Maybe the Atomic Energy Commission should have stuck with the 'light water' reactors.
That's one option, but the biggest issue with the AEC program was the focus on breeder reactors. Breeder reactors can run on plutonium and nuclear waste, which allows them to be burned off, creates power, and creates new fissile material for power reactors to turn off of. In the early years of nuclear power they greatly underestimated the amount of uranium in the world, so it was thought that the breeders would be vital for ensuring nuclear power could function.
We now know that there is a lot of fissile material for reactors to run off of, so much that we're talking about decades before it will be commercially viable to reprocess spent fuel from power reactors, and a few decades more for breeder reactors to become economical.
The AEC and its successors spent some $20 billion on breeder reactors from 1968 through to the early to mid-1980s, which consumed a massive amount of the budget. It's the largest single energy program in United States history. Spending those funds on more basic research would have given far more value.
And I think they're a classic case of a government agency having the awkward dual mission of both overseeing safety and promoting the industry.
The agency went through several splits and mergers in the 1970s. The first split was into the Nuclear Regulatory Commission and the Energy Research and Development Administration, a merger of several federal energy programs. Because the AEC was the largest energy research agency going into the merger, nuclear dominated at the ERDA. Over half the budget, programs, and personnel were carried over from the AEC, so it was really just a continuation of the research branch. Later the ERDA was reorganized into the Department of Energy, and other energy sources, as well as conservation, started to see large research dollars.
There's some argument that the split may have resulted in an NRC that took the Regulatory part of its name a bit too far. The regulations are written in such a way that you can't really build anything but a light water power reactor in the United States now (such as a gas cooled reactor, heavy water reactor, etc.), for example, or you can just add in unneeded equipment mandated by law.