Philosophers, Physicists and Cranks

[Skwinty]

With respect to the consciousness issue, have any of you folk read any of Max Tegmarks papers on "the theory of everything" , "math, matter and mind" and " the importance of quantum decoherence in brain processes".

IMHO Tegmark is a respectably published physicist and these papers indicate that this topic is being considered by physicists and although the issue poses huge problems that are far from resolution, they are valid questions and not gibberish as intimated by some.

 

[godless dave]

Hi Godless Dave,
Firstly, physics is a broad based field. I agree that a Nuclear or Particle physicist for example would not be able to answer the question of conciousness but a Biophysicist should be able to determine the fundamentals of conciousness.

Biophysics isn't a kind of physics, it's an interdisciplinary field of study involving biology. http://en.wikipedia.org/wiki/Biophysics

Secondly, Physics stems from reason and reason is but a subset of conciousness.

Reason is what scientists use to study and describe reality. But reason, and consciousness, are subsets of reality, not the other way around.

Thirdly, how can science be serious about a theory of everything, if it does not include conciousness. After all, conciousness is the fountainhead of all thought and theory.

As has been pointed out, you are misunderstanding what physicists mean by a theory of everything.

 

[godless dave]

Links to Tegmark's publications can be found here:

http://space.mit.edu/home/tegmark/

I don't think he is saying what you think he is saying.

On Math, Matter, and Mind

We discuss the nature of reality in the ontological context of Penrose's math-matter-mind triangle. The triangle suggests the circularity of the widespread view that math arises from the mind, the mind arises out of matter, and that matter can be explained in terms of math. Non-physicists should be wary of any claim that modern physics leads us to any particular resolution of this circularity, since even the sample of three theoretical physicists writing this paper hold three divergent views. Some physicists believe that current physics has already found the basic framework for a complete description of reality, and only has to fill in the details. Others suspect that no single framework, from physics or other sources, will ever capture reality. Yet others guess that reality might be approached arbitrarily closely by some form of future physics, but probably based on completely different frameworks. We will designate these three approaches as the fundamentalist, secular and mystic views of the world, as seen by practicing physicists. We present and contrast each of these views, which arguably form broad categories capturing most if not all interpretations of physics. We argue that this diversity in the physics community is more useful than an ontological monoculture, since it motivates physicists to tackle unsolved problems with a wide variety of approaches.

The importance of quantum decoherence in brain processes

Based on a calculation of neural decoherence rates, we argue that that the degrees of freedom of the human brain that relate to cognitive processes should be thought of as a classical rather than quantum system, ie, that there is nothing fundamentally wrong with the current classical approach to neural network simulations. We find that the decoherence timescales ~10^{-13}-10^{-20} seconds are typically much shorter than the relevant dynamical timescales (~0.001-0.1 seconds), both for regular neuron firing and for kink-like polarization excitations in microtubules. This conclusion disagrees with suggestions by Penrose and others that the brain acts as a quantum computer.

 

[ben m]

With respect to the consciousness issue, have any of you folk read any of Max Tegmarks papers on "the theory of everything" , "math, matter and mind" and " the importance of quantum decoherence in brain processes".

To clarify Godless Dave's Tegmark excerpt: the question Tegmark asks is "do neurons behave differently under a full Schrodinger's Equation treatment than they do under the best classical approximation", and suggesting the answer is "no". You can ask this sort of question about any system. (Household wiring? Classical is OK. Electron microscopes? Quantum mechanics needed.) This is different than the question "what's the reality underlying Schrodinger's Equation?" which is the one we've been talking about in this thread.

And, yes, Tegmark is an excellent and well-respected physicist who pays attention to the more philosophical side of things. Some of us do, as I said in (IIRC) my first reply to you.

