During the past 40 years, Cellular Automata (CA) have been extensively used in cryptographic applications, including pseudorandom number generation and symmetric primitives. However, most of the research seems to be published in non-cryptographic venues, which raises a legitimate question: are CA of any relevance to cryptographers nowadays? This talk recollects the experience of the speaker while doing research in CA-based cryptography for the past 10 years, identifying some common pitfalls in this field along with some recommendations for future research. The talk also focuses on the observation that researchers working in the CA and cryptography communities often tackle similar problems, although under different terminologies. Hopefully, acknowledging and discussing this fact could help bridging the two research communities, and open up possibilities for future cross-community collaborations.