@gravityassistus: Why does the chain rule make you MULTIPLY? What is that extra 2x actually doing? “The derivative of a composite function is the derivative of the outside, evaluated at the inside, times the derivative of the inside.” Okay. Excellent. A sentence has definitely happened. Do either of us know what it meant? Let's look at the machines. Put x into the first one. It squares x and adds one. Feed that answer into the second machine. The second machine cubes whatever it receives. x → x² + 1 → (x² + 1)³. That's a composite function. Two jobs, one after the other. “Inside” means the calculation being fed into the next calculation. Now the question changes: if I nudge x, how much does the FINAL answer move? That nudge has to pass through both machines. So measuring only what the second machine does would miss what happened in the first one. For a moment, swap them for two really simple machines: First machine: multiply by 2. Second machine: multiply by 3. One extra unit going into the first produces two extra units coming out. Each of those produces three extra units from the second. Two groups of three. Six. The combined rate is 2 × 3 = 6. Adding the rates would give five, which doesn't describe what these machines did. These are constant rates, so that example works exactly for whole steps. With curved functions, the rates can change as you move. The derivative measures the local rate at the point you're considering. Same question. Smaller and smaller nudges. Now return to the original machines: g(x) = x² + 1. f(u) = u³. The letter u is just the intermediate answer. It is what comes out of the first machine and goes into the second. Giving it a name doesn't introduce a third machine. The outside derivative is f′(u) = 3u². The inside derivative is g′(x) = 2x. But the outside machine is receiving x² + 1. So that's what belongs in its derivative: f′(g(x)) = 3(x² + 1)². Then account for how quickly that input is changing: y′ = 3(x² + 1)² × 2x. THAT is what the extra 2x is doing. It's the first machine's contribution to the rate of change of the final answer. If you stop at 3(x² + 1)², you've measured the outside's response to its own input. You haven't finished measuring its response to x. There's another detail hiding in the notation: f′(g(x)) means evaluate the outside derivative at the INSIDE VALUE. It does not mean put g′(x) inside f′. The inside derivative is the separate factor you multiply by afterward. Let's check an actual position. At x = 1, the first machine outputs 2. The second outputs 8. But those are VALUES. What about rates? The inside derivative is 2. The outside derivative, evaluated at 2, is 3 × 2² = 12. The combined derivative is 12 × 2 = 24. So the answer is 8, and its local rate of change with respect to x is 24. Different questions. Different numbers. For a tiny positive change h near x = 1, the output changes by approximately 24h. The approximation gets better in relative terms as h approaches zero. A whole step from 1 to 2 is too large to treat that local rate as constant: the output jumps from 8 to 125. You can even check the algebra without the chain rule. Expand first: (x² + 1)³ = x⁶ + 3x⁴ + 3x² + 1. Differentiate each term: 6x⁵ + 12x³ + 6x. Factor it: 6x(x² + 1)². Same answer. The chain rule just lets us follow the structure without expanding everything. At x = −1, the inside rate is −2 and the outside rate is 12. Their product is −24: a small positive nudge makes the output fall. The sign matters too. The formal rule assumes g is differentiable at x and f is differentiable at g(x). Under those conditions, the two local rates multiply. Now try a slightly different machine: y = (3x + 1)². What's the inside? What's the outside? And what factor would you forget if you only differentiated the square? Give me the derivative AND explain where that extra factor comes from. I want to know you followed both machines. #GravityAssist #Calculus #ChainRule #Math

Gravity Assist

Open In TikTok:

Region: US

Monday 19 October 2026 19:42:12 GMT