 

[Skwinty]

Hi Godless Dave
This is where I draw the link between physics and biophysics

"Because physics treats the core workings of the universe, including the quantum mechanical details which underpin all atomic interactions, it can be thought of as the foundational science, upon which stands "the central science" of chemistry, the earth sciences, biological sciences, and social sciences. Discoveries in basic physics have important ramifications for all of science."http://en.wikipedia.org/wiki/List_of_basic_physics_topics

Here is a quote from Tegmark's paper on "Decoherence and the Brain"
"In most current mainstream biophysics research on cognitive processes, the brain is modelled as a neural network obeying classical physics. In contrast Penrose and others have argued that quantum mechanics may play an essential role...............

.........Encouraged by successes in modelling memory,learning, visual processing etc, many workers in the field have boldly conjectured that a sufficiently complex neural network could in principle perform all cognitive processes that we associate with consciousness" end quote

There are numerous other references to conciousness in the paper, but as I only have pdf versions of the papers, I will not type them out, but expect that you will read them.

A quote from "On Math, Matter and Mind"

"I believe that consciousness is the way information feels when being processed. Since matter can be arranged to process information in numerous ways of vastly varying complexity, this implies a rich variety of levels and types of consciousness..................This implies that there is nothing wrong with the Matter - Mind link...........................This hypothesis has clearly not been proven.........since it can strictly speaking never be proven. I cannot even prove to my colleagues that I personally am self aware- they simply have to take my word for it ................ I have seen no hard scientific evidence against it. Rather, many objections seem to be based on a combination of human vanity and wishful thinking" end quote

A quote from "Is the theory of everything merely the ultimate ensemble theory"

"The only postulate in this theory is that all structures that exist mathematically exist also physically, by which we mean that in those complex enough to contain self aware substructures those SAS's will subjectively perceive themselves as existing in a physically real world. We find that it is far from clear that this simple theory, which has no free parameters whatsoever, is observationally ruled out" end quote.

Niow given the above quotations, what is wrong with the question "What is consciousness and can physics provide the answer, or should it be left for philosophy"

Hi Ben yes IRC. I know that I am thought of as a Witt sock puppet albeit mistakenly. I do appreciate the fact that people more highly educated than myself take the time to comment on my questions on topics of which I know very little. At very least I am prepared to read and ask questions to try and learn more. Please note the posts I make are in my opinion questions and not claims or statements proclaiming my brilliance.

I do not rely on the herb yielding seed or the philosophers stone for enlightenment.

 

[godless dave]

Hi Godless Dave
This is where I draw the link between physics and biophysics

"Because physics treats the core workings of the universe, including the quantum mechanical details which underpin all atomic interactions, it can be thought of as the foundational science, upon which stands "the central science" of chemistry, the earth sciences, biological sciences, and social sciences. Discoveries in basic physics have important ramifications for all of science."http://en.wikipedia.org/wiki/List_of_basic_physics_topics

I don't see you drawing a line there, just quoting stuff about physics. And it's true, physics studies the stuff that underlies everything else, but if you wanted to synthesize a polymer, you'd call a chemist, not a physicist. If you wanted to look at how E. coli evolves, you'd call a biologist, not a physicist.

Here is a quote from Tegmark's paper on "Decoherence and the Brain"
"In most current mainstream biophysics research on cognitive processes, the brain is modelled as a neural network obeying classical physics. In contrast Penrose and others have argued that quantum mechanics may play an essential role...............

.........Encouraged by successes in modelling memory,learning, visual processing etc, many workers in the field have boldly conjectured that a sufficiently complex neural network could in principle perform all cognitive processes that we associate with consciousness"

And if you read the abstract of the paper I quoted above, Tegmark is arguing that Penrose is wrong.

A quote from "On Math, Matter and Mind"

"I believe that consciousness is the way information feels when being processed. Since matter can be arranged to process information in numerous ways of vastly varying complexity, this implies a rich variety of levels and types of consciousness..................This implies that there is nothing wrong with the Matter - Mind link...........................This hypothesis has clearly not been proven.........since it can strictly speaking never be proven. I cannot even prove to my colleagues that I personally am self aware- they simply have to take my word for it ................ I have seen no hard scientific evidence against it. Rather, many objections seem to be based on a combination of human vanity and wishful thinking"

This really has nothing to do with the topic at hand.

