In Japan, it's a common insult to call someone "one-cellular" for being stupid, but that put down may not actually be true: In 2001, Dr. Toshiyuki Nakagaki at Riken Institute, Japan, discovered that slime mold* can solve a maze!
Solving a maze: (a) Initial condition with no nutrients. (b) After application of nutrients at two sites. Extensions retract from dead ends, but continue to connect sites possibly containing nutrients. (c) Longer connections gradually disappear. (d) Only the shortest route between the two nutrient sites remain connected. It takes half a day to move from state 'a' to 'd'.
This brings up a question whether something that has no consciousness can be considered intelligent:
I believe that such unconscious information processing mechanisms exist, to a greater or lesser extent, in all living things (for instance, the grouping tendencies of ants, or paramecium). Is this kind of information processing to be considered intelligence? On the other hand, are people with no conscious awareness of themselves, such as one in a coma, or merely asleep, to be considered unintelligent? If we can answer these questions, then we should be able to answer the question as to whether or not single-celled animals possess intelligence.
http://www.riken.go.jp/lab-www/frontier-div/NEWSLETTER/feb2001/ameboid_e.htm - via Scribal Terror
*You can argue that they're not exactly "one-cellular" in the traditional sense as swarms of slime mold Physarum polycephalum cells fuse together to form a plasmodium (a gigantic cell with multiple nuclei and no cell membranes between them). Take my word for it, slime molds are weird!