At every instant the arrow stands still — so how does it ever fly? Zeno's arrow.
Forget average speed — what if you want to know your speed at this very instant?
The speed of change at this very instant — the derivative.
Experiment
Hands-on experiment
Predict first — the average speed from a speed-camera zone and the instantaneous speed on your speedometer: are they the same thing?
An accelerating car — measure its speed at t = 1s
Position: s(t) = 5t² (m). Shrink the averaging interval step by step.
interval: 1s ~ 2.0s
15.00 m/s
Average speed over this interval
| interval width h | average speed (m/s) |
|---|---|
| 1 | 15.000 |
Read more — why it exists · insights · common mistakes · formulasExpand ▾
Why
Why does this exist?
'How far did I travel over the last hour?' — average speed answers that. But 'how fast am I going right now?' can't be answered by an average.
However finely you slice the interval, what you actually want is the value that remains as the interval approaches 0 — the rate of change at a single point.
The derivative defines this 'instantaneous slope' mathematically, unlocking problems where change itself is the point: acceleration, growth rates, marginal cost.
Insight
Insights from the video
“Thinking in rates of change, not results.”
Many problems solve better when you ask 'how fast is it changing now?' rather than 'how much is there now?'. The derivative is the lens for answering that question.
“Averages are blurry; instants are sharp.”
The wider the interval, the more the average rate of change smears out the wiggles inside it. Keep narrowing the interval and it converges to one sharp slope — the tangent. That limit is the derivative.
Misconception
Common misconceptions
A big derivative means a big function value.
The derivative is 'how fast it's changing now', not 'how big it is now'. The trail just before the summit is at near-maximum height, but its slope is close to 0.
Differentiation is the skill of memorizing rules (x²→2x).
The rules are just results. The essence is the idea 'the value the average rate of change approaches as the interval shrinks endlessly'. The process you watched shrinking h in this experiment is the whole of differentiation.
Formula
Writing it as math
As you shrank h in the experiment, the secant (orange) settled onto the tangent (teal). Write that 'endless shrinking' as a formula and you get the definition of the derivative.
Average rate of change (the orange line)
The average slope while x moves from a to a+h. It's exactly the value you controlled with the h slider.
Definition of the derivative (the teal line)
The value the average rate of change converges to as h goes to 0. It captures, in a formula, the moment the secant lay down on the tangent.
The derivative of y = x²
Why the tangent slope was always 'position × 2' wherever you clicked. The slope being 6 at x=3 is this formula.
In Real Life
Where you meet it in real life
A car's speedometer
The needle shows not an average speed but the speed 'at this instant' — position differentiated with respect to time.
Marginal cost in economics
'How much does cost rise if we make one more unit?' The instantaneous slope of the total-cost curve — the derivative — drives a firm's production decisions.
Epidemic spread rate
More than the cumulative case count, 'how fast is it growing today?' (the growth rate) is the key to public-health decisions. The slope of that news-graph curve is a derivative.
How AI learns
AI finds 'the direction in which error decreases' using derivatives (gradients) and inches along it. Gradient descent, the core of deep learning, is the derivative itself.
Try Yourself
Test yourself
Q1Using the rule you discovered: what is the tangent slope of f(x) = x² at x = 7?Show answer ▾
2 × 7 = 14. The tangent slope always being 'position × 2' in the experiment is exactly the formula f′(x) = 2x.
Q2Your speedometer needle points at 60km/h. Is that an average speed or an instantaneous speed?Show answer ▾
Instantaneous — position differentiated by time. The 'average speed' in a speed-camera zone is an average rate of change; the speedometer needle is the derivative. The difference between the two was the theme of this whole lab.
Q3If a function's derivative is 0 at every point, what shape must the function be?Show answer ▾
A horizontal line (a constant function). 'The speed of change is 0 at every moment' = 'nothing changes'. Conversely, if the derivative is 0 only at particular points, those are candidates for peaks and valleys — the key idea of optimization.
Try answering yourself before revealing it — getting it wrong is where learning starts.
Connection
Concepts connect
Previous concept
Limits
The lim inside the derivative's definition — 'send h endlessly to 0' — is exactly the limit. Without limits, the instantaneous slope has no mathematical footing.
← Limits labLeads to next
Probability
The derivative handled 'how a certain value changes'. Probability is the language for situations that aren't certain to begin with. How should we think when change can't be fully predicted?
Go to the Probability lab →Related
Labs worth exploring together
Related lab
Progressive Tax
'If income rises by a dollar, how much does tax rise?' — the marginal tax rate is the derivative of the tax function. 'Marginal' in economic news usually means a derivative.
Go to the Progressive Tax lab →Related lab
Gradient Descent
AI learns by 'walking down the slope'. The tangent slope from this lab becomes deep learning's compass.
Go to the Gradient Descent lab →