A quote from "Is the theory of everything merely the ultimate ensemble theory"

"The only postulate in this theory is that all structures that exist mathematically exist also physically, by which we mean that in those complex enough to contain self aware substructures those SAS's will subjectively perceive themselves as existing in a physically real world. We find that it is far from clear that this simple theory, which has no free parameters whatsoever, is observationally ruled out"

Niow given the above quotations, what is wrong with the question "What is consciousness and can physics provide the answer, or should it be left for philosophy"

What is wrong with that question is it is a false dichotomy. I doubt very much physics will provide the answer to consciousness. If an explanation is found for consciousness, I expect it to be found by biology and the study of complex biological systems. Philosophy can help us define consciousness, but it can't tell us about how or why it emerges. Answering questions about reality is not philosophy's bailiwick. Philosophy is for investigating how humans interact with and think about reality.

 

[Skwinty]

I don't see you drawing a line there, just quoting stuff about physics. And it's true, physics studies the stuff that underlies everything else, but if you wanted to synthesize a polymer, you'd call a chemist, not a physicist. If you wanted to look at how E. coli evolves, you'd call a biologist, not a physicist..

I was drawing a link not a line.
Discoveries in basic physics have important ramifications for all of science.
Why do they call them biophysicists then?
Pedantic semantics?
Why then are physicists like Tegmark , Penrose and others writing papers about consciousness?

And if you read the abstract of the paper I quoted above, Tegmark is arguing that Penrose is wrong. .

Yes,Tegmark argued that Penrose was wrong about the brain being a quantum computer due to the temperature of the brain and the decoherence implications. I wasn't implying that Penrose was right, merely that consciousness was being investigated by physicists even astrophysicists.

This really has nothing to do with the topic at hand..

If every is considered mathematically then self aware substructures must be part of the equation

What is wrong with that question is it is a false dichotomy. I doubt very much physics will provide the answer to consciousness. If an explanation is found for consciousness, I expect it to be found by biology and the study of complex biological systems. Philosophy can help us define consciousness, but it can't tell us about how or why it emerges. Answering questions about reality is not philosophy's bailiwick. Philosophy is for investigating how humans interact with and think about reality.

Dont physicists interact with and think about reality? If so then they must be in the philosophers bailiwick as you say.Isn't consciousness part of reality? Why do physicists look at physics with a philosophical outlook?

 

[godless dave]

Why then are physicists like Tegmark , Penrose and others writing papers about consciousness?

In my opinion, Penrose is doing it because he has a wacky idea he is trying to support. Tegmark is writing in response to Penrose, and arguing that consciousness is not in the domain of study of physics.

Yes,Tegmark argued that Penrose was wrong about the brain being a quantum computer due to the temperature of the brain and the decoherence implications. I wasn't implying that Penrose was right, merely that consciousness was being investigated by physicists even astrophysicists.

By a very small number of physicists, just Penrose and a couple other guys.

Dont physicists interact with and think about reality? If so then they must be in the philosophers bailiwick as you say.

Physicists, yes. Physics, no.

Isn't consciousness part of reality?

Yes. And it's in the part of reality studied by biologists and psychologists.

Why do physicists look at physics with a philosophical outlook?

Do they?

 

[ben m]

I was drawing a link not a line.
Discoveries in basic physics have important ramifications for all of science.

Not true. For example: the Glashow-Weinberg-Salam electroweak unification is one of the most important discoveries in "basic physics" in the past 40 years. What impact has this had on chemistry? Zero. Chemists in 1970, like chemists today, use the 80-year-old equations of quantum mechanics (or, in rare cases, the 60-year-old equations of QED) to tell them everything they need to know about atom behavior. Once you know 1930s-level atom behavior, you've got a lifetime of complex chemical phenomena to study without getting anywhere near the limits of precision that QED clears up, much less GWS. You will find zero chemistry-research conclusions that were modified, or overturned, or affected at all, by the discovery of GWS unification.

Chemistry is founded on physics, but it's not linked to basic physics the way you imply.

