Coffeinated

The idea behind the circuit is that it acts like a current regulator here. What you have described is the LM317 as a voltage regulator, and correctly so, bad that is a bad idea for laser diodes - most often, the U-I-curve isn't known and choosing a correct resistor is complicated, to say the least. Generally, a current source is considered best for Laser Diodes and normal high power LEDs, this circuit just doesn't cut it completely.

That laser driver is a really bad circuit. Two concerns:1) You will need a poti that can handle the high currents going to a laser (1 A for me)2) When the transistor is switched off (no current flowing), the LM317 is now "floating" on 12 V, with no connection to ground. When you switch the transistor on, a connection to ground will be made, thus connecting your laser diode to 12 V directly - for a very short time, but that will degrade your laser over time or pretty fast. Always try to switch off the circuit above the LM317, so it is not connected to 12 Volts anymore. This way, you can ensure that the LM317 is working before it applies 12 Volts to the laser diode. A capacitor in parallel is no bad idea either, so that it can take any current / voltage spikes.