concern about duplicate addresses

With so many addresses being generated, isn't it possible that some of them could match with someone else's address out there?

Chances are zero. Don't think too much.

That is called a collision. Let us know when you find one.

For each of the standard types of addresses, there are 2160 valid addresses. The number is so big and as long as the private key is generated randomly, in practice it's impossible that you generate an address that has already generated by someone else.

Addresses get generate using sha256 hash algorithm, each one is different from the others.

Theoretically it is possible, but the probability of it happening is almost zero. Another important fact to know is: since BTC creation, there has never been a single collision. This means it hasn't happened in the past 16 years. So never be afraid of something that hasn't even happened in 16 years, it will be hot & interesting discussion if we see it someday.

Short answer: It's possible, but it's not probable. Somewhat longer answer: It's technically possible, where the chance depends on the address type. For example, there are 2^160 possible legacy (P2PKH) address created from about 2^256 private keys. So there's about 2^96 private key for single legacy (P2PKH) address.

It is like finding an atom in the universe. From private keys to public keys, and public addresses. Finding private keys of the others is impossible.

Fun fact, your seed phrase can generate all of my addresses if you use the proper derivation path. The number of possible derivation paths is astronomically large (as people explained above). This renders the probability of finding a collision close to 0. But, as I said, our seed phrase, though they look (and are) orders of magnitude different, they can potentially generate the same addresses (i.e. we can steal each other's funds - in some parallel universe).