Ditto for biology; biologists today, like biologists in 1950, use a fairly coarse description of electron/molecule/ion behavior. It's only in special cases that you need quantum mechanics at all, much less QED, much less GWS. You will find no biology papers whatsoever that were modified/overturned/affected by the discovery of QCD, or GWS, or quarks, or anything like that. The physics phenomena which biologists care about are the ones that were discovered in the 1930s.

If biology and chemistry weren't affected by QED, nor GWS, nor QCD, nor GR, what makes you think that Grand Unification, or the ToE, will do things differently? Such theories will, after all, include a nearly-identical-to-today's list of atom-scale phenomena.

 

[Skwinty]

Godless Dave,
You really need to read the paper.
Tegmark argues that the brain works as a classical device and not as a quantum device. At no stage does he suggest that this is not the domain of physics.

There are also more than just a few other guys with wacky ideas looking at this issue.

A quote from Tegmark

"In summary, our decoherence calculations have indicated that there is nothing fundamentally quantum mechanical about cognitive processes in the brain, supporting the Hepps conjecture. Specifically, the computations in the brain appear to be of a classical nature rather than a quantum nature, and the argument by Lisewski that quantum corrections may be needed for accurate modelling of some details, eg non markovian noise in neurons, does of course not change this conclusion. This means that that although the current state of the art in neural network hardware is clearly still very far from being able to to model and understand cognitive processes as complex as those in the brain, there are no quantum mechanical reasons to doubt that this research is on the right track" end quote.


Ben

The link between physics and other sciences maybe outdated but it still exists. The information comes from .http://en.wikipedia.org/wiki/List_of_basic_physics_topics


It seems to me that the issues here are largely about semantics and linguistic baggage. Perhaps scientists should also study the language and grammar of science.

Please refresh my memory as to what I am implying, because the last time I checked it was that Physics could answer the question of consciousness, philosophy could make a better physicist and the theory of everything as you define it, is not really a theory of everything.

 

[ben m]

Godless Dave,
You really need to read the paper.
Tegmark argues that the brain works as a classical device and not as a quantum device. At no stage does he suggest that this is not the domain of physics.

... but at no point does he suggest that "consciousness" is anything other than a result of the interactions of (individually fairly simple) neuron behavior; at no point does he suggest that neuron behavior results from anything other than the (individually fairly simple) actions of atoms, molecules, and electric fields. Here's how I would summarize his argument

"The behavior of X atoms, examined at scale Y in a thermal bath, resembles a classical computer more than a quantum computer. Therefore, any suggestion that any macro-system is a "quantum computer" is incorrect. One example of a macro-system which, therefore, cannot be a quantum computer, is the brain."

This does not mean that "neuroscience is really physics". You could easily replace "brain" in the above discussion with "kidneys", "Pentium II", or "Ogallala aquifer", and Tegmark's statement would still be equally valid, but this would not mean that nephrology, computer science, and hydrology are "really physics".

The link between physics and other sciences maybe outdated but it still exists. The information comes from .http://en.wikipedia.org/wiki/List_of_basic_physics_topics

(a) I didn't say it was "outdated". Chemists usually use a set of quantum-mechanical calculations of atomic structure which are valid to (say) four or five decimal places. Since their problems are so complicated, though, they only care about (say) the first three. Meanwhile, physicists have improved our understanding of electroweak theory such that, *if they wanted to*, the chemists could pick up QED atomic calculations today which are accurate to 10 or 12 decimal places. Adding more QCD detail (in the nucleus, in virtual loops) might get another decimal place or two. Are the chemists' calculations then "outdated"? Not at all; they're using the appropriate level of precision for their problem.

It's like if you're building a house---what sort of ruler do you use to measure wall studs before cutting them? A tape measure? Sorry, your tape measure is 30 years old and NIST's calibration for the length of the meter has drifted since then. Here, take this high-finesse cavity and a comb-stabilized laser; now you've measured your wall studs to 0.1 nanometers precision and thus your house is based on physics. (Then, of course, you make a mark on the wood with a 1,000,000nm-wide pencil stub and line it up by eye with a 3,000,000nm-wide crosscut saw.) And your house looks exactly the same as mine, which I measured with the 30-year-old uncalibrated tape measure---so you couldn't call my construction "outdated".

