What are the best examples of significant harm caused by "infohazards"? (don't say Roko)
I tend to internally roll my eyes when I hear this concern, often from EA-land, and I'd like to inoculate myself against that if appropriate.
Conversation
If you believe -style optimism, I’m pretty sure you have to reject the concept of infohazards outright.
Any “hazard” caused by info can be triumphed by more knowledge
Blocking info as “hazards” instead is working against the unbounded growth of knowledge
1
4
I get what you're saying, but not sure I'm willing to bite that bullet. Path dependence does still matter: the amount of marginal knowledge needed to correct the hazard is potentially unbounded; you do care about harm done in the meantime.
I get that it is not possible in all cases to devise a policy which both minimizes interstitial harm without blocking unbounded future knowledge growth. But I guess I view that like non-decidability problems in CS: intractable in the general case; often tractable in the specific.
2
3
Yeah :) also just reminds me of the “greater good” moral argument.
Maybe you really can temporarily block knowledge growth (or make some sacrifice) to the benefit of all, but it’s risky territory.
Definitely fair! I think that principle can be more theory than practice
But it does give me pause whenever I hear the term. The info can be a temporary hazard, but no info is inherently forever evil.
1
1
Yes, I think that's part of why I tend to react negatively when it's mentioned in effective altruism contexts.
1