E.g. one could try to grow new connections. But how would that affect existing cognition? Do your models say anything about the consequences of adding new connections in mid-life?
No idea. But uninformed armchair speculation is fun so I’ll keep writing anyway. :) (Assume for the sake of argument that I’m basically right about the root cause.)
If we split symptoms into {positive, negative, cognitive}, then on this theory, the cognitive symptoms are pretty closely tied to the root cause, whereas the positive and negative symptoms are a bit downstream of the root cause.
Can we intervene on the root cause? I’m pretty skeptical. I think the typical case (see Section 5) is that the long-range connections are not there in the first place. I’d be pretty surprised if it were possible to grow new long-range connections in adulthood, although I’m not an expert. OTOH, I can imagine that there’s a subset of schizophrenia patients for which the long-range connections are present but ineffective—let’s say for lack of some protein or whatever. For that subset, I’d be more hopeful for a miracle-cure that successfully treats the root cause. But it might be a tiny subset of patients, if any, I dunno.
So I’m generally not holding out much hope about the cognitive symptoms. By contrast, the positive & negative symptoms are obviously treatable by existing drugs (not sure exactly how well they work). I think pos & neg symptoms a bit more distant from the root cause and can be mitigated by other pathways.
But again, I don’t know, that’s just off-the-cuff uninformed speculation. ¯\_(ツ)_/¯
I don’t believe uploads would be conscious, but let’s put that aside for a moment… Suppose long-range connections were somehow added to an already adult brain that had lacked them. Do your cognitive models say anything about what the effects would be?
I think everything would be fine. I don’t see any issues. As long as the relevant learning rates is set to more than zero, I think the cortical learning algorithm would “do its thing”, and it would learn newer better predictive models that bring together spatially-distant information streams, and all the schizophrenia symptoms (including cognitive) would gradually go away.
Any thoughts on what this implies for a cure?
E.g. one could try to grow new connections. But how would that affect existing cognition? Do your models say anything about the consequences of adding new connections in mid-life?
No idea. But uninformed armchair speculation is fun so I’ll keep writing anyway. :) (Assume for the sake of argument that I’m basically right about the root cause.)
If we split symptoms into {positive, negative, cognitive}, then on this theory, the cognitive symptoms are pretty closely tied to the root cause, whereas the positive and negative symptoms are a bit downstream of the root cause.
Can we intervene on the root cause? I’m pretty skeptical. I think the typical case (see Section 5) is that the long-range connections are not there in the first place. I’d be pretty surprised if it were possible to grow new long-range connections in adulthood, although I’m not an expert. OTOH, I can imagine that there’s a subset of schizophrenia patients for which the long-range connections are present but ineffective—let’s say for lack of some protein or whatever. For that subset, I’d be more hopeful for a miracle-cure that successfully treats the root cause. But it might be a tiny subset of patients, if any, I dunno.
So I’m generally not holding out much hope about the cognitive symptoms. By contrast, the positive & negative symptoms are obviously treatable by existing drugs (not sure exactly how well they work). I think pos & neg symptoms a bit more distant from the root cause and can be mitigated by other pathways.
But again, I don’t know, that’s just off-the-cuff uninformed speculation. ¯\_(ツ)_/¯
Do you agree that with sufficiently advanced technology, this would be possible?
With sufficiently advanced technology, we can upload everyone to whole-brain-emulations, and then add or subtract whatever connections we want :)
I don’t believe uploads would be conscious, but let’s put that aside for a moment… Suppose long-range connections were somehow added to an already adult brain that had lacked them. Do your cognitive models say anything about what the effects would be?
I think everything would be fine. I don’t see any issues. As long as the relevant learning rates is set to more than zero, I think the cortical learning algorithm would “do its thing”, and it would learn newer better predictive models that bring together spatially-distant information streams, and all the schizophrenia symptoms (including cognitive) would gradually go away.