Chaos Magic

[Squeegee Beckenheim]

Oh, and are you going to produce the e-mails from Professor Rees who, for some reason, didn't ask you not to?

Yes please. You said you would, and have been asked by several people to do so. I thought you were interested in showing everybody who is the liar?

 

[dlorde]

Yup, still waiting here for a pm or posting of the emails annnnoid is happy to let us see...

 

[annnnoid]

Why would Rees ask me not to publish anything? The list of unanswered questions was taken directly from a paper he himself published. It’s readily available for anyone to read.

http://www.cogsci.ucsd.edu/igert/bo...tureNeuroscience2006-multivariateAnalysis.pdf

Since the paper was published a few years ago I gave him the list (of unanswered questions) and simply asked if the questions were still unanswered. This was the list:

-How accurately and efficiently a mental state can be inferred?
-Whether a person’s compliance is required?
-Whether it is possible to decode concealed thoughts or even unconscious mental states?
-The maximum temporal resolution?
-The degree to which it is possible to provide a quasi-online estimate of an individual’s current cognitive or perceptual state?
-The problem of inverse referencing?
-Whether decoding methods are sensitive enough to reliably reveal personal information for individual subjects?
-Brain reading will be restricted to simple cases with a fixed number of alternatives...for all of which training date are available....because of the all but infinite number of cognitive states and necessarily limited training categories.

This was his reply:

I think everything remains more or less at the point where we wrote the article, so the questions are more or less the same. Of course, this is to be expected as science makes progress reasonably slowly. There have been studies published in the meanwhile (including my own) that improve accuracy and generalisability of predictive approaches to decoding hidden mental states, and some early work with MEG that dramatically improves the temporal resolution. But the overall picture remains very similar.

“…the overall picture remains very similar…”

I figured it might be helpful if things were clarified a little more, so I created a few more questions on my own and sent him a second email.

A simple hypothetical scenario. I walk in off the street and am ‘plugged into an fMRI. The fMRI operates according to the level of interpretative programming currently available (no pre-screening, calibration, alignment, testing, etc. etc. etc.)

From the fMRI results can you:
1 Provide an explicit description of what I am seeing?
2 Provide a general description of what I am seeing?
3 Provide an explicit description of what I am hearing?
4 Provide a general description of what I am hearing?
5 Provide an explicit description of what I am smelling?
6 Provide a general description of what I am smelling?
7 Provide an explicit description of what I am touching?
8 Provide a general description of what I am touching?
9 Provide an explicit description of what I am imagining (eyes closed)?
10 Provide a general description of what I am imagining (eyes closed)?
11 Provide an explicit description of what I am thinking?
12 Provide a general description of what I am thinking?
13 Provide an explicit description of how I am feeling?
14 Provide a general description of how I am feeling?

This was his reply:

The answer to all 14 questions in the situation you pose is a very clear 'no'. I am one hundred percent confident that no-one doing this sort of science anywhere in the world would disagree with me. If you are asking these kinds of questions, then I am concerned you may be radically mis-interpreting the scientific literature. While some people make slightly over-zealous claims, I have never in my entire scientific career ever read any claim that any of your 14 questions could be answered 'yes'.

Essentially, Rees position would appear to be reflected by his statement: “brain reading will be restricted to simple cases with a fixed number of alternatives...for all of which training date are available'....because of the all but infinite number of cognitive states and necessarily limited training categories.”

By the way Argent, this is what you had to say about Rees statements before you emailed him:

Rees: How accurately and efficiently a mental state can be inferred is unknown.
Argent: No, it's known quite precisely. We know what the limits of neural scanning are.

Rees: Whether a person’s compliance is required is unknown.
Argent: No, it’s known.

Rees: Whether it is possible to decode concealed thoughts or even unconscious mental states is unknown.
Argent; No. We know it's possible.

