All organisms rely on noisy molecular recognition to convey, process and store information. This stochastic biophysical setting poses a tough problem: how to construct information processing systems that are

efficient and yet error-resilient? – The challenge of molecular recognition is critical for the ribosome, the molecular engine of self-replication: In order to synthesize proteins, ribosomes have to select the correct building blocks from a large pool of similar substrates, and inaccurate or inefficient selection might be devastating to the organism. We will discuss the performance of the ribosome in this task and the possible role of its large conformational changes. The analysis suggests a generic mechanism, and we will examine it the context of other molecular recognition systems, such as homologous recombination or basically any enzymatic process in the presence of competing substrates. Finally, we will briefly mention some of challenges and possibilities of the present approach.