Keyword search looks for the exact words you typed, which fails when the right document uses different words for the same idea. Semantic search compares the meaning of your query against the meaning of documents, using numeric representations, so it can surface a relevant result that shares no words with your search.
The gain is finding things you could not name precisely; the risk is returning something related but not what you meant. Many systems now combine both, using keywords for precision and meaning for recall, because each catches what the other misses.