Ok. I have thought about it further and here is the reason I think you’re wrong. You implicitly have made an assumption that the tools available to neuroscientist today are good, and we have a civilization with the excess resources to support such an endeavor.
This is false. Today the available resources for such endeavors is only enough to fund small teams. The research that is profitable like silicon chip improvement gets hundreds of billions invested into it.
So any extrapolation is kinda meaningless. It would be like asking in 1860 how many subway tunnels would be in NYC in 1940. The heavy industry to build it simply didn’t exist so you would have to conclude it would be slow going.
Similarly, the robotic equipment to do bioscience is currently limited and specialized. It’s why a genome can be sequenced for a few thousand dollars but graduate students still use pipettes.
Oh and if you wanted to know when the human genome would be sequenced in 1920, and in 1930 learned that zero genes had been sequenced, you might make a similar conclusion.
Is the technology feasible with demonstrated techniques at a laboratory level.
Will there likely be gain to the organization that sells or deploys this technology in excess of it’s estimated cost?
Does the technology run afoul of existing government regulation that will slow research into it?
Does the technology have a global market that will result in a sigmoidal adoption curve?
Electric cars should have been predictable this way:
They were feasible since 1996, or 1990. (LFP battery is the first lithium chemistry with the lifespan to be a net gain for an EV, 1990 is the first ‘modern’ lithium battery assembled in a lab)
The gain is reduced fuel cost, maintenance cost, and supercar acceleration and vehicle performance with much cheaper drivetrains.
Governments perceive a benefit in EVs so they have subsidized the research
Yes, and the adoption curve is sigmoidal.
smartphones follow a similar such set of arguments, and the chips that made them possible were only low power enough around the point that the sigmoidal adoption started. They were not really possible much prior. Also, Apple made a large upfront investment to deliver an acceptable user experience all at once, rather than incrementally adding features like other early smartphone manufacturers did.
I will put my stake in the sand and say that autonomous cars fit all these rules:
- Feasible, as the fundamental problem of assessing collision risk for a candidate path, the only part the car has to have perfect, is a simple and existing algorithm
- enormous gain, easily north of a trillion dollars in annual revenue or hundreds of billions in annual profit will be available to the industry
- Governments are reluctant but are obviously allowing the research and testing
- The adoption curve will be sigmoidal, because it has obvious self gain. The first shipping autonomous EVs will likely produce a cost advantage for a taxi firm or be rented directly, and will be immediately adopted, and the revenue reinvested makes their cost advantage grow until on the upward climb of the adoption curve the limit is simply how fast the hardware can be manufactured.
I will take it a step further, and say that general robots that solve problems of the same form as the problem of autonomous cars also fit all the same rules, will be adopted, it will be sigmoidal, and other reports have estimated that about half of all jobs will be replaced.
Anyways, for uploading a nematode, the optical I/O techniques to debug a living neural system I think are still in the laboratory prototype stage. Does anyone have this working in any lab animal anywhere? So it doesn’t even meet condition 1. And what’s the gain if you upload a nematode? Not much. Certainly not in excess of the tens of millions of dollars it likely will cost. Governments are disinterested as nematodes are not vertebrates. And there’s no “self-gain”, upload a nematode and no one is going to be uploading nematodes all over the planet.
There still will be progress, and with advances in other areas—the general robotics mentioned above—this would free up resources and make possible something like a human upload project.
And that, if you had demonstrations of feasibility, does meet all 4 conditions.
-assume you have demonstrated feasibility with neuroscience experiments that will be performed and can “upload” a tiny section of human brain tissue.
- The gain is you get to charge each uploaded human all their assets accumulated in life, and/or they will likely have superhuman capabilities once uploaded. This gain is more like “divide by zero” gain, uploaded humans would be capable of organizing projects to tear down the solar system for raw materials, or essentially “near infinite money”.
- Governments will have to push it with all-out efforts near the end-game because to not have uploaded humans or AI is to lose all sovereignty
- Adoption curve is trivially sigmoidal.
