Computers and computing resources used to be only for professionals. Later, after the advent of the personal computer, individuals could use and operate their own computers, getting the privilege and limitation of the equipment they owned. Ethernet, Internet and other network technologies have since made limited resource sharing between home computers (and others) possible and peer-to-peer networks enable effective horizontal communication without centralized administration. The Gaia Network is the next logical step.
Right now, the home computer is still going strong. Individuals own equipment and associated computing resources. They usually share very little of it to others, yet they might browse Wikipedia with a computer that has a dual core 2 GHz processor and 4 GB of main memory, with 90 % of the 300 GB hard drive unused. In a way, they are using a supercomputer to play Space Invaders. At the same time, when they need even more resources, they need to buy more equipment; it's the only way. They might have other computers, too - desktop computers, laptops, mobile phones, media players - but they cannot offer their resources to boost the computer that is in need.
What if Internet was replaced with a universal network enabling all modern computing devices to share their resources? They would make a self-organizing network, automatically sharing everything they can do, to always provide the most effective computing ecosystem possible. They would be sharing CPU power, processes, memory, storage, specialized interfaces, users... to a common good, to enable feats that are not possible without cooperation. Wherever you would go, your devices would make new friends and assemble a powerful virtual team to do work, retrieve information, process data - to service your needs as a user. Isn't that how humans (optimally) interact?
Dropping one node out of The Gaia Network would mean less resources to share but others would take its place automatically with no visible effect to the result. Because Gaia would constantly organize itself in the most efficient way, bottlenecks around central nodes (as is typical with Internet) would not matter so much.
A turbo boost to computing. A powerful shadow of virtual power. A dream team always around. A virtual exoskeleton in a cloud.
Except that it will not happen.
The owners and operators of computing devices do not want to share their resources. The network is not ready for the amount of traffic and there is no centralized authority to design and implement Gaia. Resource allocation and sharing is not a trivial thing to do and Gaia would need it to be smooth and invisible - and effective, of course. Jealosy is another thing. If some use the power of Gaia to copy new movies, should there also be somebody (or something) to say that it is wrong, or should be allocated less resources? What about abuse, DDOS, privacy issues, copyright issues, anonymity? And think about the children, will you?
The current way of computing seems like the work of a control freak if compared to Gaia. We have our own equipment bought with our own money, used for our own purposes (even if we only use 5 % of its resources), and sharing it might even be illegal even if it was possible.
Cloud computing and distributed computing are, of course, already reality but they fall short of Gaia, a holistic computing ecosystem. The utopia.
A system among voluntary participants, focused on certains tasks, over the importance of which all agree on, is already realistic and feasible concept-wise. On the field of robust, scalable transparent distributed scheduling and load-balancing, however, requires further scientific efforts.
ReplyDeleteEven though I think this "Gaia" is utopia, some facets of it might be feasible to implement in military, for example, to assemble "battlefield supercomputers".
ReplyDeleteBut it is known that the most powerful military on this planet - that is the U.S. military - has really difficult situation with hundreds of separate, proprietary systems that do not always have a common interface.
However, certain situations would need a common framework for self-organizing communication networks and that would be a very interesting starting point for thinking about possibilities.