News IBM Ups Quantum Roadmap for Scalabity, Modularity, K-level Qubit Counts

Just in case there is a theoretical physicist who is reading this , I'm going a bit off-subject at first but I'll bring it back to quantum computing . I used to think that we could circumvent the limitation of faster-than-light communications issue by using entangled quantum particles. As you may know, entangled particles communicate with each other instantaneously without regard to distance. We change the spin on one of the entangled particles on Earth, and someone on Neptune with the other of the entangled pair sees the spin change on his particle instantaneously. Expand on this concept and soon your sending 1's and 0's back and forth and we have instantaneous communication. But any learned physicist (as well as Wikipedia) says "No, this will not work because reading the spin may actually change the spin and blah, blah, blah... science". Okay fine. I'll accept that without fully understanding. Then how is it that we can have entangled quantum particles as a key component to quantum computers from which we read values? Wouldn't the faster than light entangled particles be a much simplified version of a quantum computer with a count of just a couple of "useful" qubit pairs? What am I missing?