That's where chemistry is. Physics in the 1960s was pretty darn close to getting atom behavior exactly right. Chemistry takes those old, pretty-darn-accurate atomic models and plugs them into complex, approximate, and computationally-limited molecular models. There are so many uncertainties introduced in the chemical details; who cares if "fundamental physics" got the fifteenth digit wrong in some aspect of the atomic model?

That said, of course there's always work at every interface. There's a high-energy/atomic interface, an atomic/molecular interface, there's chemical physics and physical chemistry, and so on. But please don't use this chain to argue "everything depends on physics" unless you know what you're talking about. We've been down that road (Rutherford and his "stamp collecting" remark) and it was dumb.

(B) What's the Wikipedia link supposed to be telling us?

It seems to me that the issues here are largely about semantics and linguistic baggage. Perhaps scientists should also study the language and grammar of science.

The semantics arise when you are trying to make the word "physics" fit into impractical categories. "Anything that can be traced back to physics" you want to call "physics". Sorry, there are mainstream conventions for that, which some working physicists are trying to explain to you. If you don't like the explanations, or if they're not what you expected, this does not mean that "scientists should study the grammar of science"

Please refresh my memory as to what I am implying, because the last time I checked it was that Physics could answer the question of consciousness,

Yep. Which the consensus, Max Tegmark included, are telling you is incorrect.

philosophy could make a better physicist

Yep.

and the theory of everything as you define it, is not really a theory of everything.

Back to grammar. Well, we've been using the words "theory of everything" to refer to grand unification + gravity for two decades. That's how technical language works. I'm sorry that you don't like it, and I'm sorry that you want somebody to be working on a Theory of Absolutely Everything. Nobody is working on such a theory, because it appears to be a completely pointless, dead-end, insight-free approach to lots of basically unrelated problems.

 

[lomiller]

I think there are some areas in which science and philosophy overlap and in those cases one will have much impact on the other and it’s inevitable that some of the best people will have an understanding of both. Science will of course always be more useful then Philosophy or Religion because it focuses on making useful predictions. This means however that there are certain types of questions it can’t really answer.

The thing to watch for, IMO, isn’t Philosophy vs Science it’s when someone is talking outside their field. One of the main things that separates cranks from legitimate scientists or legitimate philosophers is that they are frequently expert in one area and assume that qualifies them to speak in a filed in which they have no real training or experience. They make “discoveries” and assume the people who study in that field have never considered and because they are expert in their own field their ego won’t allow them to listen when told the flaws of their theory.

Philosophers may be more prone to this because they have a less focused view and lack the validation cycle that exists in science, but you see this in all areas of science philosophy and religion.

 

[godless dave]

Godless Dave,
You really need to read the paper.
Tegmark argues that the brain works as a classical device and not as a quantum device. At no stage does he suggest that this is not the domain of physics.

There are also more than just a few other guys with wacky ideas looking at this issue.

A quote from Tegmark

"In summary, our decoherence calculations have indicated that there is nothing fundamentally quantum mechanical about cognitive processes in the brain, supporting the Hepps conjecture. Specifically, the computations in the brain appear to be of a classical nature rather than a quantum nature, and the argument by Lisewski that quantum corrections may be needed for accurate modelling of some details, eg non markovian noise in neurons, does of course not change this conclusion. This means that that although the current state of the art in neural network hardware is clearly still very far from being able to to model and understand cognitive processes as complex as those in the brain, there are no quantum mechanical reasons to doubt that this research is on the right track" end quote.

This supports my argument, not yours.

Please refresh my memory as to what I am implying, because the last time I checked it was that Physics could answer the question of consciousness, philosophy could make a better physicist and the theory of everything as you define it, is not really a theory of everything.

