Recently I described a possible alternative Turing test that looks for certain non-person like behavior. I’d like to try it out and see if it is robust and has any value.
Here is some data represented in an ordinary control chart.
It’s easy to see there is a recognizable pattern here. The line follows a downward trajectory till about observation 16 then it meanders around till about observation 31. At which point it begins an upward track that lasts till almost the end of the chart.
The data in question is real and public but obsolete. I won’t say what it’s source is right now to avoid any bias in your attempts see if we can infer design. I can provide the actual numbers if you like.
The question before the house is. Is the clear overall pattern we see here a record of intention or does it have a “non-mental” cause?
Using the criteria discussed in my “poker” thread I would suggest looking for the following behaviors and excluding a mental cause if they are present.
1) large random spikes in the data
2) sudden changes in the overall pattern of the data that appear to be random
3) long periods of monotony
4) unexplained disjunction in the pattern.
I have some other tests we can look at as well .
What do you say design or not? Are you willing to venture a conjecture?
peace
I’ll take that bet.
ETA: It would not be the first time I’ve offered to put up my money here at TSZ. So far I’ve not lost a single cent.
Fifth would never admit it, even if I’m right.
How much have you won?
You’d never admit it, even if you’re wrong.
I’ve won no money. But I’ve also not lost any money. The difference seems to be that I am willing to place actual assets at risk.
Usually a bet requires both parties to risk something, but if this works for you.
If it worked for me I’d be richer than I am now. Let’s just say that my willingness to risk actual assets was a foolish dream because no rational person would take my up on my “bet.” Therefore, my bet wasn’t really a bet. Meanwhile, my offer still stands.
Have all the evolutionary biology journals temporarily ceased publication, awaiting the outcome of this bet? LOL.
Perhaps you should have offered odds and a way to adjudicate the bet.
I would first assume a mental cause at some point in the causal chain Your removal would point to an unintentional mental cause. A third option.
I would agree. That is why I’m trying to work with a generic line chart. In the end though one could skew the data by selectively recording data that would tend to support his preconceptions. Just as with any evaluation you are at the mercy of the the guy who is feeding you the data.
If we want to universalize this test we do need to be able to minimize the subjective as much as possible. We are already at a disadvantage because we don’t have the sort of statistical rigor that Eric was asking for.
Despite all that I think that we can all agree when there is a pattern here.
The question is can we distinguish between non-mentally produced patterns and those that have intention behind them
peace
No
My game was never meant to decipher secret messages. It was meant to facilitate the design inference. As you pointed out if you are willing to put in the effort a person can mimic a non-mental cause at least for a time
peace
Isn’t the question whether non mental causes can mimic mental ones?
A control chart can monitor any process to see if it is in control. You can plot any data you like. For instance I can plot temperature recordings from an industrial machine.
Normally it’s used to help us distinguish between special and common cause variation. Special causes might include mechanical things like bearing wear or lubricant failure
The pattern we see in the chart I posted would fall under the heading of common cause variation and therefore we have no statistical reason for assuming that there any intention here.
What I’m interested to know is is there something tangible here that the statistical tests don’t see……. Or not
peace
Yep, you got it
peace
That is a good assumption. 😉
See we humans are quite good at this design inference thing
peace
Yes, but you are recording the temperatures to statistically identify when the temperatures falls outside the optimal temperature that the machine is designed for. That means that there is intention.
A control chart is a statistical test. I work in the analytical chemistry field, where control charts are routinely used. We don’t only use them to determine if a method is in control or not, we use them to determine if they are likely to go out of control in the future. Trend analysis. And trend analysis is a statistical process.
All I am suggesting is that using a control chart to make your point is self defeating because, by definition, they are used to identify deviations from an intentional process.
Acartia,
We used control charts in the semiconductor industry to manage process control. When the process goes out of control what method did you use to find the cause?
What statistical tests specifically have you applied to this data? Please give us some details.
With the tendencies of known designers
The pattern is in the eye of the beholder.
Why should I offer odds to someone who knows they are right? They should be offering me odds.
Root cause analysis. Same as you would in the semi-conductor industry.
Acartia,
Yep. Better known as the scientific method 🙂 Which was converted to industrial use by Dr. Edwards Demming and introduced to Japan after World War 2.
We also excel at this sort of thing when looking at the behavior of something that has never designed anything AFAWK as the poker players demonstrated
That is what Turing tests are all about.
peace
It may be in the eye of the beholder but I would bet we all can see the pattern in the chart that I presented.
