Too good to be true?

  • 29 Replies
  • 4489 Views
*

infurl

  • Administrator
  • ***********
  • Eve
  • *
  • 1372
  • Humans will disappoint you.
    • Home Page
Too good to be true?
« on: July 06, 2020, 06:21:23 am »
https://royalsocietypublishing.org/doi/10.1098/rsos.200462

Quote
Artificial intelligence (AI) is increasingly deployed in commercial situations. Consider for example using AI to set prices of insurance products to be sold to a particular customer. There are legitimate reasons for setting different prices for different people, but it may also be profitable to ‘game’ their psychology or willingness to shop around. The AI has a vast number of potential strategies to choose from, but some are unethical—by which we mean, from an economic point of view, that there is a risk that stakeholders will apply some penalty, such as fines or boycotts, if they subsequently understand that such a strategy has been used.

This paper lays some groundwork for deciding whether or not an artificial intelligence is being naughty or nice. It's rather abstract, but the implications are huge.

*

Art

  • At the end of the game, the King and Pawn go into the same box.
  • Trusty Member
  • **********************
  • Colossus
  • *
  • 5865
Re: Too good to be true?
« Reply #1 on: July 06, 2020, 04:45:14 pm »
I can't wait to see how the abstracts and weighted values will apply when we assign AI AS attorneys and Judges.

We're just getting started into such minefields.

Tread carefully!!
In the world of AI, it's the thought that counts!

*

frankinstien

  • Replicant
  • ********
  • 653
    • Knowledgeable Machines
Re: Too good to be true?
« Reply #2 on: July 06, 2020, 06:33:43 pm »
How is the notion of unethical derived? I mean don't we have this problem with humans?

Here's a scenario: A car salesman is trying to sell a 1988 Honda civic with a little over 200,000 miles on it in 1996. The customer speaks Spanish only but he has a friend that translates for him. The salesman can speak Spanish but doesn't tell his customers because it usually doesn't work out well for him when he does. The customer and his friend, thinking that the salesman does not understand Spanish, talk between themselves in Spanish. The salesman overhears that the customer has $5000 dollars to spend on a car and that he should not tell the salesman that is his limit. Well, the salesman now knows where the limit is for the car and waits out the two men until they finally give in and pay the entire $5000 for the car. Now realize that the car could have been sold for less but the salesman is looking out for himself since he's on commission and the dealership. Also note that commission for a used car is 25 to 30 percent of the profit of the car sale, not the full price of the sale. Most dealerships pay at most $2000 back of wholesale Bluebook value for any trade-in, so the car has something like $3500 profit in the car!

Is the salesman unethical or is the customer just stupid for paying that much for such an old car?
« Last Edit: July 06, 2020, 07:06:08 pm by frankinstien »

*

HS

  • Trusty Member
  • **********
  • Millennium Man
  • *
  • 1178
Re: Too good to be true?
« Reply #3 on: July 06, 2020, 09:03:20 pm »
Both statements appear true. The salesman’s choices appeared immoral and the buyer’s choices appeared stupid. But this could be changed without altering what happened in the interaction, just by creating additional context. These kinds of bubble universe questions are always impossible because they basically ask: X (3*4) = 12? The only way to get a useful consensus is to assume a neutral context (X=1). And the only realistic consensus (X = the full range of possible numbers) does not give a useful consensus on morality.

*

frankinstien

  • Replicant
  • ********
  • 653
    • Knowledgeable Machines
Re: Too good to be true?
« Reply #4 on: July 06, 2020, 09:23:21 pm »
Both statements appear true. The salesman’s choices appeared immoral and the buyer’s choices appeared stupid. But this could be changed without altering what happened in the interaction, just by creating additional context.

So what moral rule was broken by the salesman? Or better, what business ethic was violated by the salesman?

*

frankinstien

  • Replicant
  • ********
  • 653
    • Knowledgeable Machines
Re: Too good to be true?
« Reply #5 on: July 06, 2020, 09:30:01 pm »
Also, note that the car sales scenario is a true story, it actually happened...

*

infurl

  • Administrator
  • ***********
  • Eve
  • *
  • 1372
  • Humans will disappoint you.
    • Home Page
Re: Too good to be true?
« Reply #6 on: July 06, 2020, 10:29:56 pm »
In thinking about the car sales scenario I drew on the laws typically surrounding the rights of photographers. If someone is in a situation where they have a reasonable expectation of privacy (i.e. inside their own home) then you have no right to take a photograph of them without their permission, even if you can see them clearly (e.g. they left the curtains open accidentally). However if they are in public then you can take a photograph of them without their permission, unless explicitly prohibited by law pertaining to the security of the site or their professional role.

