Protanopic vision sees this To a protanope—who has two-component colorblind vision because it lacks red-sensitive cones— "normal" human color is magic.

while most others see this To ordinary human color vision—which has three components of color—the six-component color of a bird is magic

Understanding of this abstract dimensionality of color vision was one of Feynman's magical insights.  (It's a remarkably difficult concept for most ordinary people.)

Two of the three "normal" human color dimensions are the X and the  Y of this diagram.  The third dimension is Z, perpendicular to X and Y: it represents intensity.

Protanopic color—all two dimensions of it—is represented by the family of red straight lines.  The "700" point is black. Moving away from "700" on one of those lines, a protanope sees the same color, ever brighter.

This is pretty much the only kind of measurement familiar to those outside of certain specialized branches of science or technology.  Nevertheless, it is usually oversimplification.

some people are brighter than others
.

Visually follow the development, "from infancy to science-see," of the increasing sophistication and power of human concepts of measure, HERE.

Putting colors into an ordered array illustrates a simple vector measure: three components for ordinary human color, two components for protanopic colorblindness (deuteranopic, too).  Total colorblindenss is one component (scalar).  Bird color can have four, five, or six components.  Spectroscopic color has one component for each and every possible wavelength: an infinity of components.

When a surface reflects light, each wavelength will, in general, reflect a fraction of the light that falls on the surface characteristic of that particular wavelength.  That would make the "reflectivity" of the surface an infinite component vector, if that were the whole story—the vector could be represented by a graph of those fractions.  However, the surface might fluoresce: a given wavelength might also reflect at other wavelengths, too, like those bright orange detergent boxes on the supermarket shelves.  Each wavelength striking the surface needs a whole graph of its own showing the amount of all the wavelengths at which it gets reflected.  That makes the reflectivity of a surface a tensor. 

It's a rank-two, order-infinity tensor.

The rest of the real world needs attention to multi-component measures, too.  Value of a human being.  Reward to a human being.  Cost of a product.  Benefit of a product—of an action, of a law, of a belief...  These all have multiple components.  Every time we imagine a line-up by "value," every time we rank anything to tell which is better, bigger, more valuable, or whatever, we are almost certainly oversimplying.  The oversimplification can be so great as to make our ranking virtually meaningless.

Sooner or later we will need to use tensors.  Or face the consequences of oversimplification.

Food (and fuel for the car) supplies what we "run out of"; it "restores energy."  This colloquial kind of energy is a kind of capacity for doing work.

Energy...Power...Work
What in the real world are they??

Oh oh!  That makes "Energy is the capacity to do work" a logical impossibility.  But it's a logical contradiction that many textbook authors miss seeing.  Feynman not only sensed the contradiction, he saw far beyond it.

Feynman's picture of energy is something very abstract.  We have learned how to calculate a value for it in a lot of situations...and we have discovered that it is conserved.  As long as we take everything affected into account, it's total value doesn't change no matter what happens to those things.

Feynman's magic was in the relam of the theoretical.  But, for him, it was in the realm of the practical: he could apply it with ease.  He could see it in the real and concrete world that he encountered day in, day out.  For the rest of us it was "not easily understood"—and often not understood at all.

Mathematician Kieth Devlin has a very simple resolution of what many might see as dissonance in these observations.  Those who, like Devlin and Feynman, have had a lot of experience working out mathematical problems will see those deeper abstractions as having a kind of reality of their own.  The abstractions are patterns in the real world.  They are patterns of patterns.  They are patterns of patterns of patterns of...

The nearly universal pattern of learning that beginning students of science and mathematics follow is prodigous learning of specific instances of applications of the abstractions.  That makes it learning without understanding. It's hollow and almost useless.

As he read through the books, his wife complained that he repeatedly burst into shouts, "erupting like a volcano."  ALL of the books were UNIVERSALLY LOUSY. 

Those books have been the source of learned science for almost everyone in California...and probably almost everywhere else, too.  The authors of those textbooks are the teachers of our teachers.  Teachers and learners alike are learning something that goes by the name "science," but it is not science in the sense that scientists have learned...and understood.

And can use.
 

The Feynman Lectures on Physics contain many remarkable insights that will be found in very few, if any, other texts.  You will find many other references to these insights if you follow the above links leading to other pages of Explore the Physicist's Domain and Knowledge for Use.