|
|
Post by Orac on Aug 27, 2023 9:14:55 GMT
Orac, i am getting happier and happier to avoid the future you describe. The saying " I'd rather be dead" is quite right. . I think your feelings are shared by many, including myself You may have noticed that, up to about a decade or so ago, ignoring the internet was still somewhat of an option, but now our society is the internet.Someone needs to start a counter- revolution in which real communities are started that don't involve themselves in what is going to become a complete disaster.
I'm not sure what the answer is but i'm increasingly of the opinion that the Amish may have had a point |
|
|
|
Post by Vanna on Aug 27, 2023 10:46:30 GMT
Agree wholeheartedly with those sentiments. An alternative internet is long overdue. And the Amish? I would gladly join one of these things if religion could be ruled out. Other than the religious aspect, the Amish understand the principles of duty, courage, diligences, understanding, patience, and intelligent compassion. The west is fast losing these ancient values that our own ancestors held vital to functioning communities. Tradition (the wisdom of the elders) is also vanishing in the glare of woke.
If ever anyone starts an alternative internet, let us know.
|
|
|
|
Post by bancroft on Aug 27, 2023 11:13:36 GMT
@vanna how does the guest setting work?
|
|
|
|
Post by Vanna on Aug 27, 2023 13:21:04 GMT
The guest poster doesn't have a setting and every time you post only on an open thread you give a name and an assurance you are not a robot with a click box tick that you accept the board's rules.
|
|
|
|
Post by bancroft on Aug 27, 2023 13:54:39 GMT
Are you based in S.A.?
|
|
|
|
Post by Vanna on Aug 27, 2023 14:03:14 GMT
Yes
|
|
|
|
Post by Hutchyns on Sept 1, 2023 9:25:28 GMT
The algorithm genie seems to be well and truly out of the bottle, but the extent to which they are used, for good or for ill, will be the source of numerous controversies for many years to come. Just as we're currently discussing whether a lack of explicitly clarified blasphemy laws can be sustained while we create a rapidly expanding diverse society, so the 'benefits' of certain algorithms may have to be foregone as they are likely to come into conflict with 'anti prejudice' legislation that exists in many western countries. A conundrum the Dutch Police and other sectors are grappling/hamstrung with. link - Hateful Algorithms
 |
|
|
|
Post by Orac on Sept 1, 2023 9:33:59 GMT
In many cases such algorithms are quite complex and have an AI component.
It is likely that AI identified certain immutable factors as being linked to risk. Past attempts to resolve this 'bias' have included removing the immutable information. However, even when this is done, the AI often just uses a proxy for the missing information. What is of interest is whether the algorithm does its job in reducing the amount of resources needed to manage a complex task.
|
|
|
|
Post by oracle75 on Sept 1, 2023 17:14:03 GMT
Couple of thoughts...
If algorithms are essentially a synthesis of patterns of behaviour or opinion which can be used to project a successful result, and AI is the delivery of the results, what place does creativity have in creating the signposts?
For example would the work done by Paul Dirac have been ignored because it didnt conform to a pattern?
How does a system based on a pattern cope with something outside the pattern and which rejects "genius"? How reliable then is AI based on algorithms?
|
|
|
|
Post by Orac on Sept 1, 2023 18:18:52 GMT
Couple of thoughts... If algorithms are essentially a synthesis of patterns of behaviour or opinion which can be used to project a successful result, and AI is the delivery of the results, what place does creativity have in creating the signposts? For example would the work done by Paul Dirac have been ignored because it didnt conform to a pattern? How does a system based on a pattern cope with something outside the pattern and which rejects "genius"? How reliable then is AI based on algorithms? That's a tricky question. Here is a bit of of a brain teaser - AI can solve entirely novel problems by drawing comparisons with known problems and looking for similarities and relationships between the answer and the problem. How does it learn to do this? By looking at a lot examples of humans solving novel problems. Creepy huh? |
|
|
|
Post by sheepy on Sept 1, 2023 18:21:31 GMT
Couple of thoughts... If algorithms are essentially a synthesis of patterns of behaviour or opinion which can be used to project a successful result, and AI is the delivery of the results, what place does creativity have in creating the signposts? For example would the work done by Paul Dirac have been ignored because it didnt conform to a pattern? How does a system based on a pattern cope with something outside the pattern and which rejects "genius"? How reliable then is AI based on algorithms? That's a tricky question. Here is a bit of of a brain teaser - AI can solve entirely novel problems by drawing comparisons with known problems and looking for similarities and relationships between the answer and the problem. How does it learn to do this? By looking at a lot examples of humans solving novel problems. Creepy huh? Except it can do it in a millisecond. |
|
|
|
Post by oracle75 on Sept 2, 2023 8:19:56 GMT
