Someone taking a hard ‘inside view’ about AI risk could reasonably view it as better than AMF for people alive now, or during the rest of their lives. I’m thinking something like:
1 in 10 risk of AI killing everyone within the next 50 years. Spending an extra $1 billion on safety research could reduce the size of this risk by 1%.
$1 billion / (0.1 risk reduced by 1% 8 billion lives) = $125 per life saved. Compares with $3,000-7,000+ for AMF.
This is before considering any upside from improved length or quality of life for the present generation as a result of a value-aligned AI.
I’m probably not quite as optimistic as this, but I still prefer AI as a cause over poverty reduction, for the purposes of helping the present generation (and those remaining to be born during my lifetime).
That’s reasonable, though if the aim is just “benefits over the next 50 years” I think that campaigns against factory farming seem like the stronger comparison:
“We’ve estimated that corporate campaigns can spare over 200 hens from cage confinement for each dollar spent. If we roughly imagine that each hen gains two years of 25%-improved life, this is equivalent to one hen-life-year for every $0.01 spent.”
“One could, of course, value chickens while valuing humans more. If one values humans 10-100x as much, this still implies that corporate campaigns are a far better use of funds (100-1,000x) [So $30-ish per equivalent life saved].”
http://www.openphilanthropy.org/blog/worldview-diversification
And to clarify my first comment, “unlikely to be optimal” = I think it’s a contender, but the base rate for “X is an optimal intervention” is really low.
“if you are only considering the impact on beings alive today...factory farming”
The interventions you are discussing don’t help any beings alive at the time, but only affect the conditions (or existence) of future ones. In particular cage-free campaigns, and campaigns for slower growth-genetics and lower crowding among chickens raised for meat are all about changing the conditions into which future chickens will be born, and don’t involve moving any particular chickens from the old to new systems.
I.e. the case for those interventions already involves rejecting a strong presentist view.
“That’s reasonable, though if the aim is just “benefits over the next 50 years” I think that campaigns against factory farming seem like the stronger comparison:”
Suppose there’s an intelligence explosion in 30 years (not wildly unlikely in expert surveys), and expansion of population by 3-12 orders of magnitude over the next 10 years (with AI life of various kinds outnumbering both human and non-human animals today, with vastly more total computation). Then almost all the well-being of the next 50 years lies in that period.
Also in that scenario existing beings could enjoy accelerated subjective speed of thought and greatly enhanced well-being, so most of the QALY-equivalents for long-lived existing beings could lie there.
Agree with the above, but wanted to ask: what do you mean by a ‘strong presentist’ view? I’ve not heard/seen the term and am unsure what it is contrasted with.
Is ‘weak presentism’ that you give some weight to non-presently existing people, ‘strong presentism’ that you give none?
Why does this confusion persist among long-time EA thought leaders after many years of hashing out the relevant very simple principles? “Beings currently alive” is a judgment about which changes are good in principle, “benefits the next 50 years” is an entirely different pragmatic scope limitation, and people keep bringing up the first in defense of things that can only really be justified by the second.
I understand how someone could be initially confused about this—I was too, initially. But, it seems like the right thing to do once corrected is to actually update your model of the world so you don’t generate the error again. Presentism without negative utilitarianism suggests that we should focus on some combination of curing aging, real wealth creation sufficient to extend this benefit to as many currently alive people as we can, and preventing deaths before we manage to extend this benefit, including due to GCRs likely to happen during the lives of currently living beings.
As it is, we’re not making intellectual progress, since the same errors keep popping up, and we’re not generating actions based on the principles we’re talking about, since people keep bringing up principles that don’t actually recommend the relevant actions. What are we doing, then, when we talk about moral principles?
To add on to this, I think the view you’re referring to is presentism combined with the deprivationism view on death: presentism = only presently alive people matter + deprivationism = the badness of death is the ammount of happiness the person would have had.
You could be, say, a presentist (or other person-affecting view) and combined with say, Epicureanism about death. That would hold only presently alive people matter and there’s no badness in death, and hence no value in extending lives.
If that were your view you’d focus on the suffering of presently humans instead. Probably mental illness or chronic pain. Maybe social isolation if you had a really neat intervention.
But yeah, you’re right that person-affecting views doesn’t capture the intuitive badnes of animal suffering. You could still be a presentist and v*gan on environmental grounds.
And I agree that presentism + deprivationism suggests trying to cure aging is very important and, depending on details, could have higher EV than suffering relief. I’m unclear that real wealth creation would do very much due to hedonic adaptation and social comparison challenges.
