rehess
Loc: South Bend, Indiana, USA
GerryER wrote:
They should change the name from AI to IA, Intelligent Algorithm. That's all it really is, a computer algorithm , nothing "artificial" about it.
Do you know how “AI” works? It comes from attempts to have computers follow ‘rules’ of ‘thinking’ as an expert does. In that sense it
is “artificial” - it is not “real” thinking.
rehess wrote:
Do you know how “AI” works? It comes from attempts to have computers follow ‘rules’ of ‘thinking’ as an expert does. In that sense it is “artificial” - it is not “real” thinking.
I realize that; that is what an algorithm is, a formula or "rule" to follow. And, now that I think about it, whoever comes up with the algorithm may prevent it from being intelligent. as well. ( I guess that would make it AU, artificial unintelligence.) :-)
rehess
Loc: South Bend, Indiana, USA
GerryER wrote:
I realize that; that is what an algorithm is, a formula or "rule" to follow. And, now that I think about it, whoever comes up with the algorithm may prevent it from being intelligent. as well. ( I guess that would make it AU, artificial unintelligence.) :-)
My iPhone has a ‘feature’ which identifies an animal by name if it “sees” it in a photo - then tries to tell you about it. I suppose they ‘picked the brains’ of zoologists in constructing that feature.
Even from the limited view below, it identifies ‘Annie’ as a “Tortishell Cat”
Canisdirus wrote:
????
Put the bong down...
Don't scoff. It's happening in multiple disciplines.
larryepage wrote:
Don't scoff. It's happening in multiple disciplines.
Really...the loom industry from the 1800's?
Reading comprehension...
rehess
Loc: South Bend, Indiana, USA
Canisdirus wrote:
Really...the loom industry from the 1800's?
Reading comprehension...
Many professionals are being displaced, either by smarter tools {think word processing, spread sheets} or by doing things in a more efficient manner {Covid caused changes - such as zoom - to be used sooner}.
When I was retiring, Universities were just beginning to teach ‘remotely’. Now Big Ten schools like Purdue and Univ of Maryland are advertising “Global Campuses” , where one lecturer could conceivably teach the equivalent of several lecture-halls full of students.
This reduces work for Resident Hall staff, and for people who would otherwise teach / work at smaller schools.
Canisdirus wrote:
Really...the loom industry from the 1800's?
Reading comprehension...
Jacquard's loom programming cards are really not a good example for this discussion. They enabled accuracy, consistency, and repeatability. They did not design fabric patterns or even provide a safety valve against unattractive color combinations. They really didn't increase output too much above what a skilled operator could do. But they did make the job easier and reduce errors, improving product quality.
Standard computing eliminated the need for manycraft positions. The prognosis is that AI will do the same for professional positions.
This is the gist of AI. A computer is a truly dumb device - it doesn't know how to do anything except execute instructions given to it by a human programmer. Because of this, AI software will only be as good as the quality of the programming (i.e., limited to human "perfection"). And it will only be as mistake-free as the programmers can make it. That's why I see AI implemented by businesses as the ultimate demise of AI - companies that sell AI-based products will encounter bad decisions made by their AI products as the programming encounters situations no one anticipated, or the decisionmaking logic neglected troublesome unintended consequences. Companies will find themselves liable for the damages caused, and that will be the reason businesses will back away from AI.
We already have computer controlled passenger aircraft, and the inabilities of the computer programming has been responsible for a few crashes and some near misses. When AI is used to fully control automobiles, a high percentage of the mistakes the AI will make will kill people. Manufacturers won't endure that for too long.
In addition to AI, quantum computing is making significant progress with several successful trial runs by multiple different organizations.
rehess
Loc: South Bend, Indiana, USA
bflood wrote:
This is the gist of AI. A computer is a truly dumb device - it doesn't know how to do anything except execute instructions given to it by a human programmer. Because of this, AI software will only be as good as the quality of the programming (i.e., limited to human "perfection"). And it will only be as mistake-free as the programmers can make it. That's why I see AI implemented by businesses as the ultimate demise of AI - companies that sell AI-based products will encounter bad decisions made by their AI products as the programming encounters situations no one anticipated, or the decisionmaking logic neglected troublesome unintended consequences. Companies will find themselves liable for the damages caused, and that will be the reason businesses will back away from AI.
We already have computer controlled passenger aircraft, and the inabilities of the computer programming has been responsible for a few crashes and some near misses. When AI is used to fully control automobiles, a high percentage of the mistakes the AI will make will kill people. Manufacturers won't endure that for too long.
This is the gist of AI. A computer is a truly dumb... (
show quote)
There is a big difference between “computer controlled” and “AI”; don’t blame the latter for things caused by the former.
Actually, the true “gist” of AI is a machine learning from the results of its actions and from the knowledge of experts.
rehess wrote:
There is a big difference between “computer controlled” and “AI”; don’t blame the latter for things caused by the former.
Actually, the true “gist” of AI is a machine learning from the results of its actions and from the knowledge of experts.
It still needs a programmer to tell it what to learn, i.e. what is a "good" result.
rehess
Loc: South Bend, Indiana, USA
moonhawk wrote:
It still needs a programmer to tell it what to learn, i.e. what is a "good" result.
Not an accident.
Do you have any reason at all to believe that the problems with ‘self-driving’ cars is a result of “AI” as that term is normally used????
If I'm understanding your question correctly (I'm old and slow), the answer is still that the issue is who or what originally programmed the AI. It can't think for itself, it can only react to circumstances and compare them to its data base.
rehess wrote:
Not an accident.
Do you have any reason at all to believe that the problems with ‘self-driving’ cars is a result of “AI” as that term is normally used????
You could also have issues of defective technology, parts failures etc, which could alter the desired outcome.
If you want to reply, then
register here. Registration is free and your account is created instantly, so you can post right away.