Good points, I think I know what you are trying to say. My idea is incomplete or perhaps wrong. Perhaps having a representation of the attention becomes important, but if so, why? I'll have to think further about this.  Thanks

1. Sorry, I don't see the mixing up of two senses of mattering and don't understand your example.
2. Good point about weighing preferences. Yes, you can have a multidimensional goal function, as a landscape, where different directions (preferences) have different slopes (strengths). (Picture a consumer choice diagram used in microeconomics, with indifference curves.) But in this context, with strength I ment the amount of amplification of a signal. The preference to avoid a strong pain is stronger than the preference to avoid a slight pain, because the pain signal of the strong pain is more strongly amplified to get my attention. 

Perhaps a good analogy: some animals are not able to run fast on their legs to escape from predators. They have a physical limitation. Due to that limitation, those animals developped wings with which they can fly and escape from predators. Does it seem weird to suggest that the reason a bird can fly is because of its physical limitation to run fast? 

Ha, nice argument :-) It reminds me of a famous defense of predation argument in animal rights ethics: lions do not violate the rights of their prey, because lions do not have moral agency. In other words: lions have a privilege to harm, kill and use other animals, because they are too stupid to understand morality. 

To the point: although it may seem weird, I think it makes sense: if due to cognitive limitations an organism was forced to generate an attention schema that generates consciousness, that organism is more concious than e.g. an AI that does not have those cognitive limitations that cause the development of an attention schema. Furthermore, in terms of the ability to pay attention, that unfocused AI (i.e. it is not able to focus), has a cognitive limitation. It lacks a cognitive ability. Although I consider it possible and even likely that a strong AI also develops an attention schema, because that is so beneficial and efficient. Our brains are also able to run processes in parallel, devoting resources to many (unconcious) parallel processes. Yet, we benefit from having the cognitive capacity to focus, to pay attention, using an attention schema. Hence, even with our cognitive ability of parallel processing, we developed an attention schema.

Veganism is (or can be) based on animal rights, which requires a moral deontological ethic. See also https://stijnbruers.wordpress.com/2019/08/22/carbon-offsetting-versus-meat-offsetting/

"Given the complex interdisciplinary nature of societal issues, studying the basics of economics might make you overconfident that you can solve societal problems.
Take, for example, supply and demand. The standard supply and demand model will tell you that having/increasing the minimum wage will automatically increase unemployment. But if we look at actual empirical evidence it shows us that it doesn't. Overrelying on simple economic models might mislead us about which policies will actually help people, while a more holistic look at the social sciences as a whole may counter that."

I disagree with this paragraph. Economists are trained to NOT stay at the basics of economics and NOT overrely on simple economic models. They develop more complex models. I rather have the impression that many people in other social science fields, when they deal with economics issues, rely too much on basic economics with its overly simplified models. 

Note that economists - not sociologists, psychologists or anthropologists - came up with the empirical evidence that minimum wages barely increase unemployment. All your references are published in economics journals. And note that economists - not sociologists, psychologists or anthropologists - developed more complex models that explain why the competitive market equilibrium model with labor supply and demand was too simplistic.

These are valid points. I wanted to see how much time it will take me to find 10 promising, high-impact petitions. I could easily find 7 very good petitions, but the final 3 were harder to find. Overall it took me a few hours.

I partially agree with your concern about the shrimp petition lacking a ToC, although that petition still contains a clear, feasible and well-targeted ask in my opinion. But it could be that such change.org or class project petitions lack credibility and hence are less effective.

If people spend more time researching the issues, I consider that as a positive side effect. People learn about the causes. I bet people learned about AI safety by seeing me recommending those petitions. It might result in people supporting those causes in other ways in the future, for example by donating money. Especially if the petition comes from a highly effective charity. Any way, I still consider signing petitions as the cheapest altruistic thing that one can do, especially if you have a platform that offers you the most effective petitions so you don't have to look for them yourself.

Nice that you've made that website. I also had in mind a website or online platform that regularly offers you a few top-effective petitions to sign (and perhaps a newsletter that informs people when new petitions are available). The most difficult part may be regularly looking for and selecting those petitions. As I said, quickly finding 10 petitions was already a challenge, and I don't think I can easily find 10 other petitions next month. But looking for a few petitions every few months should be feasible for me in my spare time. And if more and more EA-aligned people and organizations inform me/us about new petitions, it becomes even more feasible. And if you and other effective altruists also contribute to this website by looking for effective petitions, we may have something fruitful. 

As petitions are such a low bar in terms of altruistic engagement, I also don't think it is worthwhile to have a team of (professional) effective altruists spending much time assessing petitions. It's not such a big deal, in my opinion, that a weakly tractable petition such as that shrimp petition gets selected. More problematic would be if we miss a highly effective petition. False negatives (not recommending top effective petitions) are more problematic than false positives (recommending weakly effective petitions), because signing a petition doesn't take much time and there are not many top effective petitions. There are petitions that are counterproductive and are negatively effective, but I think we are able to quickly recognize and not select those. 

You raised a good point. Yes, I guess I agree that when there is only a positive experience and no negative, the welfare is definitely positive, even if the positive experience is very small. But thinks get tricky when there are both positive and negative experiences, as is the case for almost all sentient beings, and probably also for nematodes if they are sentient. The more welfare is composed of positive and negative parts, the more difficult it becomes to compare it with a zero welfare level. Might have to do with information processing capacity. Adding up many positives and negatives is more difficult that considering a single positive or negative value. Evaluating mixed experiences (with both positive and negative parts) might require a more coarse-grained approach. The level of coarse-graining might relate to the neutral range: the more coarse-graining is used, the wider the neutral range. 

Ok, let me exaggerate a bit. Assume when state S=(X,Y,Z)=(87455.668741, -258.142567, -11024.441253), you are indifferent with nonexistence. Now consider state S'=(87455.668741, -258.142567, -11024.441153). You can confidently say that S' gives you a positive welfare? If yes: close your eyes and write down, for the given X and Y of state S, a value of Z that gives you a positive welfare lower than S'. I bet your brains are too small to do this exercise. Now consider a nematode with much smaller brains....

Perhaps we should run a survey, ask people if they have a neutral range. Can they give values of X, Y and Z such that if they would experience X, Y and Z units of some welfare determining components, they would be indifferent between that experience and non-existence, whereas if they had X+dX, Y and Z units, they would state a positive welfare and X-dX, Y and Z units would correspond with a negative welfare. I'm personally very skeptical that most people's neutral ranges are zero. You claim to have a zero neutral range?