Couple of thoughts... If algorithms are essentially a synthesis of patterns of behaviour or opinion which can be used to project a successful result, and AI is the delivery of the results, what place does creativity have in creating the signposts? For example would the work done by Paul Dirac have been ignored because it didnt conform to a pattern? How does a system based on a pattern cope with something outside the pattern and which rejects "genius"? How reliable then is AI based on algorithms? That's a tricky question. Here is a bit of of a brain teaser - AI can solve entirely novel problems by drawing comparisons with known problems and looking for similarities and relationships between the answer and the problem. How does it learn to do this? By looking at a lot examples of humans solving novel problems. Creepy huh? That then counts out the entirely unexpected novel piece of "outside the box" imaginative spark that leads us onto a new and perhaps better path. Example. A bicycle is composed of wheels, a seat, a chain, gears and handlebars. Every bicycle has these. So algorithms say these are necessary to be a bicycle. But over 100 years ago Picasso used the seat and placed handlebars on top and positioned them vertically and called it a bull. Would an algorithm have called it a bicycle? |
|
|
|
Post by Orac on Sept 2, 2023 8:42:22 GMT
That's a tricky question. Here is a bit of of a brain teaser - AI can solve entirely novel problems by drawing comparisons with known problems and looking for similarities and relationships between the answer and the problem. How does it learn to do this? By looking at a lot examples of humans solving novel problems. Creepy huh? That then counts out the entirely unexpected novel piece of "outside the box" imaginative spark that leads us onto a new and perhaps better path. Example. A bicycle is composed of wheels, a seat, a chain, gears and handlebars. Every bicycle has these. So algorithms say these are necessary to be a bicycle. But over 100 years ago Picasso used the seat and placed handlebars on top and positioned them vertically and called it a bull. Would an algorithm have called it a bicycle? I think the algorithm would be capable of noting that the re-arranged components of one thing bear a similarity to another thing. You can also tell it to do the task - to take bicycle components and re-arrange them to look like a bull. You could also set it to do this task a generality - to (say) show you lots of examples of the components of one thing being re-arranged to resemble another and present the results in the style of Rembrandt, Constable, Picasso or as (tech allowing) 'photographs' On the subject of creativity itself - here is my view. Human minds do not work at all in the same way as AI. However the world often allows more than one way to do something. I sympathise with the people saying that this is not real creativity because (in my view) it isn't. However, these people have a definition problem. Is the fact that A is different from B important when it makes no difference to the function? In my humble view we are in (or entering) a kind of emergency. People can start valuing our humanity itself and rejecting this impostor (humanity simulation) or the human race is going to become a short footnote to another story. |
|
|
|
Post by oracle75 on Sept 2, 2023 9:03:49 GMT
I take your point of AI responding to instruction, but Picasso had no such instruction.
He just saw something in a rearrangement of elements of something completely outside the result. Seeing a bull in specific elements of a bicycle has no logical connection.
Further i wonder how AI handles metaphors where the literal meaning of one thing is compared to something else which is the real meaning of the communication. And this can occur in single words or phrases.
Ie " it is raining cats and dogs".
AI might learn what that means but then encounters the phrase " it is raining stairods".
And then comes someone's new inventive metaphor.
I am beginning to see small groups of people living outside the mainstream, ignoring AI and relying solely on the natural capacity of the human brain. I suppose they could be the new neanderthals who eventually die out by giving way to more advanced technology.
It would be a shame to have one's life dominated by machine probibility. I still doubt if AI can ever account for the infinite variables life throws up.
|
|
|
|
Post by Orac on Sept 2, 2023 10:32:35 GMT
You are asking all the right questions and my ego feels i should have all the right answers - ie being a person with a lot of experience in technology. However, in all honesty, this subject cowers my ego entirely and this is not tech as it has been known
The AI receives no instruction, but it doesn't need to if it can predict (through sampling) what the instruction would likely be. This is a hard concept to get your head around but i feel it is important - if there is no aspect of the human that can't be 'adequately simulated' through sampling, then human instruction to the AI can also be simulated. I know this sounds like a cheat (and it is) but the whole thing is a cheat and this is the power of cheating. Imagine a world in which everything, no matter how complicated and badly understood, can be duplicated and re positioned into another context effortlessly. Reality as we know it no longer exists.
In another forum someone posted a nightmare they had, and it struck me as likely the most disturbing nightmare i have ever heard described.
In the dream, they were led (by someone) from their house to the garden. At the foot of their garden, there were a ring / crowd of people.
As the dreamer got closer she could see there were everyday human scenes being played out in the crowd -
A mother with her baby
A marriage - the groom kisses the bride
A family meal
Children circling each other, sword fighting with sticks
Two men arguing
people waiting for a bus
That sort of thing.
The 'guide' turned to the dreamer with a malicious smile and asks - "can you tell which is real? I bet you can't"
|
|