Someone taking a hard ‘inside view’ about AI risk could reasonably view it as better than AMF for people alive now, or during the rest of their lives. I’m thinking something like:
1 in 10 risk of AI killing everyone within the next 50 years. Spending an extra $1 billion on safety research could reduce the size of this risk by 1%.
$1 billion / (0.1 risk reduced by 1% 8 billion lives) = $125 per life saved. Compares with $3,000-7,000+ for AMF.
This is before considering any upside from improved length or quality of life for the present generation as a result of a value-aligned AI.
I’m probably not quite as optimistic as this, but I still prefer AI as a cause over poverty reduction, for the purposes of helping the present generation (and those remaining to be born during my lifetime).
That’s reasonable, though if the aim is just “benefits over the next 50 years” I think that campaigns against factory farming seem like the stronger comparison:
“We’ve estimated that corporate campaigns can spare over 200 hens from cage confinement for each dollar spent. If we roughly imagine that each hen gains two years of 25%-improved life, this is equivalent to one hen-life-year for every $0.01 spent.” “One could, of course, value chickens while valuing humans more. If one values humans 10-100x as much, this still implies that corporate campaigns are a far better use of funds (100-1,000x) [So $30-ish per equivalent life saved].” http://www.openphilanthropy.org/blog/worldview-diversification
And to clarify my first comment, “unlikely to be optimal” = I think it’s a contender, but the base rate for “X is an optimal intervention” is really low.
“if you are only considering the impact on beings alive today...factory farming”
The interventions you are discussing don’t help any beings alive at the time, but only affect the conditions (or existence) of future ones. In particular cage-free campaigns, and campaigns for slower growth-genetics and lower crowding among chickens raised for meat are all about changing the conditions into which future chickens will be born, and don’t involve moving any particular chickens from the old to new systems.
I.e. the case for those interventions already involves rejecting a strong presentist view.
“That’s reasonable, though if the aim is just “benefits over the next 50 years” I think that campaigns against factory farming seem like the stronger comparison:”
Suppose there’s an intelligence explosion in 30 years (not wildly unlikely in expert surveys), and expansion of population by 3-12 orders of magnitude over the next 10 years (with AI life of various kinds outnumbering both human and non-human animals today, with vastly more total computation). Then almost all the well-being of the next 50 years lies in that period.
Also in that scenario existing beings could enjoy accelerated subjective speed of thought and greatly enhanced well-being, so most of the QALY-equivalents for long-lived existing beings could lie there.
Mea culpa that I switched from “impact on beings alive today” to “benefits over the next 50 years” without noticing.
Agree with the above, but wanted to ask: what do you mean by a ‘strong presentist’ view? I’ve not heard/seen the term and am unsure what it is contrasted with.
Is ‘weak presentism’ that you give some weight to non-presently existing people, ‘strong presentism’ that you give none?
“Is ‘weak presentism’ that you give some weight to non-presently existing people, ‘strong presentism’ that you give none?”
In my comment, yes.
Why does this confusion persist among long-time EA thought leaders after many years of hashing out the relevant very simple principles? “Beings currently alive” is a judgment about which changes are good in principle, “benefits the next 50 years” is an entirely different pragmatic scope limitation, and people keep bringing up the first in defense of things that can only really be justified by the second.
I understand how someone could be initially confused about this—I was too, initially. But, it seems like the right thing to do once corrected is to actually update your model of the world so you don’t generate the error again. Presentism without negative utilitarianism suggests that we should focus on some combination of curing aging, real wealth creation sufficient to extend this benefit to as many currently alive people as we can, and preventing deaths before we manage to extend this benefit, including due to GCRs likely to happen during the lives of currently living beings.
As it is, we’re not making intellectual progress, since the same errors keep popping up, and we’re not generating actions based on the principles we’re talking about, since people keep bringing up principles that don’t actually recommend the relevant actions. What are we doing, then, when we talk about moral principles?
To add on to this, I think the view you’re referring to is presentism combined with the deprivationism view on death: presentism = only presently alive people matter + deprivationism = the badness of death is the ammount of happiness the person would have had.
You could be, say, a presentist (or other person-affecting view) and combined with say, Epicureanism about death. That would hold only presently alive people matter and there’s no badness in death, and hence no value in extending lives.
If that were your view you’d focus on the suffering of presently humans instead. Probably mental illness or chronic pain. Maybe social isolation if you had a really neat intervention.
But yeah, you’re right that person-affecting views doesn’t capture the intuitive badnes of animal suffering. You could still be a presentist and v*gan on environmental grounds.
And I agree that presentism + deprivationism suggests trying to cure aging is very important and, depending on details, could have higher EV than suffering relief. I’m unclear that real wealth creation would do very much due to hedonic adaptation and social comparison challenges.