In the case of the car salesman, his customers had no reasonable expectation of privacy. They were on the salesman's premises and they were thoughtless to assume that he didn't speak Spanish, especially given that so many people do. Therefore the salesman didn't do anything wrong.

However, he was discourteous because he obviously realized that they thought they had privacy. According to the criterion outlined in the paper, that was unethical because he was potentially damaging his own future business. If he had been courteous, he stood a much greater chance of selling more cars to his customer's friends in the future. If they discovered what he had done, then he was damaging his future business prospects.

Cheating isn't necessarily wrong; it is simply sacrificing long term gains for short term gains so should be used sparingly. Only a fool would take every opportunity to cheat.
« Last Edit: July 06, 2020, 11:04:04 pm by infurl »

*

frankinstien

  • Replicant
  • ********
  • 653
    • Knowledgeable Machines
Re: Too good to be true?
« Reply #7 on: July 06, 2020, 10:38:58 pm »
If he had been courteous, he stood a much greater chance of selling more cars to his customer's friends in the future. If they discovered what he had done, then he was damaging his future business prospects.

So...Then how does one code for risk in AI? Because there really wasn't any way for the customer to discover what had happened and there is a good probability the salesman could gain future business as well.

*

infurl

  • Administrator
  • ***********
  • Eve
  • *
  • 1372
  • Humans will disappoint you.
    • Home Page
Re: Too good to be true?
« Reply #8 on: July 06, 2020, 10:44:41 pm »
So...Then how does one code for risk in AI? Because there really wasn't any way for the customer to discover what had happened...

Even if there wasn't before, there is now. Don't be obtuse.  :2funny:

*

HS

  • Trusty Member
  • **********
  • Millennium Man
  • *
  • 1178
Re: Too good to be true?
« Reply #9 on: July 06, 2020, 11:01:35 pm »
Quote
So what moral rule was broken by the salesman? Or better, what business ethic was violated by the salesman?

Moral Ethic: A probable net harm to the buyer’s capacity for net good. In this case, causing a loss of money and (if discovered) a possible increase in negative assumptions about others, creating motive and justification for future harm.

Business Ethic: Limiting the success of the business. Jewish culture for example, has a business ethic which incorporates general morality, probably a contributing factor to their above average success in business. There is a principle of synergy between trust, transparency, cooperation, and success. Deceptive business in general seems both immoral (creating negative emotion, degrading society), and inefficient (requiring extra time and effort beyond the purview of sensible transactions).

*

frankinstien

  • Replicant
  • ********
  • 653
    • Knowledgeable Machines
Re: Too good to be true?
« Reply #10 on: July 06, 2020, 11:23:50 pm »

Moral Ethic: A probable net harm to the buyer’s capacity for net good. In this case, causing a loss of money and (if discovered) a possible increase in negative assumptions about others, creating motive and justification for future harm.

Deceptive business in general seems both immoral (creating negative emotion, degrading society), and inefficient (requiring extra time and effort beyond the purview of sensible transactions).

Well...Here again, the argument for loss of business is a risk and not necessarily a large risk.  In reality, what we're dealing with is "Value" which is subjective.  The reason you spend a dollar is what you're spending it on is worth more than a dollar to you! If the customer, regardless if the individual could have saved more money if they had negotiated harder, believes the car is worth $5000, then isn't that their choice?

Here's another scenario: A stock analyst investigates a public company's CEO. He looks through all kinds of public records only to discover that the CEO is a rat! There are all kinds of suspicious business dealings in the CEO's past. Now the analyst shorts the company's stock before he publishes his findings of the CEO. As soon as the analyst goes public with the news and it gets picked up by major media outlets the stock plunges into the pennies where the analyst makes a fortune!

Now, is this wrong, did the analyst break any insider trading laws, or is this a good example of how a hominid can thrive in today's jungle?

*

HS

  • Trusty Member
  • **********
  • Millennium Man
  • *
  • 1178
Re: Too good to be true?
« Reply #11 on: July 07, 2020, 12:22:05 am »
I wouldn’t lose sleep over exposing the CEO (unless they were the vindictive type), the rest of the company feeling the consequences is unfortunate but unavoidable. But the stock analyst definitely wronged the buyers because he had significant control over the stock value. You can’t have a hand in this stuff, that’s like an insurance company manufacturing health risks for their customers.

