Evolution-Driven Ethical Progression

On How to Live, What to Consider Right, What Old Societies and Super-AIs Are Like, and Why We Don’t See Them

Daniel Vallstrom

2015

NB: For the full text, see pdf.

Abstract

With an evolutionary approach, morality can be explained as adaptations to for example problems of cooperation. Reproducing AI that satisfies the conditions for evolution in a broad sense to apply will be subject to the same cooperative evolutionary pressure as biological entities. This, together with data of human progression, is used here to project that humans will become more and more cooperative and considerate, and that old societies and super-AIs are cooperative. (An ‘old society’ means here a society that has undergone more biological and cultural evolutionary changes than human society.) These projections, along with diminishing beneficial returns from material resources, suggest the possibility that societies and AIs become more and more cooperative and less and less aggressively expansive, which would explain the Fermi paradox, wondering where everybody is. It is also argued that it is likely that old societies engender, give way to, super-AIs, since it is likely that super-AIs are feasible, and fitter.

Appended is an algorithm for colonizing for example a galaxy quickly.

The Fermi paradox also tentatively suggests that certain of our approaches to AI research will be more fruitful than others, and that utilitarianism seemingly is wrong. There are also other ethical conclusions to be had.

There is also a summary discussion of traditional philosophy and ethics.

Conclusion of the Fermi Paradox Section

Looking at it from the other direction, we have all these observations and principles:

  • The Copernican or mediocrity principle

  • The equilibrium principle

  • The seemingly likely abundance of old societies

  • The observation that just a single society ought to be able to quickly explore and colonize e.g. our galaxy

  • The Fermi paradox

  • No evidence of non-benign old societies or super-AIs: our solar system hasn't been made into paper clips e.g.

  • Seemingly no Kardashev type III societies

This suggests the possibility that all old societies and super-AIs behave similarly in these regards, because of things they have in common, for example evolution and evolutionary game theory. A look at these common things, an extrapolation of human progression, and diminishing beneficial returns from material resources, indicate the possibility that societies and AIs become more and more cooperative and less and less aggressively expansive, which would explain the Fermi paradox.

Notes

[i] Caveat: The link leads to a draft where the chapter 12 epigraph is by climate contrarian Matt Ridley[R1][R2][R3][R4], and is arguably false. Following there is arguably also a related false dichotomy. (E.g. the sixth mass extinction is a catastrophe, arguably, and we ought to have started to take action against CO2 pollution a century ago, or at the very least half a century ago.[R5]) ^

[viii] To view multiple wave time series with World Values Survey's (WVS) online tool, choose e.g. the latest wave and then go to the time series tab (see instructions here under the heading "Time Series").

For a summary, see "Findings & Insights" at WVS, or https://en.wikipedia.org/wiki/World_Values_Survey#Insights or https://en.wikipedia.org/wiki/World_Values_Survey#Findings.

Below are some quotes from "Findings & Insights" at WVS:

"Norms concerning marriage, family, gender and sexual orientation show dramatic changes but virtually all advanced industrial societies have been moving in the same direction, at roughly similar speeds."

"Although a majority of the world’s population still believes that men make better political leaders than women, this view is fading in advanced industrialized societies, and also among young people in less prosperous countries."

"Since 1981, economic development, democratization, and rising social tolerance have increased the extent to which people perceive that they have free choice".

"Generally speaking, groups whose living conditions provide people with a stronger sense of existential security and individual agency nurture a stronger emphasis on secular-rational values and self-expression values." ("Self-expression values give high priority to environmental protection, growing tolerance of foreigners, gays and lesbians and gender equality, and rising demands for participation in decision-making in economic and political life.")

"With industrialization and the rise of postindustrial society, generational replacement makes self-expression values become more wide spread and countries with authoritarian regimes come under growing mass pressure for political liberalization."

Here is figure 2.5 from [20] showing Planckian progression:

WelzelFreedomRisingFigure2.5PlanckianProgression

Here is, to some extent, an other summary of WVS, and [20], by Jonathan Haidt. ^

References

[1] GiveWell, August 2015, "Potential risks from advanced artificial intelligence", http://www.givewell.org/labs/causes/ai-risk ^

[20] Christian Welzel, 2013, Freedom Rising: Human Empowerment and the Quest for Emancipation, https://www.researchgate.net/publication/259241027_Freedom_Rising_Human_Empowerment_and_the_Quest_for_Emancipation [ix] ^