In the old days the owners manual stated that if you took short trips to change the oil every 3000 miles and if you took long trips 7500 miles (if in dusty or fleet conditions change it more often). That was a hard thing for an owner to monitor accurately. The GM Oil life monitor, monitors the length of your trip, oil temp, engine RPM, coolant temp, etc using an algorithm and it makes a better determination of when to change the oil. Plus oils today are so much better than they were AND fuel injection does not contaminate the crankcase with flooding/fuel like carburators and chokes did. So to change your oil at 3000 miles is wasteful and un-necessary. Read about your OLM here:
http://www.cadillacforums.com/cadillac-tech.html#oillife
I can NOT claim credit for the following, it was written by a very knowledgable person, I had saved it in a WORD doc on my hard drive, this SHOULD clear up any apprehensions you have about the OLM:
The GM engine oil life monitor counts engine revolutions and accumulates the number for the basis of the oil life calculation. It then adds deterioration factors for operating temperature, start up temperature, soak times, ambient, coolant temperature, etc... There are a LOT of factors that "adjust" or affect the slope of the deterioration but the fundamental deterioration is traced back to the ZDP depletion that is inescapable with engine revolutions. The specific rate of ZDP depletion is readily measurable for any given engine so that is the fundamental item that is first calibrated for the oil life algorithm to tailor it specifically to that engine.
You would obviously like to get the oil out of the engine before the ZDP concentration gets so low that it is ineffective at being at the right place at the right time and preventing engine wear so that becomes the long term limit on oil life for that application.
The other things that determine oil life such a acid build up, oxidation, petane insuluables such as silicon from dust/dirt, carbon or soot build up from the EGR in blowby, water contamination, fuel contamination, etc.... are all modeled by the multipliers or deterioration factors that "adjust" the immediate slope of the line defined by the engine revolution counter as those items can be modeled in other ways and accounted for in the immediate slope of the ZDP depletion line.
The algorithm was developed over the course of many years by several lubrication experts at GM Fuels and Lubes, spearheaded by Doctor Shirley Schwartz who holds the patents (with GM) for the algorithm and the oil life montitor. I had the luck of working directly with Dr. Schwartz when the idea of the oil life monitor first progressed from the theoretical/lab stage to real world testing/development/validation. There were fleets of cars operated under all conditions that deteriorate the oil life for any and every reason and , thru oil sampling and detailed analysis of the oil condition, the algorithm was developed, fine tuned and validated to be the most accurate way invented yet to recommend an oil change interval by. As just one example, I have seen cars driven side-by-side on trips, one towing a trailer and one not, for instance, to prove the effectiveness of the oil life monitor in deteriorating the oil at a faster rate just because of the higher load, higher average RPM, higher temps, etc...and it works flawlessly.
The oil life monitor is so effective because: it is customized for that specific vehicle/engine, it takes everything into account that deteriorates the oil, it is ALWAYS working so as to take into account THAT INDIVIDUALS driving schedule, and it tailors the oil change to that schedule and predicts, on an ongoing basis, the oil life remaining so that that specific individual can plan an oil change accordingly. No other system can do this that effectively.
One thing is that I know personally from years of testing and thousands of oil analysis that the oil life algorithm works. There is simply no argument to the contrary. If you don't believe me, fine, but, trust me, it works. It is accurate because it has been calibrated for each specific engine it is installed on and there is considerable testing and validation of the oil life monitor on that specific application. NOt something that oil companies do. They generalize....the oil life monitor is very specific for that application.
Oil condition sensors in some BMW and Mercedes products are useful, also. They have their limitations, though, as they can be blind to some contaminates and can, themselves, be contaminated by certain markers or constituents of certain engine oils. Oil condition sensors can only react to the specific oil at that moment and they add complexity, cost and another potential item to fail. One other beauty of the GM oil life monitor is that it is all software and does not add any mechanical complexity, mass, wiring or potential failure mechanism.
There is considerable safety factor in the GM oil life monitor. Typically, I would say, there is a 2:1 safety factor in the slope of the ZDP depletion curve....in other words, zero percent oil life per the ZDP depletion is not zero ZDP but twice the concentration of ZDP considered critical for THAT engine to operate under all conditions reliably with no wear. This is always a subject of discussion as to just how low do you want the ZDP to get before the oil is "worn out" if this is the deciding factor for oil life. We would tend to be on the conservative side. If the oil life is counting down on a slope that would recommend a 10K change interval then there is probably 20K oil life before the ZDP is catostrophically depleted....not that you would want to go there...but reason why many people are successful in running those change intervals.