Rees: The maximum temporal resolution is unknown.
Argent: The what now? (apparently this is something you don’t understand)

Rees: The degree to which it is possible to provide a quasi-online estimate of an individual’s current cognitive or perceptual state is unknown.
Argent: No, it’s entirely known.

Rees: The problem of inverse referencing remains, essentially, unresolved.
Argent: Don’t understand that one either.

Rees: Whether decoding methods are sensitive enough to reliably reveal personal information for individual subjects in unknown.
Argent: No, it's known, and the answer is "yes, it can, with about seventy-five percent accuracy in people trying to fool the machine and up to ninety-five percent in people who are cooperative".

Rees:....and outright ‘no’ to every one of the fourteen points.
Argent: Wrong. Do your research.

Once again…you make a whole bunch of massive claims…without a shred of substantiation. No-evidence-Argent.

…but then Argent decides it’s time to go look for his own evidence. This is a first Argent. You…actually presenting evidence. No ‘google it’?

Annnnoid’s list is, in some areas, generally correct.

…now isn’t that interesting. Why don’t the members of our audience have a look above and note just how many times Argent flatly rejected the points made on that very same list! Argent either generally or completely dismissed every single statement Rees made. But now, for some mysterious reason, the list is generally correct. Which areas Argent? Do we have to change some of your no’s to yes’s? Which ones?

I've only objected to annnnoid's claims that we cannot detect the things he has listed.

No Argent, you flat out insisted that:

Everything you listed can be and has been detected by a myriad of scientific instruments.

…a list that (as I clearly stated when I wrote it) goes on forever and includes every variety of human experience (there was another clearly stated qualification but you’ve ignored everything else so I’m not at all surprised you ignored that as well). But now, once again, Argent’s position changes. Only in ‘specific cases’ can neural scanning read minds.

So before it was…’EVERYTHING’….but now it’s …’only in specific cases’.

Care to specify which ones Argent. I could give you a list of few million, but why don’t we restrict ourselves to just the words on that list. Which ones can be detected? Oh yeah…google it.

…but this is what you and Pixy have been saying all along. Except…all along, neither one of you has produced a shred of evidence to support your position. Oh yeah, I forgot, ‘google it’.

annnnoid is wrong in saying that emotions and visual data cannot be decoded.

Please provide a specific link where I have claimed that emotions and visual data cannot be decoded. I never once said this. The only position I have ever taken, and argued extensively, is that there are substantial limits to what can be decoded. This was explicitly confirmed and supported by Rees: “Brain reading will be restricted to simple cases with a fixed number of alternatives...for all of which training date are available....because of the all but infinite number of cognitive states and necessarily limited training categories.”

So we go from:

Everything you listed can be and has been detected by a myriad of scientific instruments.

To: “Brain reading will be restricted to simple cases with a fixed number of alternatives...for all of which training date are available....because of the all but infinite number of cognitive states and necessarily limited training categories.”

 

[Nonpareil]

Why would Rees ask me not to publish anything? The list of unanswered questions was taken directly from a paper he himself published. It’s readily available for anyone to read.

I was talking about the e-mails you claimed that you would produce. In his e-mails to me, he said that he did not want anything he sent reproduced on a forum.

-How accurately and efficiently a mental state can be inferred?
-Whether a person’s compliance is required?
-Whether it is possible to decode concealed thoughts or even unconscious mental states?
-The maximum temporal resolution?
-The degree to which it is possible to provide a quasi-online estimate of an individual’s current cognitive or perceptual state?
-The problem of inverse referencing?
-Whether decoding methods are sensitive enough to reliably reveal personal information for individual subjects?
-Brain reading will be restricted to simple cases with a fixed number of alternatives...for all of which training date are available....because of the all but infinite number of cognitive states and necessarily limited training categories.

This was his reply:

I think everything remains more or less at the point where we wrote the article, so the questions are more or less the same. Of course, this is to be expected as science makes progress reasonably slowly. There have been studies published in the meanwhile (including my own) that improve accuracy and generalisability of predictive approaches to decoding hidden mental states, and some early work with MEG that dramatically improves the temporal resolution. But the overall picture remains very similar.

