avturchin's Comments

ACDT: a hack-y acausal decision theory
But it would be similarly convenient to have uncertainty about the correct decision theory.

Yes, this is really interesting for me. For example, if I have the Newcomb-like problem, but uncertain about the decision theory, I should one box, as in that case my expected payoff is higher (if I give equal probability to both outcomes of the Newcomb experiment.)

Since figuring out human values is hard, what about, say, monkey values?

There is a couple of followup articles by the authors, which could be found if you put the title of this article in the Google Scholar and look at the citations.

Since figuring out human values is hard, what about, say, monkey values?

Yes. See:

Mammalian Value Systems

Gopal P. Sarma, Nick J. Hay(Submitted on 28 Jul 2016 (v1), last revised 21 Jan 2019 (this version, v4))

Characterizing human values is a topic deeply interwoven with the sciences, humanities, art, and many other human endeavors. In recent years, a number of thinkers have argued that accelerating trends in computer science, cognitive science, and related disciplines foreshadow the creation of intelligent machines which meet and ultimately surpass the cognitive abilities of human beings, thereby entangling an understanding of human values with future technological development. Contemporary research accomplishments suggest sophisticated AI systems becoming widespread and responsible for managing many aspects of the modern world, from preemptively planning users' travel schedules and logistics, to fully autonomous vehicles, to domestic robots assisting in daily living. The extrapolation of these trends has been most forcefully described in the context of a hypothetical "intelligence explosion," in which the capabilities of an intelligent software agent would rapidly increase due to the presence of feedback loops unavailable to biological organisms. The possibility of superintelligent agents, or simply the widespread deployment of sophisticated, autonomous AI systems, highlights an important theoretical problem: the need to separate the cognitive and rational capacities of an agent from the fundamental goal structure, or value system, which constrains and guides the agent's actions. The "value alignment problem" is to specify a goal structure for autonomous agents compatible with human values. In this brief article, we suggest that recent ideas from affective neuroscience and related disciplines aimed at characterizing neurological and behavioral universals in the mammalian class provide important conceptual foundations relevant to describing human values. We argue that the notion of "mammalian value systems" points to a potential avenue for fundamental research in AI safety and AI ethics.
Another argument against cryonics

You can donate your brain to a brain bank, where it will be preserved for long time and studied. This combines benefits of donation and cryonics.

Might humans not be the most intelligent animals?

Interestingly, we created selection pressure on other species to create something like human intelligence. First of all, dogs, which were selected for15 000 years to be more compatible with humans, which also includes a capability to understand human signals and language. Some dogs could understand few hundreds words.

The Counterfactual Prisoner's Dilemma

The policy is better than opportunity in the legal filed. If one implements a policy "never steal", he wins against criminal law. If one steal only when there is no chance to be caught, that is, he acts based on opportunity, he will be eventually caught.

Inductive biases stick around
I think that this setup will naturally yield a double descent for noisy data: first you get a “likelihood descent” as you get hypotheses with greater and greater likelihood, but then you start overfitting to noise in your data as you get close to the interpolation threshold. Past the interpolation threshold, however, you get a second “prior descent” where you're selecting hypotheses with greater and greater prior probability rather than greater and greater likelihood. I think this is a good model for how modern machine learning works and what double descent is doing.

Reminded me about Ptolemean system and heliocentric system

The Vostok Ice Cores Revisited pt. II

Interesting thing is that ocean has 100 times more dissolved CO2 than atmosphere. All anthropogenic CO2 will eventually dissolve in oceans (but it will take like 1000 years because of slow mixing of deep ocean layers). Currently ocean absorbs 1 ppm a year. More: https://en.wikipedia.org/wiki/Ocean_storage_of_carbon_dioxide

Changes of the ocean temperature could result in CO2 emissions into atmosphere, which could explain the observed historical correlation between CO2 and temperature. Not sure if it is true, but you may look deeper in the direction of changes of ocean's CO2 content,

A dilemma for prosaic AI alignment

It was just an example of the relation between language and the world model. If I have an AI, I can say to it "Find the ways to deflect asteroids". This AI will be able to create a model of Solar system, calculate future trajectories of all dangerous asteroids etc. So it could make a relation between my verbal command and 3D model of the real world.

The same is true if I ask an AI to bring me coffee from the kitchen: it has to select in its world model right kitchen, right type of coffee and right type of future activity.

Humans also do it: any time we read a text, we create a world model which corresponds to the description. And back, if we see a world model, like a picture, we could describe it words.

A dilemma for prosaic AI alignment

In my opinion, such language model should be able to create equivalence between the map of a territory and its verbal description.

In that case, an expression like "the red rose is in the corner" gets meaning as it allows to locate the rose on the map of the room, or otherwise, if the rose is observed in the corner, it could be described as "the rose is in the corner".

Thus natural language could be used to describe all possible operations above world maps, like "all asteroids should be deflected".

Load More