[ExI] Why stop at glutamate?

Mon Apr 10 02:27:14 UTC 2023

I want to think in more concrete terms and in fact something that in real
life could be done in a real experiment instead of a bunch of words like
quality and stuff like that.
I do the following experiment. I show 1000 people a screen with the large
square with what I recognize as the color red. I measure something tangible
and usually associated with brain activity. For example blood flow via fMRI
or brain waves measured with an EEG or both preferably. I could add other
measurable quantities like the level of neurotransmitters or even measure
single neuron activity with a grid of very fine electrodes implanted in the
brain. The more the better. But also reducing the number of inputs and see
how much information I can extract with the minimum input could be
interesting. I then train I do the same with an image of the color green.
Then I train a neural network of given sophistication to make associations
between neural activation and other brain correlates in the two cases.
There are a lot of ways to process the information and interpret the data.
I know by experience from similar studies (not necessarily focused on
color) that one interesting way to show the data is to plot areas in the
brain that were activated when the 2 colors were shown. What you will see
is that there is a lot of overlap between people activity when they see the
same color. We can also show that the activation for green is different
than red. You could also create what is called a phase diagram. For
example, you could show the value of activity (some level of blood flow or
whatever) in region A which is mostly associated with red and region B
which is mostly associated with green. Then each dot in the diagram
represents level of activity in region A vs region B in each individual. We
can color code (with red and green for example) the different experimental
conditions for each individual and their value of A and B for each
condition. What you would see in the end is 2 blobs in this space that are
not perfectly separated but mostly so. Maybe there is a region that
overlaps. You can then take another sample of people and do the same
experiment and "guess" what color they are seeing by plotting their
particular values of A and B on this map. If that dot (defined by
coordinates A and B) falls squarely in the region associated with the red
experience by most people than I can say with confidence (that can be even
calculated mathematically as a probability) that person 1234 is seeing red
vs green. Let's imagine now that there is subject 1359 that when shown what
you and I recognize as red on the screen he responds with a set of (A,B)
that falls in the green area. Not even in the overlapping region just all
the way there where it is supposed to be a strong green response but on the
screen we know we have red.
What is the interpretation of this?
This is a valid way to ask this question. It could be a problem with our
measurement. We go back double check everything, we calibrate everything
again and we get the same result. It could be that some rare condition
changes the wiring of this person and it "translates" the physical
information of red into a green pattern in his brain. Anyway, it is an
interesting anomaly and we could focus on this anomaly and try to
understand what happens. But most of the time science throws away such
anomalies because they are outliers and we care more about averages and
things in common among a large set of samples. But it still valid to focus
on a single anomaly and figure out what happens.
Now, another way to explore this in a scientific way would be to see what
happens when I stimulate the brain areas in question. This is an important
approach because we know in science we often look at correlation but
correlation doesn't imply necessarily direct causation.
I dealt with such a problem in neuroscience. In our case, we looked at slow
waves that are associated with memory consolidation. We knew, from other
studies, that slow waves are associated with memory performance. I can give
a memory test to multiple people in the evening, give them a score and then
test them in the morning on the same test but randomized in terms of the
order of things to remember. People in general score better than in the
evening showing that sleep has improved memory. I then can create a graph
showing the average amplitude of the wave during the night vs their memory
score (difference between performance in the evening vs morning). You can
see there is a nice correlation between the 2 parameters.
So slow waves are associated with the improvement in memory due to sleep
but do they have a causal role?
One way to answer this question would be to manipulate the waves and see if
this manipulation gives a memory benefit that is different from a night of
sleep without manipulation. We did exactly that and I have a patent for a
device that can improve memory in people that is based on this particular
manipulation of slow waves (we used short bursts of pink noise that are
synchronized to the individual brain waves in real-time).
So this is how you think about these problems in a scientific way instead
of vague philosophical-like conundrums that are confusing and don't help us
to understand how reality works.

Giovanni

On Sun, Apr 9, 2023 at 6:34 PM Brent Allsop <brent.allsop at gmail.com> wrote:

>
>
> On Sun, Apr 9, 2023 at 7:17 PM Giovanni Santostasi <gsantostasi at gmail.com>
> wrote:
>
>> I'm saying that science is about map making. This is all what we can do
>> and all what we should do.
>> It is called modelling. It is the core of science.
>> It is the most important concept to understand about science. We make
>> models, good models are useful not necessarily true. In fact, a good model
>> needs to have less information than what it represents to be useful. Apply
>> all what I'm saying above to this entire business of understanding redness.
>>
>
>
>
>> Science to explain red needs only to make a good model of what redness
>> is, not to give the subjective experience of redness to somebody as if was
>> some kind of enlightenment experience. I can use the model to make you
>> experience red if I wanted by extracting what is essential in redness and
>> reproduce that with some artificial mean (for example electrodes in your
>> brain) even if you optical nerve was severed.
>>
>
> I think this is exactly the same thing I'm trying to say.  Are you
> saying we could come up with a model of your redness, and a model of your
> greeness.  Then if you objectively observed someone else's brain as they
> experienced their redness, yet you objectively matched this up with your
> model of greenness, you could then know that his redness was like your
> grenness?
>
> This is how you know we do understand what the heck red is. This is how we
>> make airplane flight or this is how we communicate on these computers.
>> Do you get what I'm trying to say?
>>
>
>
>
>
> On Sun, Apr 9, 2023 at 7:17 PM Giovanni Santostasi <gsantostasi at gmail.com>
> wrote:
>
>> I'm saying that science is about map making. This is all what we can do
>> and all what we should do.
>> It is called modelling. It is the core of science.
>> It is the most important concept to understand about science. We make
>> models, good models are useful not necessarily true. In fact, a good model
>> needs to have less information than what it represents to be useful. Apply
>> all what I'm saying above to this entire business of understanding redness.
>> Science to explain red needs only to make a good model of what redness
>> is, not to give the subjective experience of redness to somebody as if was
>> some kind of enlightenment experience. I can use the model to make you
>> experience red if I wanted by extracting what is essential in redness and
>> reproduce that with some artificial mean (for example electrodes in your
>> brain) even if you optical nerve was severed.
>> This is how you know we do understand what the heck red is. This is how
>> we make airplane flight or this is how we communicate on these computers.
>> Do you get what I'm trying to say?
>>
>>
>> On Sun, Apr 9, 2023 at 6:08 PM Brent Allsop <brent.allsop at gmail.com>
>> wrote:
>>
>>>
>>>
>>> On Sun, Apr 9, 2023 at 6:59 PM Giovanni Santostasi <
>>> gsantostasi at gmail.com> wrote:
>>>
>>>> *how do you eff the ineffable nature of colorness quality experience*s.
>>>> You don't because it is not what knowledge or science is about. You are
>>>> confusing the territory and the map. We already discussed this.
>>>> You are basically putting the finger on a blue line on a map and saying
>>>> "this is not a river because my finger is not wet". It is a freaking map.
>>>> Or you are looking at the drawing of an engine in a very detailed blue
>>>> print and say "I don't hear the engine noise or it doesn't move". Do you
>>>> understand what I try to tell you?
>>>>
>>>
>>> I'm trying to understand what you mean here, but it makes no sense to
>>> me.  I'm also trying to point out that rednes is the "map" and not the
>>> territory (the light, or the thing that reflects the light).   So why are
>>> you saying you can't know what that map (which is not the territory)
>>> subjectively like?
>>>
>>>
>>>
>>>
>>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20230409/ccf1bb74/attachment-0001.htm>