I don’t know when all the conditions will be met, but uploading humans is a project similar to nuclear weapons, in terms of gain and how up until just 29 months! before detonation the amount of fission done on earth by humans was zero. In 1900 you might feel safe in predicting no fission before the end of the century like you do now.
Also, you can use this method to disprove personal jetpacks or fusion power.
feasible- personal jetpacks, no, rocket jetpacks of the 1960s had 22 second flight times.
fusion power—no, no experiments were developing significant fusion gain without a fission device to provide the containment pressure
gain—no. Jetpacks would guzzle jet fuel even with more practical forms that worked more like a VTOL aircraft, and the value of this fuel is going to exceed the value of the time saved to almost all users. Fusion power is a method to boil water using high energy laboratory equipment and is unlikely to be cheaper than the competition over any feasible timescale.
government- no. Jetpacks cause extreme noise pollution and extra fires and deaths from when they fall out of the sky. Fusion is a nuclear proliferation risk as a fusion reactor provides a source of neutrons that could be used to transmute to plutonium.
sigmoidal—no, you can’t have this without large gain. Maybe this criterion is redundant.
If you got this far in reading, one notable fault of this proposed algorithm is it does not predict technologies requiring a true breakthrough. You could not predict lasers, for instance, as these were not known to be feasible until the 1960s when the first working models existed. That’s a real breakthrough. The distinction I am making is that if you do not know if physics will allow something to work, or if physics will allow something to work well, then you need a breakthrough to get it working.
Ditto argument for math algorithms, neural networks I would say are another real breakthrough, as they are much “better” for how little we know what we are doing than they should be.
We do know that physics will allow us to build a computer big enough to emulate a brain, to scan at least the synaptome of a once living brain, and get some detail on the weights. We also know that learning means we do not really have be all that exact.
This doesn’t give any real help in guessing the timing. But I think the curve to imagine is much closer to a step function than it is to a linear slope. So not seeing an increase just means we haven’t reached the step, not that their is a linear slope that is too small to see.
Ok. I have thought about it further and here is the reason I think you’re wrong. You implicitly have made an assumption that the tools available to neuroscientist today are good, and we have a civilization with the excess resources to support such an endeavor.
This is false. Today the available resources for such endeavors is only enough to fund small teams. The research that is profitable like silicon chip improvement gets hundreds of billions invested into it.
So any extrapolation is kinda meaningless. It would be like asking in 1860 how many subway tunnels would be in NYC in 1940. The heavy industry to build it simply didn’t exist so you would have to conclude it would be slow going.
Similarly, the robotic equipment to do bioscience is currently limited and specialized. It’s why a genome can be sequenced for a few thousand dollars but graduate students still use pipettes.
Oh and if you wanted to know when the human genome would be sequenced in 1920, and in 1930 learned that zero genes had been sequenced, you might make a similar conclusion.
Do you have a better way of estimating the timing of new technologies that require many breakthroughs to reach?
I’ll try to propose one.
Is the technology feasible with demonstrated techniques at a laboratory level.
Will there likely be gain to the organization that sells or deploys this technology in excess of it’s estimated cost?
Does the technology run afoul of existing government regulation that will slow research into it?
Does the technology have a global market that will result in a sigmoidal adoption curve?
Electric cars should have been predictable this way:
They were feasible since 1996, or 1990. (LFP battery is the first lithium chemistry with the lifespan to be a net gain for an EV, 1990 is the first ‘modern’ lithium battery assembled in a lab)
The gain is reduced fuel cost, maintenance cost, and supercar acceleration and vehicle performance with much cheaper drivetrains.
Governments perceive a benefit in EVs so they have subsidized the research
Yes, and the adoption curve is sigmoidal.
smartphones follow a similar such set of arguments, and the chips that made them possible were only low power enough around the point that the sigmoidal adoption started. They were not really possible much prior. Also, Apple made a large upfront investment to deliver an acceptable user experience all at once, rather than incrementally adding features like other early smartphone manufacturers did.
