A 25% chance to receive $1 is worth something. A 1% chance for $2 is worth something too, as is a 100% chance of $0.01. Everyone here can work out which of these is worth more than others, even though there's a strong chance that any of these bets could "turn out" worthless.
The problem is that options are harder to value: there are many more result scenarios and the math is a little more complicated. As a result, many people opt out of doing the math entirely, and they think either "it's basically worthless" or "it's basically money in the bank."
The latter error is much worse than the former, but here are two consequences of the former: (1) one is less likely to work at a startup which might (in reality) be the best choice, and (2) if one is working at a startup, one's less likely to be nearly as alert to the risks and targets that may make the option worth less or more (at least as a probabilistic matter). The latter is why the options are a win-win for the company and employee.
By the way, I do understand that different people have different tolerances for risk and for different kinds of risk. But even if someone is biased very heavily towards low-risk situations (which is perfectly fine), it's best to do the math, understand the risk/reward profile of the option and then apply that against your own risk tolerance.