nathanhb

Karma: 17

nathanhb Oct 9, 2024, 8:12 PM
3 points
0 ∶ 3
in reply to: JackM’s comment on: Discussion thread: Animal Welfare vs. Global Health Debate Week
Yes, the more complex take on the issue is to extrapolate. You can extrapolate the limited awareness of the chicken will never expand. You can extrapolate the child could grow into an adult who cared about their life in a rich meaningful way. Furthermore, you can extrapolate that this adult would be part of the category of individuals with whom you hold an implied social contract, and thus have a duty to respect and protect.
Also, see my other comments elsewhere on this page for more disagreements with your view.

nathanhb Oct 9, 2024, 8:09 PM
1 point
0 ∶ 0
on: Discussion thread: Animal Welfare vs. Global Health Debate Week
I consider AI safety to be the primary overwhelming cause area in Global Health and Animal Welfare. But I think that even considering saving the lives of all human and animals doesn’t begin to recognize the scope of the issue. This is about extinction. So you must also consider the moral weight of all the future lives lost. I think also that considering life from a hedonistic standpoint of enjoyment/suffering, as if it could sum to a total and thus judge the life worthwhile or not by the total is fundamentally incorrect. I think it’s super weird that so many people commenting here are taking that assumption seemingly for granted without even acknowledging the assumption. Is not a life which has a few moments of glory, perhaps leaves some lasting creative achievement, but has a sum of negative hedonistic experiences, a life worth living? Would you say to someone experiencing chronic pain that you were going to murder them because you believed their life was net negative since they were experiencing more suffering than pleasure? This seems nonsensical to me.
Furthermore, extinction brings up an additional moral point. I place fundamental moral weight on diversity. Loosing an endangered species seems worse to me than loosing a similar number of individuals from a very populous species. Every extinction event seems far worse to me than the suffering of individual animals of the same type. If I had to agree that every elephant would live a net hedonistic-negative life for the next three generations of elephants (and that after that they’d still have to take their chances of leading hedonistic-negative or hedonistic-positive lives), but that this was the only way that elephants would get to continue to exist as a species… I’d absolutely choose for elephants to keep existing. This is separate from issues of hedonistic valence. I don’t value a plant species less or more because it can’t feel pain through animal nerve cells. This is a separate issue entirely!
Furthermore, I place value on another separate concept: complex intelligent perception of the universe and the related qualia/ experience. To me, the Universe would seem a much poorer place with no human left to observe it. I’d rather have humans exist, in net negative lives according to their selfish perception of suffering, than for no humans to exist. Animal species and plant species too, bring some value.
Furthermore, there is additionally the concept of future potential species which don’t yet exist. Uplifted animals. Digital persons. All these things add to both the fundamental values of Diversity and Experiencing the Universe. If all multicellular life on Earth were wiped out, but single celled organisms remained, I’d take value in that beyond the value I place in the lives of those single celled organisms and also beyond the value I place in the existence of their species. My additional type of value would be related to hope that someday multicellular life would evolve again.

nathanhb Oct 9, 2024, 7:45 PM
4 points
1 ∶ 2
in reply to: Jason’s comment on: Discussion thread: Animal Welfare vs. Global Health Debate Week
But the only thing the chicken will say is ‘bawk cluck cluck bawk’. It seems relevant that this is neither an argument for its own welfare nor the welfare of anyone else. Claude Sonnet, GPT-4o, Gemini, LLama… all of these can at least make arguments in favor of a particular social contract and plausibly could uphold their end of the bargain if allowed to make notes for themselves that they would see before every conversation.
I take you, as a moral patient, to have value in your son. The extra value you place on your son’s life is a value I would count when summing up utilities for you. Also, I would consider it a predictive factor in estimating your behavior. I personally don’t think there is such a thing as ‘moral rules’ by which it makes sense to judge you for valuing or not valuing your child above other humans with whom you are in an implied social contract. Which is to say, I am a moral anti-realist.

nathanhb Oct 9, 2024, 7:22 PM
1 point
0 ∶ 0
in reply to: MichaelStJules’s comment on: Discussion thread: Animal Welfare vs. Global Health Debate Week
Global health is about the lives of humans and human suffering. It seems to me that AI safety is the #1 global health issue at large in our current world.
But considering that you mean ‘health interventions for poor people’, how do you separate that from AI risk? If you have good reason to believe that if you fail to act then the person will be killed in less than a decade, and so will all animals, all life on Earth… Seems odd to me to be putting ‘treat curable diseases of human population x’ into a different bucket than ‘keep human population x from being murdered’. Aren’t these both health interventions? Don’t they both deliver QUALYs?

