Thank you very much. Both EA and Less Wrong are exceptional places. To some extent, the model of what Internet shall be, and we lost somewhere in 2010s. We will recover this open high quality Intenet at a large scale. Meanwhile we have this. 

I wrote this las year: 

https://forum.effectivealtruism.org/posts/uHeeE5d96TKowTzjA/world-and-mind-in-artificial-intelligence-arguments-against

Now, cheap optimization does not mean that economics is no longer interesting. Economists still have to develop interactive settings (let’s name them “games”) and compare the equilibrium (=ergodic distribution) properties of the game under consideration with some real world phenomenon of interest. 

This introduction has a single goal: to direct the attention of the reader into a critical detail in the development of Artificial Intelligence: intelligence is not only a property of the optimizing kernel of a system: it emerges in the interaction between optimizing agent and the world it inhabits. 

In a world with Zeros, programming is simply the design of interesting worlds.

More here:

https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4971482

"Evidence" about consciousness is always a red flag. The whole point of consciousness is that we observe ours (that is all we observe, by the way) and postulate that of others. At the end the fundamental intuition about consciousness is that you need neural complexity (information integration) and insects has a modest number of nodes in their neural network. 

https://www.lesswrong.com/posts/3FEz94Hs5dEsgRJYg/arthropod-non-sentience

"The (Darwinian) reinforcement learning process that has led to our behavior imply strong rewards and penalties and being products of the same process (animal kingdom evolution), external similarity is inevitable. But to turn the penalty in the utility function of a neural network into pain you need the neural network to produce a conscious self. Pain is penalty to a conscious self. Philosophers know that philosophical zombies are conceivable, and external similarity is far from enough to guarantee noumenal equivalence."

At some point you really to engagge with the problem of consciousness noumenality. The fact that (so far, probably forever) we can only postulate consciousness not measure it. 

I leave here my two posts about insect consciousness:

https://forum.effectivealtruism.org/posts/3nLDxEhJwqBEtgwJc/arthropod-non-sentience

and about Naturalistic Dualism:

https://forum.effectivealtruism.org/posts/5zbmEPdB2wqhyFWdW/naturalistic-dualism

If fish move across properties, your own overfishing affects fish density in neighbouring properties. It is like two oil wells extracting from a common reservoir. Of course, both "privatization" and "Coasian bargaining" are better than Pigouvian taxation, but none of this mechanism is necessarily available.

I remind you the entire sequence: 

1. Theorems of welfare economics: under the hypotheses of the theorem all Pareto optimal outcomes can be obtained by market clearing and [tailored] lump sum taxation. Unfortunately, to "lump sum" tax you need private information on productivity, so the best you get is "pareto optimal" with the usual deadweight loss of income taxation.
2. Property is not perfect: there are externalities. Then you try to use "Coasian bargaining", for small cases where externalities and property is well defined.
3. Multilateral bargaining is too complex, or property rights are not easy to establish: Pigouvian taxation on the externality as long is easy to measure and you have some sovereign to impose it.

Now, a funny thing is that on one side you complain on the lack of Pigouvian mechanisms, and then even for the canonical case of the carbon tax, soon you find arguments against it (!). Yes, of course, the total value of the externality is the world average impact of carbon emissions: there are winners and losers. The consensus based on detailed simulation is that the global externality of an additional molecule of CO2 is negative (at least given the current location of human population: given how costly and destabilising is large human reallocation, better not to remove that hypothesis). 

"Instead of setting up regulations to prevent overfishing, the oceans could be privatised, and then the companies owning them would have an incentive to prevent the collapse of fish stocks"

If you do not put physical barriers, fish would move across different properties, making overfishing profitable anyway. It is like two "private" oil fields over the same oil reservoir.

"I think global warming may well be beneficial in many regions. However, at least for countries wanting to decrease it, I suppose taxing CO2eq would make sense". 

It is the canonical case for an immediate Pigovian tax: the externality is global, uniform, circulates in the atomosfere...  Regarding imports, you can charge a carbon tariff.

I find this criticism not so good in general, because there are many externalities and "measuring" them means nothing. To some extent an externality is simply "what the market does not measure for us", so Pigovianism is more a framework than a theory.

On the other hand, the lack of Pigovian taxes on carbon (the canonical case where the framework is almost a theory by itself) and the incredible roundabouts to avoid the simple and well known solution proves the utter disgrace that are our social systems.

Intelligence is the only chance of some redemption to the massive suffering probably associated to the emergence of consciousness. 

This is the age of danger, we are the first species on Earth that has figured out morality, so we shall survive at allmost all cost. 

All those proxies tell us they have the wires to feel the pain. But what abour the self? You need the side of penalty and the side of self to have real pain. Pain shall inflicted to a conscious mind.

With their ridiculously small brains, how likely is a self on the receiving side of penalty?

[I post here the same than in the original Asterisk Magazine article [https://asteriskmag.substack.com/p/yes-shrimp-matter/comment/96031154]

I completely disagree: their brains are very simple neural networks, and their degree of consciousness is in the same range as electronic devices.

https://forum.effectivealtruism.org/posts/3nLDxEhJwqBEtgwJc/arthropod-non-sentience

All arguments based on behavioral similarity only proof we all come from evolution: "we are neural networks trained by natural selection. We avoid destruction and pursue reproduction, and we are both effective and desperate in both goals. The (Darwinian) reinforcement learning process that has led to our behavior imply strong rewards and penalties and being products of the same process (animal kingdom evolution), external similarity is inevitable. But to turn the penalty in the utility function of a neural network into pain you need the neural network to produce a conscious self. Pain is penalty to a conscious self. Philosophers know that philosophical zombies are conceivable, and external similarity is far from enough to guarantee noumenal equivalence."

Now, regarding how much information is integrated, supperativity implies that the ammount of resources devoted to the shrimp shall be propotional to their number, but (at most!) to their brain mass:

"As a rule, measures of information integration are supper additive (that is, complexity of two neural networks that connect among themselves is far bigger than the sum of the original networks), so neuron count ratios (Shrimp=0.01% of human) are likely to underestimate differences in consciousness. The ethical consequence of supper additivity is that ceteris paribus a given pool of resources shall be allocated in proportion not to the number of subjects but (at most!) to the number of neurons. "