Preventing artificial intelligence from taking on negative human traits.

J_B_

I have answers to questions no one ever asks.
May 15, 2020
1,258
365
Midwest
✟101,755.00
Country
United States
Faith
Christian
Marital Status
Private
I think it depends on the 'depth' of the AI - for example, Alpha Go came up with a move, called 'miraculous' and 'sublime' by the world's top players (some hyperbole, perhaps), that no-one had envisaged, that no human player would have played ...

We can't know that no human would have ever conceived of those moves. And people are impressed by beautiful rock formations they never would have conceived, calling them sublime. This conversation is somewhat humorous. As one person alluded, the question of responsibility for AI actions very closely resembles arguments over God & the problem of evil. This particular line of the conversation resembles arguments about intelligent design.

Before fully answering what you've said, I need some clarification on what you think intelligence is? In short, if an entity is not free to think its thoughts, is that intelligence? What the AI is doing is just a more sophisticated version of an optimizer searching a nonlinear space. It's not having independent thoughts. It's searching that space according to the rules forced upon it. It does not invent the space, nor can it ever decide to leave the space. It's just searching the space.
 
Upvote 0

durangodawood

Dis Member
Aug 28, 2007
23,411
15,558
Colorado
✟427,906.00
Country
United States
Faith
Seeker
Marital Status
Single
Yeah there's 2 different notions of "intelligence" out there, and the conflation of them is really bogging down discussion.

1. a capacity to perform intellectual tasks. (a calculator has a rudimentary intelligence here)

2. An understanding consciousness. C3PO is written as an intelligence. (no real signs of this sort of AI yet, afaik.)
 
  • Agree
Reactions: J_B_
Upvote 0

FrumiousBandersnatch

Well-Known Member
Mar 20, 2009
15,258
8,056
✟326,229.00
Faith
Atheist
I don't think it's a matter of computer beings (A.I.'s) being "intelligent", but rather are they really truly fully "conscious", do they act with their own "will", are they really "alive", etc...?
They're not conscious in the sense that cognitively advanced animals are; they don't have either the complexity or the functional organization for it and they don't need it.
 
Upvote 0

sjastro

Newbie
May 14, 2014
4,852
3,887
✟273,723.00
Faith
Christian
Marital Status
Single
Grandmasters can execute short term tactical chess moves which AlphaZero would also make.
Tactical chess moves on average require grandmasters to think up to 4-5 moves ahead (8-10 ply).
Does it mean grandmasters can ever hope to play at AlphaZero’s level?
The answer is no as strategic chess on the other hand which becomes more significant as the standard of play increases involves long term planning and decision making well beyond the depth of tactical chess.
AlphaZero’s vastly superior strategic chess knowledge over grandmasters allows it evaluate at depths way beyond the capacity of the human brain.
What can be stated is grandmasters can now conceive some of these long term strategic moves by learning from AlphaZero.

Since ‘intelligence’ has come into the discussion a chess player’s ELO rating is a measurement of 'chess IQ'.

chessA.gif

The very best human grandmaster ELO ratings are over 2800.

1*MTioUb6v19q4m-4wkySBMg.jpeg

The graph is somewhat out of date as the best chess engines including Leela Chess Zero are now around 3500.

ELO rating differences provide a probability of the winning percent rate.
The difference between the best grandmasters and chess programs is around 700.
This means a chess program has around a 98% chance of beating the human world champion; in other words a no contest.
 
Last edited:
Upvote 0

J_B_

I have answers to questions no one ever asks.
May 15, 2020
1,258
365
Midwest
✟101,755.00
Country
United States
Faith
Christian
Marital Status
Private
This means the a chess program has around a 98% chance of beating the human world champion; in other words a no contest.

Yep. Just like an automobile can beat a human in a head-to-head race every time - just not because the automobile is "intelligent".
 
Upvote 0

J_B_

I have answers to questions no one ever asks.
May 15, 2020
1,258
365
Midwest
✟101,755.00
Country
United States
Faith
Christian
Marital Status
Private
The humans werent so upset when machines finally overtook them in ditch digging and hay raking and so on. Why now alla sudden?

You've never heard of John Henry?

... which leads to a digression. In my history program I studied a lot of African American history. I proposed a hypothesis to my professor that institutional slavery didn't really end because of some moral victory, but because of the success of the Industrial Revolution. He didn't agree, but I still think it would be a great thesis.
 
Upvote 0

FrumiousBandersnatch

Well-Known Member
Mar 20, 2009
15,258
8,056
✟326,229.00
Faith
Atheist
We can't know that no human would have ever conceived of those moves.
The point is that the AI came up with a strategy that was totally novel and left-field - according to the world's top players. That doesn't mean someone wouldn't have come up with it eventually, but, more likely, that the AI's strategic developments were not constrained by the conventional wisdom of human strategies. Top human players do not learn strategy from scratch, but rely to a considerable extent on the accumulated knowledge of previous greats - who, in turn, learned from earlier masters.

Before fully answering what you've said, I need some clarification on what you think intelligence is? In short, if an entity is not free to think its thoughts, is that intelligence?
I consider intelligence to be a measure of the ability to efficiently solve problems. There is domain-specific intelligence and general intelligence.

I don't know what you mean by 'free to think its thoughts'. If by 'thoughts' you mean information processing, all information processors are constrained in some way or other; e.g. by their processing resources and available data.

What the AI is doing is just a more sophisticated version of an optimizer searching a nonlinear space. It's not having independent thoughts. It's searching that space according to the rules forced upon it. It does not invent the space, nor can it ever decide to leave the space. It's just searching the space.
In a sense that's true of all information processors - in some the space is larger and more inclusive than in others.

