We don’t know anything about derivatives that allows us to compute the derivatives of exponential functions without getting our hands dirty. Let’s do a little work with the definition of the derivative:
There are two interesting things to note here: We are left with a limit that involves \(h\) but not \(x\), which means that whatever \( \lim _{h\to 0} (a^h-1)/h\) is, provided it exists, we know that it is a number, or in other words, a constant. This means that \(a^x\) has a remarkable property:
The derivative of an exponential function is a constant times itself.
Unfortunately it is beyond the scope of this text to compute the limit
However, we can look at some examples. Consider \((2^h-1)/h\) and \((3^h-1)/h\):
While these tables don’t prove that we have a pattern, it turns out that
Moreover, if you do more examples, choosing other values for the base \(a\), you will find that the limit varies directly with the value of \(a\): bigger \(a\), bigger limit; smaller \(a\), smaller limit. As we can already see, some of these limits will be less than \(1\) and some larger than \(1\). Somewhere between \(a=2\) and \(a=3\) the limit will be exactly \(1\). This happens when
We will define the number \(e\) by this property in the next definition:
Using this definition, we see that the function \(e^x\) has the following truly remarkable property.
Hence \(e^x\) is its own derivative. In other words, the slope of the plot of \(e^x\) is the same as its height, or the same as its second coordinate. Said another way, the function \(f(x)=e^x\) goes through the point \((a,e^a)\) and has slope \(e^a\) at that point, no matter what \(a\) is.