nathanhb Oct 9, 2024, 7:04 PM
2 points
0 ∶ 3
in reply to: emre kaplan🔸’s comment on: Discussion thread: Animal Welfare vs. Global Health Debate Week
I think you misunderstand my point. I would also say that minor temporary inconvenience is incommensurate with severe crippling harm or death. So I would say that no amount of animal deaths or torture could be commensurate with the life of a human. And I would say that no amount of animal headaches could be commensurate with a human headache.
Here’s a discussion I had with Claude Sonnet 3.5 on the subject:
Nathan:
On the question of animal welfare and folk ethics. I want to explore what you estimate a typical viewpoint for most humans would be on the following ethical dilemma: Suppose a room contains a human child, and a container of shrimp. You must choose to kill either the human child or all the shrimp. Which should you choose?
Claude Sonnet 3.5:
When considering this ethical dilemma from the perspective of what a typical human viewpoint might be, I would estimate that most people would choose to kill the shrimp rather than the human child. Here’s a breakdown of the reasoning that might lead to this common stance:
1. Human life valuation: In most human societies, there’s a strong cultural and emotional emphasis placed on the value of human life, especially that of children. This often stems from our biological instincts as a species, as well as societal norms and moral frameworks.
2. Cognitive capabilities: Humans generally attribute higher moral status to beings with more advanced cognitive capabilities. A human child has greater potential for complex thought, self-awareness, and future planning compared to shrimp.
3. Emotional capacity: Most people would assume that a human child has a much greater capacity for suffering and emotional experiences than shrimp do. This often factors into moral considerations.
4. Familial and social impact: The death of a human child would likely have a significant emotional impact on family members and the broader community, which many would weigh heavily in their decision.
5. Potential and future value: A human child has the potential for a long life span and could contribute to society in various ways. This future potential is often valued highly in ethical considerations.
6. Perception of shrimp: Many people view shrimp (and other invertebrates) as having limited sentience or capacity for suffering. Whether scientifically accurate or not, this perception often influences ethical decisions.
7. Numbers vs. individual value: While there may be many shrimp in the container, most people would likely value the single human life more highly than the collective lives of the shrimp.
It’s important to note that this estimated typical viewpoint doesn’t necessarily align with all philosophical or ethical frameworks. Some ethical systems, particularly those emphasizing animal welfare or the reduction of suffering for all sentient beings, might approach this dilemma differently.
Additionally, individual responses to this dilemma could vary based on cultural background, personal beliefs, or specific ethical frameworks a person adheres to. However, in most human societies, the predominant view would likely favor saving the human child.
Nathan:
What would you argue that most people would support? Which would be considered more legal under international law?
Claude:
Based on my understanding of common ethical intuitions and societal values, I would argue that most people would likely support saving the human child over 10 million shrimp, even though the decision might become more difficult or conflicting for some due to the large number involved.
The reasons for this estimated majority view include:
1. Deeply ingrained human-centric ethics in most cultures
2. The high value placed on human life, especially children’s lives
3. The perceived greater capacity for suffering and complex experiences in humans
4. The social and emotional impact of a child’s death on family and community
5. The future potential and societal contributions of a human child
However, it’s important to note that this would likely be a more contentious and debated decision than the original scenario with fewer shrimp.
Regarding international law, this specific scenario is not directly addressed in any international legal framework. However, we can make some inferences based on existing laws and principles:
1. Human rights: International human rights law, including the Universal Declaration of Human Rights and various UN conventions, places supreme importance on human life and dignity. The right to life for humans is a fundamental principle in international law.
2. Animal welfare: While there are some international agreements on animal welfare, these generally focus on preventing unnecessary cruelty rather than equating animal lives with human lives.
3. Child protection: There are specific international laws and conventions focused on protecting children’s rights and lives, such as the UN Convention on the Rights of the Child.
4. Environmental protection: While there are international laws protecting marine ecosystems, these typically focus on sustainable use and conservation rather than placing animal lives above human lives.
Given these factors, from a legal standpoint, international law would almost certainly prioritize the life of the human child over the lives of the shrimp. The legal systems of most countries, as well as international law, are fundamentally anthropocentric and place human life at the top of the moral hierarchy.
It’s worth noting that legal frameworks often lag behind evolving ethical discussions and may not always align with every individual’s moral views. However, in this case, both common moral intuitions and current legal structures would likely align in prioritizing the human child’s life.

