What a JPG file actually is
JPEG has been the default photographic format since 1992. It divides the image into 8×8 blocks and discards the high-frequency detail within each one, which is why it compresses photographs so well and why hard edges and flat colour — screenshots, logos, text — pick up the smeared halos known as ringing artifacts.
Almost every camera, scanner and non-Apple phone writes JPG, and it is what a website will hand you when you save an image. If a file came from somewhere and you have no idea what produced it, JPG is the safe bet.
What you get from PNG
PNG output is lossless, so the saved pixels are exactly the pixels that went in and re-saving costs nothing in quality. It keeps a full 8-bit alpha channel, which means soft shadows and anti-aliased edges survive against any background.
Universal since the early 2000s, and the format every design tool, browser and operating system handles without argument. It is also the safest format to hand to a printer or a colleague when you do not know their software.
What changes when you convert JPG to PNG
PNG is lossless, so the conversion preserves exactly what the JPG decoded to — but not the detail the JPG had already discarded. You are freezing the current state, not recovering the original, and a common misunderstanding is that saving to PNG somehow restores quality. Expect the file to grow, often several times over, because lossless compression cannot match what lossy compression already did.
There is no transparency to preserve — JPG files do not carry an alpha channel, so every pixel is opaque to begin with. Converting to a format that supports transparency does not create any; it only means you could add it later in an editor.