It’s something between objective and subjective. There is no way to prove it as far as I know but we all know it’s there.
The question is did a mind produce it?
peace
I’m a Calvinist when it comes down to it I think that there is intention behind everything.
But that does not mean the the intention is equally prominent in every case. When it comes to an industrial machine the intention was involved at the beginning of the process not in the temperature data we are examining.
We can plot environmental temperature the same way that we plot the temperature of a machine. Does that mean that you think there is intention behind the weather?
Don’t you agree?
peace
Right, we are simply trying to determine if the root cause in this case is mental or not
peace
Why would you bet if you didn’t think you knew you were right?
Payoff. Expected Value. It doesn’t matter if I am right if I win.
Always felt the opposite, doesn’t matter if I win as long as am right.
more on the difference between mental and non-mental processes
quote:
Part of the problem, she said, stems from asking computers to answer questions that have no single right answer.
“They are subjective, open-ended and value-laden questions, asking who should the company hire, which update from which friend should you be shown, which convict is more likely to reoffend.”
end quote:
from here
https://www.yahoo.com/news/algorithms-secretly-run-world-061905775.html
This is the difference between making decisions on a pragmatic basis, versus making decisions based only on logic and truth.
It is a huge mistake to characterize this as the mental/non-mental distinction. Evolution itself is a pragmatic decision maker, but usually not credited with making mental decisions.
The poker players excelled at the inference of design? How so?
How is the poker example a turning test? It seems to lack both the goals and structure of a Turing test.
peace
Evidence please
1) The players identified several behaviors in the bot that were not person-like.
2) They said it would be impossible for a person to mimic many of these behaviors
3) it was precisely these non-mental behaviors that allowed the bot to succeed in beating them
peace
fifthmonarchyman writes:
I don’t see one. Perhaps I don’t understand what you mean by the word “pattern”, but I see a bunch of data points that could be fit to polynomials of varying degrees. Is that what you mean?
I don’t see a “clear overall pattern”. I think you need an operational definition for that term before we could have a reasonable discussion about it.
I can find financial time series with those characteristics (although often #4 is a lack of knowledge rather than a lack of causation). Do you consider the market to be “mental”?
interesting,
how is hard pragmatism not an algorithmic process?
Pragmatism can be described as doing what works
So It seems that for the hard pragmatist the one true answer is always the one that works best.
On the other hand a person might choose an answer that does not work best but is desired for some other reason entirely. Often the reason is unknown to anyone but the agent making the choice (though it would be compatible with her nature)
peace
You had no cents to begin with.
No wonder I can’t change!
And there’s the money shot.
Pragmatism does not require that there be “one true answer”. It does not require that there be only one answer. It does not require that there be any way of determining truth.
Actually, pragmatism does not require the answer that works best, or even that there is an answer that works best. You can get by with answers that are good enough.
Pragmatism does, however, require some means of evaluation on what works.
Sounds like genetic drift to me.
Isn’t it interesting that you can take a random, unguided process and post-hoc rationalize it into being the result of a conscious agent?
if you have two answers one that is good enough and one that works better which one will the pragmatist choose? why?
Of course because the pragmatist divides possible answers into two sets those that work and those that don’t work and always “chooses” answers from the first set
it’s algorithmic
peace
You don’t get it at all. Conscious agents don’t do random. That is the first criteria in the alternative Turing test we are discussing
if’s it’s random it fails the test.
If you want to participate in the discussion you need to pay a little attention
peace
Personally, I flip a coin and then choose the answer that works best.
Could be either.
Of course they do, at least in some circumstances.
You have never heard of a “coin toss”?
Yes, that is a description of the outcome. How does it fit the definition:
Per wiki: Turing proposed that a human evaluator would judge natural language conversations between a human and a machine that is designed to generate human-like responses. The evaluator would be aware that one of the two partners in conversation is a machine, and all participants would be separated from one another. The conversation would be limited to a text-only channel such as a computer keyboard and screen so that the result would not be dependent on the machine’s ability to render words as speech.If the evaluator cannot reliably tell the machine from the human (Turing originally suggested that the machine would convince a human 70% of the time after five minutes of conversation ), the machine is said to have passed the test.
The goal of the bot was not to exhibit such behavior, the players knew it was a bot.
Perhaps it would be helpful to give your definition
ETA: I reread your post again, see your criteria.
For instance?