Bernd
05/30/2020 (Sat) 21:02:25
No.37076
del
05/30/2020 (Sat) 21:02:25
No.37076
del
There was also an experiment with a machine which was actually a system of many machines (64) joined together. So despite the miniaturization, size may grow, and with that energy consumption too.
The growing number of parts means that more parts can fail. In a human brain redundancy is so large, that a death of a cell is not noticeable. But in a human constructed machine the failure has to be noticeable so they can change the part, fix the machine. A great challenge is that how can be constructed a reliably working machine out of unreliable parts?
A second problem they faced back then, is the overcomplication. They had to avoid creating one that too complex to use. So ease of use is one of the points they try to plan along.
For the next 20-25 years the road is paved - says the article written in 1972 - and back then in laboratories those computers or at least their plans which we were going to use in the '90s had already existed, although it's acknowledged that unexpected discoveries might change things.
AI! Right after the birth of computers the guesswork started if we could make a machine think. Enthusiasm was big. Bunch of stuff were taught to computers, from playing chess to translating languages, but these weren't thinking, just executing algorithms. This set back the expectations, they realized, while the brain doesn't speak mathematics, computers only speak that. There were happy accidents, when they tried to translate they realized linguists don't know their languages deep enough. Then computational linguistics was born (the irony: I think that's the proper translation, the text says mathematical linguistics). They realized the key question isn't if the machine thinks, but how the human brain thinks, and the steps can be programmed for machines to follow.
The author says, the use of computers was decided by the realistic needs of the societies, and the future will remain the same. Exceptions will exist, solving theoretical problems, or smaller scale projects of those who dare to dream big, but the use on large scale will be the chief deciding factor. Computers will take off the load from humans, physical or intellectual routine work will be relegated to them, and the man will use his brain what computers cannot do: thinking. We can only wish.
The growing number of parts means that more parts can fail. In a human brain redundancy is so large, that a death of a cell is not noticeable. But in a human constructed machine the failure has to be noticeable so they can change the part, fix the machine. A great challenge is that how can be constructed a reliably working machine out of unreliable parts?
A second problem they faced back then, is the overcomplication. They had to avoid creating one that too complex to use. So ease of use is one of the points they try to plan along.
For the next 20-25 years the road is paved - says the article written in 1972 - and back then in laboratories those computers or at least their plans which we were going to use in the '90s had already existed, although it's acknowledged that unexpected discoveries might change things.
AI! Right after the birth of computers the guesswork started if we could make a machine think. Enthusiasm was big. Bunch of stuff were taught to computers, from playing chess to translating languages, but these weren't thinking, just executing algorithms. This set back the expectations, they realized, while the brain doesn't speak mathematics, computers only speak that. There were happy accidents, when they tried to translate they realized linguists don't know their languages deep enough. Then computational linguistics was born (the irony: I think that's the proper translation, the text says mathematical linguistics). They realized the key question isn't if the machine thinks, but how the human brain thinks, and the steps can be programmed for machines to follow.
The author says, the use of computers was decided by the realistic needs of the societies, and the future will remain the same. Exceptions will exist, solving theoretical problems, or smaller scale projects of those who dare to dream big, but the use on large scale will be the chief deciding factor. Computers will take off the load from humans, physical or intellectual routine work will be relegated to them, and the man will use his brain what computers cannot do: thinking. We can only wish.