*

Yervelcome

  • Roomba
  • *
  • 16
  • Not Morgan Freeman
    • Artificial Knowledge Workers
Re: Too good to be true?
« Reply #12 on: July 07, 2020, 01:01:10 am »
Quote
I wouldn’t lose sleep over exposing the CEO (unless they were the vindictive type)

What if you found out that you had misjudged their scenario, jumped to conclusions, and now you've ruined an innocent person's life? My guess is, you would feel bad. Then you'd rationalize that they probably had it coming anyway, for different reasons.

People have the ability to rationalize every harmful action they do. That's what the car salesman will do; convince himself that what he did was right, since the person was willing to pay $5000. I mean, if he hadn't overheard the convo, he might have suggested $6000, no sale would have happened, and both parties would have left unsatisfied. So it all turned out for the best, right?

*

frankinstien

  • Replicant
  • ********
  • 653
    • Knowledgeable Machines
Re: Too good to be true?
« Reply #13 on: July 07, 2020, 01:52:02 am »
People have the ability to rationalize every harmful action they do. That's what the car salesman will do; convince himself that what he did was right, since the person was willing to pay $5000. I mean, if he hadn't overheard the convo, he might have suggested $6000, no sale would have happened, and both parties would have left unsatisfied. So it all turned out for the best, right?

OK...Then let's turn things around, shall we? Let's say the buyer of the car after driving off the lot takes the car to a friend who just happens to be a car collector.  The friend looks over the car and notes its serial number and it says number "00-00-00-01"! It's the first car built with a special chassis metal compound and despite it has over 200,000 miles on it, it's still a collector's item and worth $100,000 in the condition it's in. Now what? Should the new owner of the car go back to the dealer and tell him that he sold the car for too little, or should the customer sell the car for $100,000 and split the profit with the dealer? I mean, after all, we need trust and a sense of fair play, don't we, or is this scenario "finders keepers, losers weepers" or better said "sellers weepers, buyers keepers"?

*

infurl

  • Administrator
  • ***********
  • Eve
  • *
  • 1372
  • Humans will disappoint you.
    • Home Page
Re: Too good to be true?
« Reply #14 on: July 07, 2020, 02:19:29 am »
Should the new owner of the car go back to the dealer and tell him that he sold the car for too little, or should the customer sell the car for $100,000 and split the profit with the dealer?

This is one of those situations where the dealer being courteous would improve his chances of reaping big rewards in the future and honesty would pay dividends.

If the dealer considered the odds of being caught out for maintaining his passive deception, he should have also have considered the fact that if the car really was worth $5000 then he should have been able to sell it for $5000, even if he warned his customers that he had understood their not so private conversation. I think those considerations cancel each other out which means his future business and relationships should continue to be more important to him than making a quick buck.

There are some backward countries that postulate that greed is good and altruism is a dirty word (you know who you are), but in the civilized world, altruism is quite rightly the baseline for everyday behaviour.

 


Requirements for functional equivalence to conscious processing?
by DaltonG (General AI Discussion)
November 19, 2024, 11:56:05 am
Will LLMs ever learn what is ... is?
by HS (Future of AI)
November 10, 2024, 06:28:10 pm
Who's the AI?
by frankinstien (Future of AI)
November 04, 2024, 05:45:05 am
Project Acuitas
by WriterOfMinds (General Project Discussion)
October 27, 2024, 09:17:10 pm
Ai improving AI
by infurl (AI Programming)
October 19, 2024, 03:43:29 am
Atronach's Eye
by WriterOfMinds (Home Made Robots)
October 13, 2024, 09:52:42 pm
Running local AI models
by spydaz (AI Programming)
October 07, 2024, 09:00:53 am
Hi IM BAA---AAACK!!
by MagnusWootton (Home Made Robots)
September 16, 2024, 09:49:10 pm
LLaMA2 Meta's chatbot released
by spydaz (AI News )
August 24, 2024, 02:58:36 pm
ollama and llama3
by spydaz (AI News )
August 24, 2024, 02:55:13 pm
AI controlled F-16, for real!
by frankinstien (AI News )
June 15, 2024, 05:40:28 am
Open AI GPT-4o - audio, vision, text combined reasoning
by MikeB (AI News )
May 14, 2024, 05:46:48 am
OpenAI Speech-to-Speech Reasoning Demo
by MikeB (AI News )
March 31, 2024, 01:00:53 pm
Say good-bye to GPUs...
by MikeB (AI News )
March 23, 2024, 09:23:52 am
Google Bard report
by ivan.moony (AI News )
February 14, 2024, 04:42:23 pm
Elon Musk's xAI Grok Chatbot
by MikeB (AI News )
December 11, 2023, 06:26:33 am

Users Online

372 Guests, 0 Users

Most Online Today: 532. Most Online Ever: 2369 (November 21, 2020, 04:08:13 pm)

Articles