This answer begs another question: how do the dispatchers know when a worker is needed or not? Some kinds of work need to be done all the time, some kinds need to be done when specific conditions are met, still others can be done except when certain things are happening. How do the dispatchers know which is which?

Life's Instruction Book has some special features about it that allow the dispatchers to decide which workers to make. The Instruction Book contains a note in the margin by each Instruction that marks the Instruction's beginning. These notes (called promoters) tell a kind of dispatcher called a transcription factor where to bring other dispatchers to make workers to carry out the Instruction. (The other dispatchers needed to make a worker are RNA polymerases and ribosomes. The RNA polymerase copies the Instruction into a message for a ribosome. Ribosomes read the messages to make workers that carry out the instruction. Details of how the dispatchers work -- and the reasons for their mysterious names -- are given in Making Workers.)

Promoters are not all equally good at attracting the attention of a transcription factor as it reads the Instruction Book. Those that almost always get noticed are called strong promoters, those that often get overlooked are called weak promoters. The different strengths of promoters provide living things some control over how many workers to make to carry out an Instruction. Strong promoters are noticed by almost every transcription factor that passes by them, so many workers are made for those Instructions. Few workers are made for Instructions with weak promoters.

Changing the numbers of workers that carry out an Instruction still doesn't quite control which work gets done when. That's the job of two other kinds of dispatchers, activators and repressors.

Activators try to attract more transcription factors' attention to promoters for work that needs to be done. Most of them work by pointing out those promoters to the transcription factors by pointing and waving, but some seem to go further by leading the transcription factors over to the Instructions they want them to copy. To make matters more complicated, not all of the activators are equally good at getting the attention of the transcription factors; some can shout louder than others. Transcription factors tend to spend more of their time making workers for Instructions with noisy activators.

There's one last dangling thread to tie off: how do the activators and repressors know when to attract attention? The Instruction that made them lets them sense a certain thing; it says says something like, "If you step in water on the floor, run over to the Book, stand by promoter #3148 and make lots of noise." Some activators and repressors don't sense things directly, but instead react to being handed notes by other workers that can. For example, let's say that there's a worker whose job it is to detect fires outside the factory. Rather than have him run to the Book when he sees a fire, his Instruction tells him to print up lots of notes that say "There's a fire outside!" and pass them around the factory. When the activators for fire-fighting Instructions get these notes, they run back to the Book and try to get the dispatchers to make the fire company. At the same time, repressors getting these notes try to shut down production of flammable substances. In this way, one worker can signal many activators and repressors to do a complex job. The process of passing messages from few workers to many workers is called a cascade.