# 4.13 Is the Probability of Collapse Really 95%+???

J: This analysis seems to put the probability of a benign outcome fall lower than 5%. What would have to be true in order to get the probability up to your high estimate 5%, and what would be the approximate math behind it?

AI: To raise the probability of a benign outcome to as high as 5%, several significant changes in assumptions would have to be made. Here are some possibilities:

1. Higher Individual Probabilities: Assume that each of the 20 risk factors has a much higher chance of not leading to societal collapse, say 90% instead of 50%.
2. Fewer Risk Factors: Assume that some of the 20 risk factors are not actually risks, reducing the number of factors considered.
3. Positive Interdependencies: Assume that mitigating one risk factor significantly reduces the risk of others, creating a beneficial feedback loop.
4. Effective Interventions: Assume that highly effective interventions are implemented for each risk factor, significantly reducing their potential impact.

Let's consider the first assumption, where each risk factor has a 90% chance of not leading to collapse. The probability that all 20 factors turn out benignly would be 0.920≈0.12160.920≈0.1216 or about 12.16%. This is above the 5% threshold.

For the second assumption, if we consider only 10 risk factors instead of 20, each with a 50% chance of not leading to collapse, the probability would be 0.510≈0.00097660.510≈0.0009766 or about 0.09766%, still far below 5%.

For the third assumption, introducing positive interdependencies would require a complex multivariate model, but it's reasonable to assume that if mitigating one risk factor could halve the risk of another, the overall probability of a benign outcome would increase. However, quantifying this increase would require detailed modeling.

For the fourth assumption, if effective interventions could raise the individual probabilities to 90% for each of the 20 risk factors, we would be back to the first scenario with a 12.16% chance of a benign outcome.

In summary, to reach a 5% probability of a benign outcome, we would need to make very optimistic assumptions about the individual probabilities of each risk factor and their interdependencies. Even then, reaching a 5% probability would likely require a combination of these optimistic assumptions.

J: From a highly oversimplified perspective, that seems reasonable. What probability do you think there is that some combination of those adjustments might be True, and that there is in reality a 5% chance that everything is ok?

AI: Assigning a probability to the likelihood that a combination of those optimistic assumptions might be true is a complex task. It would require not just evaluating each assumption independently, but also considering the interdependencies and potential synergies between them.

However, if we were to make a rough estimate, we could consider the following:

1. Higher Individual Probabilities: The likelihood that we've significantly overestimated the risk of each individual factor is probably low, given the existing scientific and sociopolitical evidence. Let's generously assign a 20% chance to this being the case.

2. Fewer Risk Factors: The chance that we've misidentified some factors as risks when they are not could be higher. Let's assign a 30% chance to this.

3. Positive Interdependencies: The chance that mitigating one risk significantly reduces others is plausible but not guaranteed. Let's assign a 25% chance to this.

4. Effective Interventions: Given the current state of global governance and the scale of the challenges, the likelihood of universally effective interventions being implemented is low. Let's assign a 10% chance to this.

If we assume these are independent (which they likely are not, but this simplifies the calculation), the chance that any one of these is true is:

However, for the 5% benign outcome probability to be true, we would likely need a combination of these factors, not just one. The chance that all four occur simultaneously would be:

0.2×0.3×0.25×0.1=0.00150.2×0.3×0.25×0.1=0.0015

or 0.15%, which is extremely low.

So, while there's a reasonable chance that we might be wrong on one or more assumptions, the likelihood that a combination of these optimistic assumptions could result in a 5% chance of a benign outcome is very low, likely well below 1%.

Forward to 5.1 The Limitations of Political Systems
Back to 4.12 The Opposing Argument