The answer to this question can be found in Chapter 5 of Guns, Germs, and Steel. There, Diamond discusses a number of ways in which this transition differed from our assumptions.
First, Diamond says that farming would not necessarily have been better than hunting and gathering. We tend to assume that farming is much easier than hunting and gathering, but that is because few of us farm, and even our farmers do not have lives as hard as early farmers would have. Diamond says (on p. 104 and 105) that “most peasant farmers and herders … aren't necessarily better off than hunter-gatherers.” This goes against our assumptions.
Second, Diamond says (on p. 105) that we assume that farming was discovered or invented all at once. We believe that people figured out how to farm and decided to do so. According to Diamond, this is incorrect. Instead, hunter gatherers gradually did new things that slowly evolved to the point where they found themselves farming. They did not consciously decide to start farming and they did not transition to farming all at once. It was a slow evolution that they probably did not really even notice all that much.
Third, on p. 106, Diamond says that we assume that there is “necessarily a sharp divide between nomadic hunter-gatherers and sedentary food producers.” Again, Diamond says this is not true. Instead, he argues that many societies are or have been hybrids. There have been sedentary hunter gatherers and there have been nomadic farmers. These are not clear cut boxes that we can put societies into. Instead, there can be many different economic systems with various mixes of farming and hunting and gathering.
In these ways, Diamond says, the transition to farming did not really happen in the way that we think it did.