, or can (non-conscious) intelligence be fairly claimed (as in the above) to possess/make them?
 
#83
Having worked in IT, AI has always interested me although I think the label 'artificial intelligence' is very misleading. No real intelligence has yet appeared despite very clever algorithms and incredible processing speeds just good simulations. Here is a quote from an article in Bulletin of Atomic Scientists which may interest folks:

"Although it has failed so far to realize the dream of intelligent machines, artificial intelligence has been one of the greatest intellectual adventures of the last 60 years. In their quest to understand minds by trying to build them, artificial intelligence researchers have learned a tremendous amount about what intelligence is not. Unfortunately, one of their major findings is that humans resort to fallible heuristics to address many problems because even the most powerful physically attainable computers could not solve them in a reasonable amount of time. ... As a consequence, both the peril and the promise of artificial intelligence have been greatly exaggerated."
 
#84
Having worked in IT, AI has always interested me although I think the label 'artificial intelligence' is very misleading. No real intelligence has yet appeared despite very clever algorithms and incredible processing speeds just good simulations.
Fully agree. Its a complete misnomer in my view. If you want to keep the acronym, at least swap "Artificial" with "Automated". I'm not wild about the use of "Intelligence" at all.
 
#85
Having worked in IT, AI has always interested me although I think the label 'artificial intelligence' is very misleading.
From an IT point of view, the term "artificial stupidity" has also been used. That at least gives a caricature of the gap between what is termed "intelligence" in human and machine usages.

It is also a term which is very hard to define. I.Q. has been defined as "that which is measured by I.Q. tests", there is a certain circularity to it.

Years ago I held the view that intelligence was a survival characteristic. But under that definition the diesel engine would also be intelligent, since its ability to survive has been pretty strong. I'm not sure what a good definition would be .
 
#86
Having worked in IT, AI has always interested me although I think the label 'artificial intelligence' is very misleading. No real intelligence has yet appeared despite very clever algorithms and incredible processing speeds just good simulations. Here is a quote from an article in Bulletin of Atomic Scientists which may interest folks:

"Although it has failed so far to realize the dream of intelligent machines, artificial intelligence has been one of the greatest intellectual adventures of the last 60 years. In their quest to understand minds by trying to build them, artificial intelligence researchers have learned a tremendous amount about what intelligence is not.
Unfortunately, one of their major findings is that humans resort to fallible heuristics to address many problems because even the most powerful physically attainable computers could not solve them in a reasonable amount of time. ... As a consequence, both the peril and the promise of artificial intelligence have been greatly exaggerated."
I absolutely agree with the part of your quote that I have emphasised! It is a great shame that this understanding about what intelligence isn't was not fully appreciated because it embarrassed too many researchers who had assumed AI would be easy once the computer power was available.

I find myself less in agreement with the rest of what you wrote - though clearly we do take some risky short-cuts to reach conclusions -, although I recognise it as a common excuse for the failure of AI. If you want to pick a specific example, we could discuss it more concretely.

My feeling is that AI researchers endlessly take conceptual ideas - like 'tall' for instance - and morph them into something far too concrete. For example, there was the craze for 'fuzzy sets' where a word like 'tall' became a function of someone's height that returned a number between 0 and 1. I don't think this captured the essence of the meaning of the word 'tall', which means something like 'of a surprisingly large height given the circumstances'. For example 'tall' applied to a child of 5, or maybe someone with dwarfism, would obviously not mean the same as applied to an adult.

David
 
Last edited:
#87
I would prefer the term 'simulated intelligence' to AI. I am no expert in the field and I do think current 'AI' applications like helper bots etc are a good thing when applied sensibly.

I think time will prove that the materialist concept of real AI and immortality through uploading our consciousness are just fiction. Reality is much more interesting as NDEs suggest.
 
#88
I would prefer the term 'simulated intelligence' to AI. I am no expert in the field and I do think current 'AI' applications like helper bots etc are a good thing when applied sensibly.
OK - but do they differ really from ordinary programs? I mean, this is the point, for the concept of AI to be meaningful, it has to do something special, and even then, you have to ask if t is 'intelligent' in any way. In the 80's, some AI programs came out that would do algebra and calculus - something that certainly looked impressive. However, symbolic differentiation can be done by a simple recursive algorithm, and symbolic integration by pattern matching - accepting that the program will fail to spot some cases. However, nowadays there are much more powerful programs - such as Mathematica - that make no pretence of being based on AI.
I think time will prove that the materialist concept of real AI and immortality through uploading our consciousness are just fiction. Reality is much more interesting as NDEs suggest.
Agreed!

David
 
#91
From an IT point of view, the term "artificial stupidity" has also been used. That at least gives a caricature of the gap between what is termed "intelligence" in human and machine usages.

It is also a term which is very hard to define. I.Q. has been defined as "that which is measured by I.Q. tests", there is a certain circularity to it.

Years ago I held the view that intelligence was a survival characteristic. But under that definition the diesel engine would also be intelligent, since its ability to survive has been pretty strong. I'm not sure what a good definition would be .
Hmmm... Try leaving a diesel engine entirely to its own devices and let me know how well it survives :D;)
 
#95


http://www.businessinsider.com/robot-lawyer-chatbot-2017-7
The free robot lawyer that appealed $3 million in parking tickets is now available in the US
...
The bot is based on a conversation algorithm, meaning it uses keywords, pronouns, and word order to understand a user's issue. Browder previously told BI that the more people use the bot, the more intelligent it becomes. The algorithm can quickly analyze large amounts of data while improving itself in the process. Browder told The Verge that the bot has helped strike down 375,000 parking tickets in the last two years.
The video shows it does a lot more than help with parking tickets.

https://donotpay-search-master.herokuapp.com/
 
#96
It's interesting that even that famous (or infamous) skeptic of the paranormal Martin Gardner was really a Mysterian. Perhaps the only area in which I agree with him:

.....no computer of the sort we know how to build — that is, one made with wires and switches — will ever cross a threshold to become aware of what it is doing. No
chess program, however advanced, will know it is playing chess anymore than a washing machine knows it is washing clothes. Today’s most powerful computers differ from an abacus only in their power to obey more complicated algorithms, to twiddle ones and zeroes at incredible speeds.
(In Do Loops Explain Consciousness?, review of "I Am a Strange Loop" by Douglas Hofstadter, at http://www.ams.org/notices/200707/tx070700852p.pdf .)
 
Last edited:
The real IT/SiliconValley world is after AI, like they are starving animals of prey. There is something they want; to consume for growth and evolution. AI/Brain interfaces are what they are seemingly after.
Google’s artificial-intelligence guru, Demis Hassabis, has unveiled the company’s grand plan to solve intelligence by unraveling the algorithms, architectures, functions, and representations used in the human brain. But is that all there is to it?

No one disputes the basics of artificially intelligent neural networks, namely that brain neurons are connected by synapses with “weights” that grow stronger (learn) the more they are used and atrophy (forget) when seldom used. The European Union’s Blue Brain project, for instance, hopes to simulate on a supercomputer even the smallest details of how the brain works, so as to unravel the mechanisms behind maladies such as Parkinson’s and Alzheimer’s, as well as to build AI systems.
no hidden conspiracy going-on here. this is right out in the open
http://www.eetimes.com/document.asp?doc_id=1332058
 
Top