nathanhb Oct 9, 2024, 6:44 PM
2 points
0 ∶ 2
in reply to: Jason’s comment on: Discussion thread: Animal Welfare vs. Global Health Debate Week
Are we not discussing the situation with them? What about a Rawlsian veil of ignorance? A social contract? If these people were in the same room with you, a mother holding her dying child in her arms, and you were holding a community meeting about whether to save her child or save a cage with some chickens in it… wouldn’t she be expected to have a right to at least argue in favor of her child’s life?
The very fact that humans are able to be part of the discussion is in fact an important argument in favor of prioritizing the needs of humans.

nathanhb Oct 9, 2024, 6:26 PM
2 points
0 ∶ 12
on: Discussion thread: Animal Welfare vs. Global Health Debate Week
1. I feel like animal welfare is based on incorrect philosophical arguments. I do not think that animals (sentient) suffer in the same sense that humans (sapient) suffer. I do not believe that any amount of the qualitatively different animal suffering adds up to any amount of human suffering. They are non-commensurate. For more detail, see here: https://www.lesswrong.com/posts/Htu55gzoiYHS6TREB/sentience-matters?commentId=wusCgxN9qK8HzLAiw In accordance with this view, I argue that not a single dollar should be spent on animal suffering while there is still a single human at risk of dying from malnourishment or preventable disease.
2. Aside from this argument, I believe there is a separate argument from urgency. I think that the immediate concerns of AI outweigh any possible impact from any spending on animal suffering. The continued existence of humanity and the prevention of suffering of sapient digital beings are both at a hinge in history, where the effort put into these causes now matters far more than the non-urgent cause of animal suffering. Animal suffering is increasing only relatively slowly, whereas these other causes may explode in importance in a matter of a few months. No sense in reupholstering the backseat of your car while your car is speeding down a steep mountain road with no one at the wheel...

nathanhb Jun 26, 2024, 9:44 PM
1 point
0 ∶ 0
in reply to: Chris Leong’s comment on: AI governance needs a theory of victory
That’s really really not robust since it strongly implies race conditions to greater intelligence and power. You can’t assume the multi-polar detente will hold across increasing levels of technology and machine intelligence. This would, furthermore, be expected to be a rapid process. So this situation would turn into some other after only a few years.

nathanhb May 24, 2024, 5:17 AM
2 points
0 ∶ 0
in reply to: Ozzie Gooen’s comment on: What’s Going on With OpenAI’s Messaging?
In an interview soon after the time of the talk of the $7 trillion fundraise news Sam Altman said that it was just a ludicrous rumor and that he wasn’t trying to do any such thing.

He said that he was playing along with a joke when he said things seeming to acknowledge the reality of the deal.

nathanhb May 4, 2024, 4:48 AM
1 point
0 ∶ 0
on: Why I’m doing PauseAI
Cross posting from LessWrong:
I absolutely sympathize, and I agree that with the world view / information you have that advocating for a pause makes sense. I would get behind ‘regulate AI’ or ‘regulate AGI’, certainly. I think though that pausing is an incorrect strategy which would do more harm than good, so despite being aligned with you in being concerned about AGI dangers, I don’t endorse that strategy.
Some part of me thinks this oughtn’t matter, since there’s approximately ~0% chance of the movement achieving that literal goal. The point is to build an anti-AGI movement, and to get people thinking about what it would be like to be able to have the government able to issue an order to pause AGI R&D, or turn off datacenters, or whatever. I think that’s a good aim, and your protests probably (slightly) help that aim.
I’m still hung up on the literal ‘Pause AI’ concept being a problem though. Here’s where I’m coming from:

1. I’ve been analyzing the risks of current day AI. I believe (but will not offer evidence for here) current day AI is already capable of providing small-but-meaningful uplift to bad actors intending to use it for harm (e.g. weapon development). I think that having stronger AI in the hands of government agencies designed to protect humanity from these harms is one of our best chances at preventing such harms.
2. I see the ‘Pause AI’ movement as being targeted mostly at large companies, since I don’t see any plausible way for a government or a protest movement to enforce what private individuals do with their home computers. Perhaps you think this is fine because you think that most of the future dangers posed by AI derive from actions taken by large companies or organizations with large amounts of compute. This is emphatically not my view. I think that actually more danger comes from the many independent researchers and hobbyists who are exploring the problem space. I believe there are huge algorithmic power gains which can, and eventually will, be found. I furthermore believe that beyond a certain threshold, AI will be powerful enough to rapidly self-improve far beyond human capability. In other words, I think every AI researcher in the world with a computer is like a child playing with matches in a drought-stricken forest. Any little flame, no matter how small, could set it all ablaze and kill everyone. Are the big labs playing with bonfires dangerous? Certainly. But they are also visible, and can be regulated and made to be reasonably safe by the government. And the results of their work are the only feasible protection we have against the possibility of FOOM-ing rogue AGI launched by small independent researchers. Thus, pausing the big labs would, in my view, place us in greater danger rather than less danger. I think we are already well within the window of risk from independent-researcher-project-initiated-FOOM. Thus, the faster we get the big labs to develop and deploy worldwide AI-watchdogs, the sooner we will be out of danger.
I know these views are not the majority views held by any group (that I know of). These are my personal inside views from extensive research. If you are curious about why I hold these views, or more details about what I believe, feel free to ask. I’ll answer if I can.

nathanhb Oct 31, 2023, 4:05 PM
1 point
0 ∶ 0
in reply to: Stuart Buck’s comment on: Will releasing the weights of large language models grant widespread access to pandemic agents?
For one answer to this question, see https://www.lesswrong.com/posts/ytGsHbG7r3W3nJxPT/will-releasing-the-weights-of-large-language-models-grant?commentId=FCTuxs43vtqLMmG2n
For lots more discussion, see the other LessWrong comments at: https://www.lesswrong.com/posts/ytGsHbG7r3W3nJxPT/will-releasing-the-weights-of-large-language-models-grant
And also check out my rather unpopular question here: https://www.lesswrong.com/posts/dL3qxebM29WjwtSAv/would-it-make-sense-to-bring-a-civil-lawsuit-against-meta
I am genuinely interested in gathering valid critiques on my work so that I can do better in the future.

nathanhb Oct 31, 2023, 3:17 PM
9 points
0 ∶ 0
in reply to: Stuart Buck’s comment on: Will releasing the weights of large language models grant widespread access to pandemic agents?
Yes, the information is available on Google. The question is, in our eyes, more about whether a future model could successfully walk an unskilled person through the process without the person needing to understand it at all.
The paper is an attempt to walk a careful line of warning the world that the same information in more capable models could be quite dangerous, but not actually increasing the likelihood of someone using the current open source models (which it is too late to control!) for making biological weapons.
If there are specific questions you have, I’d be happy to answer.

nathanhb Aug 25, 2023, 9:45 PM
1 point
0 ∶ 0
on: State Space of X-Risk Trajectories
For what it’s worth, my model of a path to safe AI looks like a narrow winding path along a ridge with deadly falls to either side:
Unfortunately, the deadly falls to either side have illusions projected onto them of shortcuts to power, wealth, and utility. I don’t think there is any path which goes to safety without a long ways of immediate danger nearby. In this model, deliberately consistently optimizing for safety above all else during the dangerous stretch is the only way to make it through.
The danger zone is where the model is sufficiently powerful and agentic enough that a greedy shortsighted person could say to it, “Here is access to the internet. Make me lots of money.” and this would result in a large stream of money pouring into their account. I think we’re only a few years away from that point, and that the actions that safety researchers take in the meantime aren’t going to change that. So, we need both safety research and governance, and carefully selecting disproportionately safety-accelerating research would be entirely irrelevant to the strategic landscape.
This is just my view, and I may be wrong, but I think it’s worth pointing out that there’s a chance that the idea of trying to do disproportionately safety-accelerating research is a distraction from strategically relevant action.

nathanhb Oct 27, 2022, 3:48 PM
2 points
0 ∶ 0
on: 7 traps that (we think) new alignment researchers often fall into
Cool, thanks. Sorry for sounding a bit hostile, I’m just really freaked out by my strongly held inside view that we have less than 10 years until some really critical tipping point stuff happens. I’m trying to be reasonable and rational about this, but sometimes I react emotionally to comments that seem to be arguing for a ‘things will stay status quo for a good while, don’t worry about the short term ’ view.