You've offered nothing to support the first implication, little to support the second, and no one disagrees with the third.

 

[Skwinty]

Hi Ben
Thanks for the comprehensive answer to my last post. I realise now, that my interpretation of certain words are prone to
be misconstrued, for example outdated. I should perhaps just said dated or old. The word everything is another. I
interpreted it literally to mean everything, whereas scientists used the term perhaps in a jocular manner.
The other has been my misuse of the word physics, although no-one answered my question as to why there are
biophysicists. I should have used the word science but then it would have been said that this term was probably too
generic. Remember, I am not a scientist, only an engineer. My bad.
I guess that is why mathematics is the appropriate language for science because it rules out the linguistic baggage. I still
need some help here with why you say that Tegmark is saying that consciousness is not the domain of physics. I have
read the paper on Decoherence and the Brain numerous times and can't see the implication that consciousness is not the
domain of physics. If the brain is likened to a classical computer which owes its existence to physics, then is that not
sufficient correlation? As far as me not liking or being unsatisfied with the technical language, the reception of Null Physics
on this forum, the issue of consciousness or that no-one is working on the absolute theory of everything, its just that I dont
see the limitations that are imposed which prevent the full realisation of mans quest for ultimate knowledge. Now,I can
understand that if the Universe is truly random, then absolute knowledge is impossible, but if it is only pseudorandom,
then why not? I think that the quote below from Tegmark best sums up how I feel about these issues.

Quote from Shut up and Calculate by Max Tegmark
"Evolution endowed us with intuition only for those aspects of physics that had survival value for our distant ancestors,
such as the parabolic trajectories of rocks. Darwins theory thus makes the testable prediction that whenever we look
beyond the human scale, our evolved intuition should break down. We have repeatedly tested this prediction, and the
results overwhelmingly support it: our intuition breaks down at high speeds, where time slows down: on small scales,
where particles can be in two places at once: and at high temperatures, where colliding particles change identity. To me,
an electron colliding with a positron and turning into a z boson feels about as intuitive as two colliding cars turning into
a cruise ship. The point is that if we dismiss seemingly weird theories out of hand, we risk dismissing the the correct
theory of everything, whatever it may turn out to be. If the mathematical universe hypothesis is true, then it is great news for
science, allowing the possibility that an elegant unification of physics and mathematics will one day allow us to understand
reality more deeply than most dreamed possible. Indeed, I think the mathematical cosmos with its multiverse is the best
theory of everything that we could hope for, because it would mean that no aspect of reality is off limits from our scientific
quest to uncover regularities and make quantitive predictions" end quote

 

[Skwinty]

This supports my argument, not yours.




You've offered nothing to support the first implication, little to support the second, and no one disagrees with the third.

Hi Godless Dave,
I must say you are a master of the one liner.

 

[ben m]

If the brain is likened to a classical computer which owes its existence to physics, then is that not
sufficient correlation?

In that sense, everything "owes its existence to physics", from chemistry to biology to sociology to art. Similarly, biology/sociology/aesthetics "owe their existence to chemistry" since they're all concerned with things made out of interacting atoms. Similarly, sociology/political science/literature "owe their existence to biology" since they're all concerned with humans. So what? This is true in some sense, but this truth is boring, obvious, and unproductive. You learn exactly zero about sociology by reminding yourself that societies are made of humans which are made of atoms, and even less by specifying that those atoms' nuclear self-gravity energy term comes from broken-mSUGRA rather than Randall-Sundrum-II.

I'm not making wild anti-science claims here, Skwinty---this is standard stuff. Dennett calls your view "greedy reductionism".

As far as me not liking or being unsatisfied with the technical language, the reception of Null Physics
on this forum, the issue of consciousness or that no-one is working on the absolute theory of everything,
its just that I dont
see the limitations that are imposed which prevent the full realisation of mans quest for ultimate knowledge.

You're making presumptions of your own about what, and how, "everything" should be explained. You're presuming, for example, that ordinary neuroscientists, mathematicians, and computer scientists are NOT going to explain consciousness perfectly well, entirely using "neurons" and "action potentials" and "receptors" as building blocks, with zero reference to the gritty physics underlying these things. This presumption-of-failure is completely groundless.

