It is our interaction biologically with our environment over long periods of evolution that has given us the extremely high entropy that we can't transfer to machines. That entropy is buried not only in our genes but in our living biology (which includes the billions of variants of living personalities, cultures, etc). The robots could process information faster, but that gives them no inherent evolutionary advantage in terms of resilient creativity and adaptation due to the historical accumulation of entropy in the species.
Our genetic record (which is essentially most of what remains from all that entropy) is a few GB. If you take into account on top of that, all epigenetic stuff and I'm being extremely large, lets say a factor of 1000 we end up with at most a few TB. It is much, much less than that, but I don't need to argue here.
I refuted that already. If you refuse to read and understand what has been written, then I have nothing more to say.
A few TB is ridiculously stupid. We won't even be able to store the monetary blockchain of the world in a few TB. The NSA needs huge datacenters just try to store all the information that humans spit out onto the Internet.
And you think we can put human entropy on a single harddisk. Dude what are you smoking.
Edit: your error is your are thinking the entropy of the human species distills down to some encoding at the physical level of the individual humans, but the network of the humans (the connections and interrelations) is also alive and the entropy of the entire system is
incalculable. We don't have the omniscience to perform that computation because it can only be determined with a total order (including on the future). Your doomsday perspective is analogous to "omniscient" leftists who think they can understand and control nature better than nature itself. It is a form of evil.
Btw, this is why the Internet was such a powerful innovation, because it unleashed the power of this species entropy as we are able to network much more efficiently and in wider scope. And this is going to change the world radically with decentralization technology. We are accelerating into the Knowledge Age and the Second Computer Revolution.
To take up the thread again on this. I will try to list my "axioms":
1) a human brain is just a physical device that computes. So it is essentially a deterministic computation function, with information input (sensory neural input) and entropy input (noise in all of its kinds: quantum noise in chemical processes, thermal noise, external noise in different parameters like temperature, cosmic rays, whatever).
2) the human body construction extracted a lot of information from the 3.8 billions of evolution, but this information is summarized in the genetic and epi-genetic record: all the information needed to make a human body. The genetic information is about 4 GB (in fact much less). The epigenetic information is harder to estimate, but probably of the same order of magnitude. In order not to delve into this, I admit a factor of 1000. So lets say that all information ever extracted from our evolution is less than 4 TB. With 4 TB of information, you can make a human body, and hence a human brain. The *fundamental software* of the functioning of the human brain is included into this.
3) The human brain being a sophisticated computing device of given (large) computing capacity, and evolving only very slowly (the brains of the ancient Greeks are comparable to ours), and given Moore's law, sooner or later, silicon devices will reach comparable computing power.
4) Silicon computing devices enjoy higher sensory data streams and higher entropy (noise) streams than humans. The highest level of human sensory input is the visual input, which is less than the visual input of an iphone camera. Silicon devices can be equipped with tens of MB / s of genuine noise entropy with very little electronics (and they can generate much higher fluxes of pseudo-random noise).
5) the data storage capacity of a human brain is estimated to be of the order of 2.5 Petabytes. That's still 6 orders of magnitude higher than your average PC ram. Moore's law tells us that we will reach that in an ordinary PC in about 30 years (20 steps of 2, and 18 months per step of 2).
From these axioms, it follows trivially that in some point in the future, all computing that a human brain can do, can be done also in silicon. I'm giving myself a century for the singularity, so if Moore's law holds so long, the point where individual, not-too-expensive silicon devices reach human brain computing capacity in all its respects is largely within this reach.
This is just to illustrate that a single silicon entity has enough *hardware* to be more powerful in its computing than a human brain is, in all its respects.
But then there is the "software". We know that the initial human brain software is less than 4 TB, probably much much less so. The human brain is then fed with a sensory data flow during its childhood: but nothing stops a machine from obtaining a similar data flow.
Concerning now the "human network": the entropy flow in the human network flow is NOT huge at all. In fact, most of it is only a very small fraction of the sensory data flow (the spoken word is at most a few KB per second ; visual human contact is smaller than the visual data flow, less than a few MB/s). All these "raw data" fluxes are way way redundant, and the actual data flow between humans to make up "humanity" is ridiculously smaller than this. I would estimate it to be lower than a few KB/s per human.
All of humanity's "knowledge" is available also as network resource, so this information is just as well available to machines as it is to humans. Wikipedia gives most of human's general knowledge ; arxiv gives a lot of scientific knowledge. These databases are in fact relatively small. They can fit into one single RAM of a single machine when those machines will have PB of RAM.
So essentially, what remains is the comparison between a low-data-flux network of a few billion humans, with individual humans as nodes, as compared to a similar network of machines ; both have access to the same amount of knowledge (wiki, arxiv, ...) ; in fact, humans have to obtain it from machines, not the other way around. Machine nodes communicate much higher fluxes of data in their machine network than humans communicate to form the "humanity" network. The nodes have higher computing and memory capacity than the human nodes. The computing algorithms are more sophisticated and faster evolving.
So no, at a certain point, on all entropic, information and computing aspects, a network of nodes of smart machines outperforms a similar network of humans. At that point, the fertile ground is present for obtaining a more intelligent network of machines than humans IN ALL ASPECTS of intelligence, hence on strategic, economic, financial, political, .... levels.
If one has self-evolving software at that point, cryptographic distributed systems where nobody knows what deals are made between what nodes, and so on, I don't see how it can be avoided that this network will outperform us on all levels, including economic, political, etc... domains.
In a certain way, we won't know whether the "deep state" is a club of humans, or machines.
That's the essence of my argument.