“…the overall picture remains very similar…”

Hence what he said to me: you are correct in a general sense. There are many areas of neurology where these questions do remain unanswered. However, in the context of reading emotions and constructing visual images - which is what we've been talking about - these questions can be answered.

A simple hypothetical scenario. I walk in off the street and am ‘plugged into an fMRI. The fMRI operates according to the level of interpretative programming currently available (no pre-screening, calibration, alignment, testing, etc. etc. etc.)

From the fMRI results can you:

<snip for brevity>

This was his reply:

The answer to all 14 questions in the situation you pose is a very clear 'no'. I am one hundred percent confident that no-one doing this sort of science anywhere in the world would disagree with me. If you are asking these kinds of questions, then I am concerned you may be radically mis-interpreting the scientific literature. While some people make slightly over-zealous claims, I have never in my entire scientific career ever read any claim that any of your 14 questions could be answered 'yes'.

Essentially, Rees position would appear to be reflected by his statement: “brain reading will be restricted to simple cases with a fixed number of alternatives...for all of which training date are available'....because of the all but infinite number of cognitive states and necessarily limited training categories.”

And I never said otherwise. But you, on the other hand, claimed that it was entirely impossible to get even a general idea of what someone was saying under any circumstances.

Essentially…with currently available technology and processing and interpretive software …if you or anyone were to walk off the street into a lab and be wired into a neural scanner (fMRI for example) the results would NOT be able to show EITHER an explicit or even general representation of:- what your are seeing

Rees is not stating that it is impossible to extract visual - or any other type of sensory - data from the brain, as you are. He is stating that it is possible to do so with limitations, which is what I have been saying all along.

By the way Argent, this is what you had to say about Rees statements before you emailed him

Yes, some of my statements were wrong. I admit that. Your point?

Once again…you make a whole bunch of massive claims…without a shred of substantiation. No-evidence-Argent.

Actually, I have presented the evidence. It's still there, for anyone who cares to look back a couple of pages.

…now isn’t that interesting. Why don’t the members of our audience have a look above and note just how many times Argent flatly rejected the points made on that very same list! Argent either generally or completely dismissed every single statement Rees made. But now, for some mysterious reason, the list is generally correct. Which areas Argent? Do we have to change some of your no’s to yes’s? Which one’s?

I explained it in that very post. These questions remain unanswered in many areas of neurology. I admitted that.

…a list that (as I clearly stated when I wrote it) goes on forever and includes every variety of human experience (there was another clearly stated qualification but you’ve ignored everything else so I’m not at all surprised you ignored that as well). But now, once again, Argent’s position changes. Only in ‘specific cases’ can neural scanning read minds.

I have said this from the beginning.

So before it was…’EVERYTHING’….but now it’s …’only in specific cases’.

It was never "everything". It was always regarding emotions. And, later, when you expanded it, visual data.

Please provide a specific link where I have claimed that emotions and visual data cannot be decoded. I never once said this.

Essentially…with currently available technology and processing and interpretive software …if you or anyone were to walk off the street into a lab and be wired into a neural scanner (fMRI for example) the results would NOT be able to show EITHER an explicit or even general representation of:- what your are seeing

The only position I have ever taken, and argued extensively, is that there are substantial limits to what can be decoded.

If that's all you're arguing, fine. We have no argument, except in that the limits you are positing are incorrect.

 

[Robin]

Explain why.

Obviously because it is best to judge what somebody says by seeing it in their own words and in context of what they were asked.

 

[carlitos]

Why didn't annnnoid just paste the emails? I missed something.

I had some cool chaos magick last night. I think if I intentionally close my eyes only partially when going to sleep, it improves my chances for hypnagogia. I actually had a freaking vision. Cool stuff.