liberal academia is doing the wrong thing by emphasizing the importance of language & word choice. 

Claim: why do i think that it is bad when someone calls me a faggot? The fact that my feelings get hurt is secondary to the fact that the person is revealing their mindset of bigotry & high propensity to do active harm to my material interests. 

^That one I’m almost ready to fight about. The following idea is a newer thought that I’m still workshopping:

TL;DR: Keep working to understand minds better, and care about them enough for that understanding to inform your decisions. Keep getting better at thinking and doing. Keep working to own and fix your flaws and problems. These efforts will gradually reveal more and more things as not ethical. What remains is the best approximation of ethics you can have, so far.

Long answer:

My philosophical foundation for ethics is summarized in my "From First Principles" post. A couple relevant elaborations and responses to typical objections to those ideas can be found at:

Some posts about the Experience Axiom

Some posts about the Empiric Definition of Logic

Sadly some other relevant clarifications and additions are scattered across my thousands of posts. They should mostly or entirely be tagged with the relevant tags though.

Of course, this only takes care of the more philosophically challenging part of justifying the existence of a likely external reality and other minds. I have not written up concise posts for all the steps beyond that. I think the broad strokes are intuitive enough from there, not enough to be rigorous but to take whatever seems obvious from there and run with it. I think I wrote a couple posts exploring what I call "the first value judgement", but basically:

On the one extreme, once mind interferometry lets you see that other minds seem to exist with roughly comparable experiences, you can decide that it is arbitrary to a priori value any mind over any other, and so you give all minds and their experiences equal ethical weight based on the empirical observation that they are all roughly the same kind of thing.

On the other extreme, you have unapologetic selfishness, based on the empirical observation that you only experience your experiences, so everyone else is a different kind of thing than you. Even when evidence is pretty strong that we are all having experiences, it kinda requires a motivated axiomatic leap to say "experiences which I never experience should be given as much weight as my experiences".

A third extreme is self-excluding altruism - just like the second and unlike the first, this is rooted in the empirical difference between the self and all others, but here the focus is flipped - the axiomatic choice is made the other way: the singular experiencer gives itself special insignificance, valuing its own experiences less or not at all.

Obviously you can make this initial axiomatic choice by putting the boundary somewhere other than just self/others, or even by putting down multiple boundaries or a gradient for how different minds are weighed.

What's cool is that you can see every one of these in different humans' moral intuitions, which is very validating of all my bullshit up to this point because it all fits - that's exactly what you'd expect if I'm right about how human brains empirically struggle their way through something like the progression from raw experiences which I outline in "From First Principles", but with distortions and errors added in by personal and circumstantial variation and the need to do it with naturally selected heuristics more efficiently within the constraints we have or had in our life or in our ancestral environment.

What's even cooler is that for several years now I keep realizing that each of these must converge onto the same ethics as you integrate the future and factor in the constraints that we are subject to and all the messy stuff in life, so I don't think which of the above choices you make changes the ultimate destination as much as it changes your initial naive ethically clumsy fumbling until you mature. Like, you'll make different mistakes and need different epiphanies along the way, or some of the same mistakes in different order, and you might have different dead-end branches of ethical development you might go into for a while, but I think there is something inevitable about being forced towards the direction of a certain natural ethics. The caveat is that many people don't seem to get nearly the large enough range of experiences to get very far with that progress, and it is progress which has a ceiling based on your mental abilities and how much you put in the work. Which also includes things like growing coping skill and healing your traumas and insecurities and so on (otherwise certain things are more unbearable than others, and this naturally distorts our ethics).

To be clear, I think you can grow all of the relevant abilities. I don't like the idea of there being some biological predestined cap on how far you can develop your ethics. It's just that it takes time and effort which is not very rewarded or supported or subsidized socially. I have the privilege of having the time and energy to think about this bullshit pretty much all the time, and thus to keep training the thinking "muscles" almost as a side-effect. I had the privilege of being raised by smart and fairly mature and compassionate people who kinda managed to cause or force me to develop certain mental skills. So even though I think wisdom and pragmaticism and uncertainty seem to combine to create pressures for ethics to look a certain way, because they tend to make lots of choices worse than other choices regardless of your base values, I think in practice at any given moment we're all capped in one way or another from seeing all those considerations all the way to the end, or even far enough to see if it's going to be a false or dead-end path.

if you can do a bad thing that is beneficial to you, and you know you will not experience any negative consequences for it, you shouldn't.

It would've been nice and convenient if such an argument existed. I certainly wanted such an argument to exist for most of my life, and I pursued it pretty eagerly. Humanity has poured much motivated reasoning into finding one. But at this point I think it is misguided to try.

You're not building that up from first principles, that's for sure. It requires axiomatically chosing some value judgements as the right ones out - arbitrarily, out of all the other conceivable ones.

