Let's consider the following scenario: 1. There is a group of people who has 99% of chances that they are going to be the first to create superintelligence. 2. Let's suppose hypothetically that superintelligence gives you 100% control of the world for the rest of the remaining time (which is far from true, but let's assume that). 3. They have 1% of chances that they lose and someone else will create superintelligence before them. 4. Let's suppose that there is 1% chance that people will become close to immortal, as a result of improving technology. Let's suppose that the company has a chance of joining the equality norm that requires them to distribute resources equally with the rest of the coalition. In exchange, the rest of the coalition (which has 1% of winning the superintelligence race) will distribute the resources with them, if they win. Would the group of people that has 99% of chances of creating superintelligence first be better off joining the coalition? The answer is yes, because... Simple answer: 1. If they don't join and they win, they will be in a great situation. 2. If they join the equality norm, they will also be in a great situation. In the long-term, their situation will be only minimally worse than in case of being the sole winner. 3. If they don't join and they lose, then they will be in a terrible situation for a very long time. 4. So, it's better to join. Mathematical answer: 1. If they don't join, their expected utility is equal: $0.99 \cdot log_b m + 0.01 \cdot 0$ which can be simplified to $0.99 * log_b m$, where $m$ is the number of resources that they will acquire through the entire existence. 1. Note: if they don't join the coalition, it will end with winner takes all scenario. For simplicity, I don't count what will happen before that winner takes all scenario manifests because if we assume that people will be close to immortal, then what will happen before the time when there is only one winner will be a very small fraction of time and is almost irrelevant. That's why, in the above formula there is $0$ for the case if they lose. 2. If they join, their expected utility is equal: $log_b \frac{m}{n}$, where $n$ is the number of people in the coalition. 3. Now, if we assume that they will exist for close to infinity, then $m$ is super large because the number of resources will grow exponentially with time (and time is almost infinity). 4. For a sufficiently large $m$, it is true that $log_b \frac{m}{n} > 0.99 * log_b m$. And $m$ is super high, in case of people becoming close to immortal. Therefore, if we assume that there is 100% chance that people will become close to immortal, then it's better for the group of people to join the norm. 5. If we assume that there is 1% of possibility that people will become close to immortal, then that case will dominate other cases because it will have significantly more long-lasting consequences, therefore it is also better for the group of people to join the norm, even if it's just 1%.