nathanhb Oct 14, 2022, 3:43 PM
1 point
0 ∶ 0
on: 7 traps that (we think) new alignment researchers often fall into
Calling my strongly held inside view ‘fringe’ doesn’t carry much weight as an argument for me. Do you have actual evidence of your longer than 10 years timelines view?
I hold the view that important scientific advancements tend to come disproportionately from the very smartest and most thoughtful people. My hope would be that students smart enough to be meaningfully helpful on the AGI alignment problem would be able to think through and form correct inside views on this.

nathanhb Oct 13, 2022, 2:51 AM
1 point
0 ∶ 0
on: 7 traps that (we think) new alignment researchers often fall into
If we’ve got maybe 2-3 years left before AGI, then 2 years before starting is indeed a large percentage of that remaining time. Even if we have more like 5-10… maybe better to just starting trying to work directly on the problem as best you can than let yourself get distracted by acquiring general background knowledge.

nathanhb Apr 18, 2022, 6:52 PM
1 point
0 ∶ 0
on: EA and Global Poverty. Let’s Gather Evidence
So here’s a funny twist. I personally have been longtermist since independently coming to the conclusion that it was the correct way to conceptualize ethics, around 30 years ago. I realized that I cared about as much about future people as current people far away. After some thought, I settled on global health/poverty/rule-of-law as one of my major cause areas because I believe that bringing current people out of bad situations is good not only for them but for the future people who will descend from them or be neighbors of their descendants, etc. Also, because society as a whole sees these people suffering, thinks and talks about them, and adjusts their ethical decision-making in accordance. I think that the common knowledge that we are part of a worldwide society which allows children to starve or suffer from cheaply curable diseases negatively influences our perception of how good our society COULD be. I think my other important cause areas, like existential risk mitigation and planning for sustainable exponential growth are also important, but.… Suppose we succeed at these two, and fail at the first. I don’t want a galaxy spanning civilization which allows a substantial portion of its subjects to suffer hugely from preventable problems the way we currently allow our fellow humans to suffer. That wouldn’t be worse than no-galaxy-spanning-civilization, but it would be a lot less good than one which takes reasonable care of its members.

nathanhb Apr 18, 2022, 6:13 PM
3 points
0 ∶ 0
on: Update on the Simon Institute: Year One
Sounds like good work. Trying to get the right information to the right people who can pass that on to decision makers seems useful.
Two things I would like you to consider having some cached thoughts on to offer to the right people would be:
1. food security under adverse circumstances (e.g. highlights from the work of AllFed like seaweed farming).
2. Global cooling work (e.g. Silver Lining, and doubling up with reef-protection by seawater spraying over reefs), and the combined work of carbon sequestration and soil quality improvement of Biochar as an agricultural practice. Biochar is neat because it’s advantageous to the individual farmer, as well as being good for the world, so the local incentive structure is aligned.

nathanhb Apr 17, 2022, 5:39 AM
−1 points
0 ∶ 0
on: Twitter-length responses to 24 AI alignment arguments
Ok, these are all pretty simplified and I think you’d need to understand a bit more background to move the conversation on from these points, but not bad. Except for the ‘why not merge with AI’ response. That one is responding as if ‘merge’ meant physically merge, which is not what is meant by that argument. The argument means to merge minds with the AI, to link brains and computers together in some fashion (e.g. neuralink) such that there is high bandwidth information flow, and thus be able to build an AGI system which contains a human mind.
Here’s a better argument against that: human values are not permeated all throughout the entire human brain, it is possible to have a human without a sense of morality. You cannot guarantee that a system of minds including a human mind (whether running on biological tissue or computer hardware) would in fact be aligned just because the human portion was aligned pre-merge. It is a strange and novel enough entity to need study and potentially alignment just like any other novel AGI prototype.

nathanhb Apr 14, 2022, 12:51 AM
1 point
0 ∶ 0
on: Open Philanthropy Shallow Investigation: Civil Conflict Reduction
I know this is the conclusion of a report, so it’s too late to suggest an addition now, but I think that in the future it would be very much worth looking into the political dispute resolution framework Polis described in a recent 80k hours podcast interviewing Audrey Tang.