AI's like Alpha Go are relatively simple domain-specific learning systems with more conventional modules associated. The more sophisticated biological brains are (roughly) complex networks of domain-specific and general-purpose learning systems with various processing 'accessories'.

I'm curious to know what you mean by 'independent thoughts'.
 
  • Agree
Reactions: sjastro
Upvote 0
This site stays free and accessible to all because of donations from people like you.
Consider making a one-time or monthly donation. We appreciate your support!
- Dan Doughty and Team Christian Forums

FrumiousBandersnatch

Well-Known Member
Mar 20, 2009
15,258
8,056
✟326,229.00
Faith
Atheist
Yeah there's 2 different notions of "intelligence" out there, and the conflation of them is really bogging down discussion.

1. a capacity to perform intellectual tasks. (a calculator has a rudimentary intelligence here)

2. An understanding consciousness. C3PO is written as an intelligence. (no real signs of this sort of AI yet, afaik.)
Yes; Sentience and awareness are another two terms in this field that have very different meanings in different contexts. Then there are the various metaphorical uses...
 
Upvote 0

J_B_

I have answers to questions no one ever asks.
May 15, 2020
1,258
365
Midwest
✟101,755.00
Country
United States
Faith
Christian
Marital Status
Private
The point is that the AI came up with a strategy that was totally novel and left-field - according to the world's top players.

Sure. AI came up with a new strategy. I've never disputed that, and though that may be your point, it didn't seem the point of anyone else, but rather the focus on the "novel" part - an attempt to imply a human would never come up with that idea, thereby proving AI is actually intelligent. It's that part I don't buy.

Top human players do not learn strategy from scratch, but rely to a considerable extent on the accumulated knowledge of previous greats - who, in turn, learned from earlier masters.

Now we're into Thomas Kuhn space and Scientific Revolutions. So, again, sure. Paradigms can cause some very real myopia. But Holy Computers, Batman ... Archimedes, Aristotle, DesCartes, Leibniz, Newton, Einstein (as the list goes on) - them dudes had some VERY novel ideas.

I consider intelligence to be a measure of the ability to efficiently solve problems. There is domain-specific intelligence and general intelligence.

Thanks for the definition. A minimalist you are, then.

I'm curious to know what you mean by 'independent thoughts'.

As @durangodawood said earlier, the ability to step outside the game. A computer that would look at flat space and think, "Hmm, what about curved space? What about time as a dimension that results in spacetime?" In the case of Chess, a computer that would try to cheat - intimidate, fatigue, distract its opponent.

Let's be honest. No doubt women are intellectually equal in an ability to play chess, but if I were a single guy with Anya Taylor-Joy sitting across from me, it would be darned difficult to concentrate.
 
Upvote 0

J_B_

I have answers to questions no one ever asks.
May 15, 2020
1,258
365
Midwest
✟101,755.00
Country
United States
Faith
Christian
Marital Status
Private
I'm surprised you missed the point.
Who or what is performing the decision making when driving, the programmer or the AI?

It's just a different example with the same conclusions. I was being glib.

Depends on your definition of "decision". The word can be defined such that the answer is either.
 
Upvote 0

sjastro

Newbie
May 14, 2014
4,852
3,887
✟273,723.00
Faith
Christian
Marital Status
Single
It's just a different example with the same conclusions. I was being glib.

Depends on your definition of "decision". The word can be defined such that the answer is either.
response1.jpg
Not relevant.
Who or what is performing the decision making when driving, the programmer or the AI?
 
Upvote 0
This site stays free and accessible to all because of donations from people like you.
Consider making a one-time or monthly donation. We appreciate your support!
- Dan Doughty and Team Christian Forums

SelfSim

A non "-ist"
Jun 23, 2014
6,154
1,953
✟174,600.00
Faith
Humanist
Marital Status
Private
On a fixed obstacle course in the absence of other vehicles that would possible even without using AI.
Explain how a programmer can anticipate the random conditions a driverless car will encounter in city driving.
Surely its more a question of reduced reaction times to a given set of potentially hazardous events(?) more than random conditions?
 
Upvote 0

sjastro

Newbie
May 14, 2014
4,852
3,887
✟273,723.00
Faith
Christian
Marital Status
Single
Surely its more a question of reduced reaction times to a given set of potentially hazardous events(?) more than random conditions?
That is part of the decision making AI has to make.
A driverless car is also expected to safely overtake, brake at a safe distance and know how to park which depends on the position of other vehicles and is random.
 
Upvote 0
This site stays free and accessible to all because of donations from people like you.
Consider making a one-time or monthly donation. We appreciate your support!
- Dan Doughty and Team Christian Forums

J_B_

I have answers to questions no one ever asks.
May 15, 2020
1,258
365
Midwest
✟101,755.00
Country
United States
Faith
Christian
Marital Status
Private
On a fixed obstacle course in the absence of other vehicles that would possible even without using AI.
Explain how a programmer can anticipate the random conditions a driverless car will encounter in city driving.

You accuse me of obfuscation. I will simply note I am in good company. What I repeatedly ask you to do, and what you repeatedly refuse to do, is define your terms. For someone who claims to be a physicist, it perplexes me that you won't do it. Regardless, in this video Feynman is asked to answer whether machines are intelligent. Almost the first comment he makes is that in order to answer, intelligence must be defined. (see 0:28)


You deferred on that point. So, I took the liberty of doing it myself. Since, however, you declined to share, I likewise decline to share how I came to that conclusion ... though if you read my earlier replies, you'll see that actually I've already answered the question.

Good day, sir.

Surely its more a question of reduced reaction times to a given set of potentially hazardous events(?) more than random conditions?

True.
 
Upvote 0