The Nonlinear Library EA - The Value of Consciousness as a Pivotal Question by Derek Shiller
Jul 3, 2024
12:28
Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: The Value of Consciousness as a Pivotal Question, published by Derek Shiller on July 3, 2024 on The Effective Altruism Forum.
Context
Longtermists point out that the scale of our potential for impact is far greater if we are able to influence the course of a long future, as we could change the circumstances of a tremendous number of lives.
One potential avenue for long-term influence involves spreading values that persist and shape the futures that our descendants choose to build. There is some reason to expect that future moral values will be stable. Many groups have preferences about the world beyond their backyard. They should work to ensure that their values are shared by those who can help bring them about. Changes in the values that future groups support will lead to changes in the protections for the things we care about.
If our values concern how our descendants will act, then we should aim to create institutions that promote those values. If we are successful in promoting those values, we should expect our descendants to appreciate and protect those institutional choices.
What values should we work to shape so that the future is as good as it might be? Many of humanity's values would be difficult to sway. Some moral questions, however, might be open to change in the coming decades. It is plausible that there are some questions that we haven't previously faced and for which we have no vested interest. We may be pressed to establish policies and precedents or commit to indifference through inaction.
The right policies and precedents could conceivably allow our values to persist indefinitely. These issues are important to get right, even if we're not yet sure what to think about them.
Controversy
Foremost among important soon-to-be-broached moral questions, I propose, is the moral value that we attribute to phenomenal consciousness (having a 'what-its-like' and a subjective perspective). Or, more particularly, whether mental lives can matter in the absence of phenomenal consciousness in anything like the way they do when supplemented with conscious experiences.
What we decide about the value of phenomenal consciousness in the coming few centuries may not make a difference to our survival as a species, but it seems likely to have a huge effect on how the future plays out.
To get a grip on the problem, consider the case of an artificial creature that is otherwise like a normal person but who lacks phenomenally conscious experiences. Would it be wrong to cause them harm?
Kagan (2019, 28) offers a thought experiment along these lines:
Whatever you feel about this thought experiment, I believe that most people in that situation would feel compelled to grant the robots basic rights.
The significance of consciousness has become a recent popular topic in academic philosophy, particularly in the philosophy of AI, and opinions among professionals are divided. It is striking how greatly opinions differ: where some hold that phenomenal consciousness plays little role in explaining why our lives have value, others hold that phenomenal consciousness is absolutely necessary for having any intrinsic value whatsoever.
One reason to doubt that phenomenal consciousness is necessary for value stems from skepticism that proposed analyses of consciousness describe structures of fundamental importance.
Suppose that the global workspace theory of consciousness is true - to be conscious is to have a certain information architecture involving a central public repository - why should that structure be so important as to ground value? What about other information architectures that function in modestly different ways? The pattern doesn't seem all that important when considered by itself.
If we set aside our preconceptions of consciousness, we wouldn't recognize that architecture as having...
Context
Longtermists point out that the scale of our potential for impact is far greater if we are able to influence the course of a long future, as we could change the circumstances of a tremendous number of lives.
One potential avenue for long-term influence involves spreading values that persist and shape the futures that our descendants choose to build. There is some reason to expect that future moral values will be stable. Many groups have preferences about the world beyond their backyard. They should work to ensure that their values are shared by those who can help bring them about. Changes in the values that future groups support will lead to changes in the protections for the things we care about.
If our values concern how our descendants will act, then we should aim to create institutions that promote those values. If we are successful in promoting those values, we should expect our descendants to appreciate and protect those institutional choices.
What values should we work to shape so that the future is as good as it might be? Many of humanity's values would be difficult to sway. Some moral questions, however, might be open to change in the coming decades. It is plausible that there are some questions that we haven't previously faced and for which we have no vested interest. We may be pressed to establish policies and precedents or commit to indifference through inaction.
The right policies and precedents could conceivably allow our values to persist indefinitely. These issues are important to get right, even if we're not yet sure what to think about them.
Controversy
Foremost among important soon-to-be-broached moral questions, I propose, is the moral value that we attribute to phenomenal consciousness (having a 'what-its-like' and a subjective perspective). Or, more particularly, whether mental lives can matter in the absence of phenomenal consciousness in anything like the way they do when supplemented with conscious experiences.
What we decide about the value of phenomenal consciousness in the coming few centuries may not make a difference to our survival as a species, but it seems likely to have a huge effect on how the future plays out.
To get a grip on the problem, consider the case of an artificial creature that is otherwise like a normal person but who lacks phenomenally conscious experiences. Would it be wrong to cause them harm?
Kagan (2019, 28) offers a thought experiment along these lines:
Whatever you feel about this thought experiment, I believe that most people in that situation would feel compelled to grant the robots basic rights.
The significance of consciousness has become a recent popular topic in academic philosophy, particularly in the philosophy of AI, and opinions among professionals are divided. It is striking how greatly opinions differ: where some hold that phenomenal consciousness plays little role in explaining why our lives have value, others hold that phenomenal consciousness is absolutely necessary for having any intrinsic value whatsoever.
One reason to doubt that phenomenal consciousness is necessary for value stems from skepticism that proposed analyses of consciousness describe structures of fundamental importance.
Suppose that the global workspace theory of consciousness is true - to be conscious is to have a certain information architecture involving a central public repository - why should that structure be so important as to ground value? What about other information architectures that function in modestly different ways? The pattern doesn't seem all that important when considered by itself.
If we set aside our preconceptions of consciousness, we wouldn't recognize that architecture as having...