I will put my stake in the sand and say that autonomous cars fit all these rules:
- Feasible, as the fundamental problem of assessing collision risk for a candidate path, the only part the car has to have perfect, is a simple and existing algorithm
- enormous gain, easily north of a trillion dollars in annual revenue or hundreds of billions in annual profit will be available to the industry
- Governments are reluctant but are obviously allowing the research and testing
- The adoption curve will be sigmoidal, because it has obvious self gain. The first shipping autonomous EVs will likely produce a cost advantage for a taxi firm or be rented directly, and will be immediately adopted, and the revenue reinvested makes their cost advantage grow until on the upward climb of the adoption curve the limit is simply how fast the hardware can be manufactured.
I will take it a step further, and say that general robots that solve problems of the same form as the problem of autonomous cars also fit all the same rules, will be adopted, it will be sigmoidal, and other reports have estimated that about half of all jobs will be replaced.
Anyways, for uploading a nematode, the optical I/O techniques to debug a living neural system I think are still in the laboratory prototype stage. Does anyone have this working in any lab animal anywhere? So it doesn’t even meet condition 1. And what’s the gain if you upload a nematode? Not much. Certainly not in excess of the tens of millions of dollars it likely will cost. Governments are disinterested as nematodes are not vertebrates. And there’s no “self-gain”, upload a nematode and no one is going to be uploading nematodes all over the planet.
There still will be progress, and with advances in other areas—the general robotics mentioned above—this would free up resources and make possible something like a human upload project.
And that, if you had demonstrations of feasibility, does meet all 4 conditions.
-assume you have demonstrated feasibility with neuroscience experiments that will be performed and can “upload” a tiny section of human brain tissue.
- The gain is you get to charge each uploaded human all their assets accumulated in life, and/or they will likely have superhuman capabilities once uploaded. This gain is more like “divide by zero” gain, uploaded humans would be capable of organizing projects to tear down the solar system for raw materials, or essentially “near infinite money”.
- Governments will have to push it with all-out efforts near the end-game because to not have uploaded humans or AI is to lose all sovereignty
- Adoption curve is trivially sigmoidal.
I don’t know when all the conditions will be met, but uploading humans is a project similar to nuclear weapons, in terms of gain and how up until just 29 months! before detonation the amount of fission done on earth by humans was zero. In 1900 you might feel safe in predicting no fission before the end of the century like you do now.
Also, you can use this method to disprove personal jetpacks or fusion power.
feasible- personal jetpacks, no, rocket jetpacks of the 1960s had 22 second flight times.
fusion power—no, no experiments were developing significant fusion gain without a fission device to provide the containment pressure
gain—no. Jetpacks would guzzle jet fuel even with more practical forms that worked more like a VTOL aircraft, and the value of this fuel is going to exceed the value of the time saved to almost all users. Fusion power is a method to boil water using high energy laboratory equipment and is unlikely to be cheaper than the competition over any feasible timescale.
government- no. Jetpacks cause extreme noise pollution and extra fires and deaths from when they fall out of the sky. Fusion is a nuclear proliferation risk as a fusion reactor provides a source of neutrons that could be used to transmute to plutonium.
sigmoidal—no, you can’t have this without large gain. Maybe this criterion is redundant.
If you got this far in reading, one notable fault of this proposed algorithm is it does not predict technologies requiring a true breakthrough. You could not predict lasers, for instance, as these were not known to be feasible until the 1960s when the first working models existed. That’s a real breakthrough. The distinction I am making is that if you do not know if physics will allow something to work, or if physics will allow something to work well, then you need a breakthrough to get it working.
Ditto argument for math algorithms, neural networks I would say are another real breakthrough, as they are much “better” for how little we know what we are doing than they should be.
We do know that physics will allow us to build a computer big enough to emulate a brain, to scan at least the synaptome of a once living brain, and get some detail on the weights. We also know that learning means we do not really have be all that exact.
This doesn’t give any real help in guessing the timing. But I think the curve to imagine is much closer to a step function than it is to a linear slope. So not seeing an increase just means we haven’t reached the step, not that their is a linear slope that is too small to see.