Of course, in practice there is no such thing as absolute certainty that there will never be any bad consequences. But the probabilistic rephrasing is also seemingly impossible to argue to a mind that does not already share relevant values with you.

Cognition habitualization: Most cognition can be trained by repetition to be more automatic, less effortful, and eventually less conscious; or untrained by finding ways to either avoid triggering that cognition or to redirect to other cognition, until it becomes less automatic and more effortful. This is well-established in psychology and neurology, but I figured out how to do this on my own as a little kid, so I've seen first-hand that you can take this much further than most people realize. Brains do what they train.

Cognemes: Cognition can be thought of as being made up of pieces that are subject to natural selection, and this is useful for understanding some things in the world and in human experience. There are at least three interacting dimensions of natural selection which are very relevant to us - in biology, in culture, and in the minds - each with different "selectable pieces" - genes, memes, and cognemes.

Cognetic openings: We can end up with arrangements of habitualized cognition that seem fine to us and normally work fine and as expected, but which can go wrong or motivate the wrong action with just a tiny change - such as a single cogneme added or missing at the wrong moment. A lot of problems in people's cognition and choices can be seen as cognetic openings to some cognemes causing bad consequences, and some problems can only be seen with that lens.

The empiric definition of logic: Philosophically, this elegantly frees logic from depending on any arbitrary axioms, and enables what I think is the most sound and justifiable "from first principles" build up to any other right idea. Psychologically, it reveals that our sense of logic is also at least in large part empirically developed. For me, this idea was key to even conceiving of countless other ideas.

Prediction trees: Our brains - and in particular almost all of our emotions, motivations, and reactions - can be explained almost entirely by the brain learning to treat some things as predictive of other things, and then reacting to those predictions preemptively and automatically, before or entirely without any critical review of whether those predictions are accurate in the current situation.

Decision shares: If we iteratively improve on democratic governance and try to solve problems with it, or optimize a benevolent non-democracy that cares about its people, or take money further as a mechanism of coordination, they all converge on a way that we can do thoroughly better as a society in almost every way imaginable.

There is concilience with it in the form of my thoughts about overlapping and composite minds, and my thoughts on multiplicity, and my thoughts about how minds are more of a label we draw over... "concentrations" of cognition and the stuff that hosts/executes it, and the bit in the mind interferometry post about how all we can ever have evidence of is cognition processing inputs.

It also feels promising for making per-mind ethics trivial to philosophically justify from first principles. Maybe other things too.

Utilitarianism is unique in a very important way.

Every other approach to ethics requires taking on additional axioms beyond the experience axiom.

The experience axiom is inescapable - we have to take something about our raw experience as axiomatic because no amount of doubting it will negate the experiencing itself.

All other axioms are arbitrary and unjustified until proven necessary.

The idea that preferred/pleasant experiences are ethically more valuable than less preferred/pleasant experiences also looks like an arbitrary axiom if we just look at it from our normal, day-to-day, higher-level perspective.

But if we build from the ground up it is qualitatively less arbitrary/axiomatic: raw experience itself already provides that value function. Defining ethics on top of it is just going with what is already inescapably there. It makes sense precisely because any other choice would be more arbitrary/axiomatic.

Of course, the only thing our raw experience stream provides is a solipsistic value function - giving us utilitarianism in a world where the only mind is you.

It was possibly very sweepingly wrong to say that utilitarianism is unique among ethical systems in not requiring additional axioms.

It would definitely be very sweepingly wrong to consider that conclusion correctly reached by just what I have written on this blog so far.

It is still possible that it is a correct statement, I simply do not have vivid vision of a comprehensive end-to-end justification for it.

I know I have thought this through much more in the past, but I have only vague memories of the arguments and conclusions I reached.

It is entirely possible that when I wrote "Solipsistic Utilitarianism" I was actually channeling a cached idea which I had simply forgotten disproving, due to not having done enough mental passes over the disproof to lastingly update my thinking about it.

There is a reason why "From First Principles", which was written when I was much more careful about not saying possibly wrong things, ends precisely where it does.

Because the distance between

sketching out a rigorous foundation for ideas like logic, probability, and theory of mind, and

proving sweeping claims about all ethical systems

I assert that the minimum axioms need to justify anything justifiable is just the single experience axiom. Then:

Make some tedious pedantic definitions to talk about the experiences and empirically define logic.

Derive notions like probability, etc, now that you have logic empirically built on the idea of correct predictions.

Infer that other minds probably exist, because some objects in our experience stream are best explained and most reliably predicted by supposing they have experiences/cognition too.

Observe that some experiences are more or less positive/pleasant/preferred than others to us.

Note that preference and pleasure are only definable in terms of each other and boil down to the same key facets of experience/cognition.

Observe that this qualitative relative positivity or negativity is part of the experiences themselves and thus already included in the experience axiom.