Whats'UPIT News, AI Guides & Interactive Math
Tested and written up firsthand

IT news, AI guides, and math you can touch

We rewrite tech news from Korea and abroad fact-first, always linking the original source. We share hands-on ways to use AI tools like ChatGPT and Claude in real work. And we unpack the 'why' of math with interactive visualizations you can play with.

132 hands-on guides 3 categories Only what I've actually used

Recent posts

10 posts

How the definite integral was born — from rectangle approximation to the long S (∫)

The area under a curve has no memorized formula the way a triangle or a circle does. So to measure it, people long ago filled the space under the curve with many very thin rectangles and added up all their areas. A rectangle pokes above or falls short of the curve, so it carries an error, but the finer you slice the pieces, the smaller that overshoot becomes. Slice the pieces infinitely fine and push the error all the way to zero, and the value you reach is the true area. This article confirms, on the easy example of a triangle under a straight line whose answer we already know, that the method really does converge to that answer, and shows with an interactive how the sum of infinitely many thin slices hardened into a long S-shaped symbol, giving birth to the definite integral. The scene where the Greek letter sigma, which once meant addition, transforms into the integral sign meaning the sum of infinitely thin slices is the heart of this article.

August 12, 20269 min
04

Why an n-th order differential equation gets exactly n arbitrary constants — the number of differentiations to unwind

Solve a differential equation and a mysterious constant tags along at the end of the answer. Sometimes there is one, sometimes two. The rule is astonishingly simple. The highest number of differentiations sitting inside the equation, its order, is exactly the number of constants. An equation containing a once-differentiated expression gives one constant, a twice-differentiated one gives two. Why? Because one differentiation erases one layer of information from a function, and to restore that erased layer you must integrate once, and every integration brings back exactly one arbitrary constant. So an equation built by differentiating twice needs two integrations to unwind back to the original function, and it carries two constants. This article integrates a simple twice-differentiated equation by hand, watches the constants appear one at a time, and confirms with an interactive that the two constants mean the starting height and the starting slope.

August 11, 20268 min
Math with AI
07

Why the answer to a differential equation is a function, not a number — and why you may move dx around like a fraction

Solve an ordinary equation and the answer comes out as a few numbers. Ask what number doubles to four and the answer is simply two. But a differential equation answers with a function, that is, a whole curve. What is more, the answer is not one curve but a whole family of curves stacked above and below one another, and only after you fix a single starting point does one of them get picked out. This article first pins down why the answer is a function rather than a number, and then shows that the common shortcut in separation of variables — peeling the tiny change attached to the slope symbol off and moving it to the other side as if it were a piece of a fraction — is really an honest calculation justified by integration by substitution. On a field carpeted with slope arrows you will click a starting point and watch a single solution curve get drawn, and follow the shortcut path and the rigorous path arriving at the very same answer side by side.

August 10, 202610 min
Math with AI
10

Why a complicated fraction splits exactly into a sum of smaller ones — partial fractions and the logarithm that appears when the numerator is the denominator's derivative

To integrate a complicated fraction whose denominator is a product, you first break it into a sum of a few smaller fractions. This is called partial fraction decomposition, and remarkably the split always comes out exact. Why does it always divide perfectly? Because the number of unknowns and the number of conditions to match are built to be equal from the start. And as you integrate each of those pieces, you meet a special shape where the numerator is exactly the derivative of the denominator, and then the answer is simply the logarithm of the denominator, with no computation at all. This is also why the integral of tangent turns out to be the logarithm of cosine. The two stories are really one. With sliders you overlap the sum of two piece-fractions onto the original curve, and by color-matching the numerator against the denominator's derivative you watch a fraction fold into a logarithm.

August 9, 20269 min
Math with AI