As a practical matter: "the full realization of knowledge", if possible, has to be actually sought. As a practical matter, each of us has to sit down and do something every day that gets us closer to knowing something new. Why, then, aren't neuroscientists spending their days reading string theory---or, for that matter, QED, or chemical physics? They could if they wanted to, but they don't. Because they don't think it'll help.

Now,I can
understand that if the Universe is truly random, then absolute knowledge is impossible, but if it is only pseudorandom,
then why not?

Answer: (a) You've misunderstood "random", if you're referring to the quantum sort of randomness. The randomness of quantum mechanics doesn't mean "absolute knowledge is impossible"---at least for very large classes of interesting knowledge. (c) "Why not"? Complexity, chaos, emergence. Read about them. Reductionism won't always work.

 

[INRM]

What does "Self Energy" of an electron and "Born Energy" mean?


INRM

 

[lenny]

so what about a "how should we do it" question, say, in the statistical design of experiments.

i generally hate biological examples in physics arguments, but i fear this one fits best:

you are testing for the effectiveness of a new drug.

you design a clinical trial, but wish to stop the trial early if the drug is shown to be either useless or harmful. doing the statistics is straight-forward once you know the number of patients the trial impacts, and there is the rub: is the "number of patients" the number of people in the trial -or- the number of people who will have this disease in the next 50 years -or- other.

answering this question is critical to the design of the scientific experiment, yet it is arguably not a scientific question.

philosophers demonstrate utility here, for example.

I'm not sure of the validity of this example, a lot of work will have been done to arrive at the point of a phase II study. This research would not have been performed if there wasn't some sort of gain to be made.

all we can have before human trials is the expectation of "some sort of gain".

the example is real, and i expect "valid". i am told some drug trials in Europe have different stopping rules than in the US, this due in part to the differences in the definition of the "number of patients".

there is a large literature on when to stop trials early, which i do not know. but the basic question is simple: a balance is sought between doing harm to the patients in the study and stopping the study early due to a statistically meaningless fluctuation.

if we only count the individuals in the test set, then we will stop the trial earler than if we consider the potential benifit lost by future patients IF we stop the trial "too early" and in fact fail to recognize an effective treatment as such.

continuing the trail puts those in (an ensemble of) trail(s) at risk.

stopping too early puts those that would have benefited from the incorrectly rejected treatment at risk.

the question of who to consider is not one of statistics or science, yet it plays a major role in the experimental design.

i hope that is a bit clearer.

 

[lenny]

and when you cannot repeat the experiment a "bizzillion times"?
what do they mean then?

Not much, mathematically speaking. And yet I'm perfectly willing to make a bet once if I know the odds, so apparently they mean something.

the odds or the probabilities?

but back to your remark: in the case of weather, we can never expect to reproduce the experient twice, so i agree that, mathematically speaking, the probabilities do not mean much. yet almost all decision support tools assume we have probablities... if you do not have access to probabilities, what do you do?

 

[lenny]

I have never understood what those probabilities are supposed to mean. Is it that the model certainly predicts rain, but with enough theoretical/measurement uncertainty to bring the likelihood down to 80%? That it predicts rain at some point, but only an 80% chance it will be tomorrow? That it predicts rain over 80% of the area in question? Or rain, but only for 80% of the day? Or rain over 80% of a large area that includes Heathrow?

that is a great question. and the answer depends on which country you live in. (the definition of "it rained" depended on which country you lived in until recently, it may still!)

That it predicts rain at some point, but only an 80% chance it will be tomorrow?

that one is the closest, by european standards: the objective probability that it will rain at a given point in space-time.

suppose we run the forecast model 100 times with slightly different initial condution, the distribution of initial conditions reflecting our uncertainty in the current state of the atmosphere. and 80 of those 100 had rain at the space-time point in question. we'd like to say there was an 80% chance of rain there-then.

make sense?