Experiment on the Cloud

Due to undecidability of the halting problem, outcomes of many computations can not be predicted by a fixed Turing machine. Direct computational experiments are often employed to study properties of specific computations. Many computations are done locally, but it would be nice to perform them on the cloud so that they are accessible everywhere. Wolfram Programming Lab is a good example of computational experimentation on the cloud. Since computations can be written into apps, Microsoft Azure, Google App Engine, Apple Xcode Cloud, etc. may also be employed to experiment with code.

Legendary particle physics computations are a great place to start with. Eventually, general computations may attract experimenters' attention, due to intricate patterns computation may produce. It's difficult to formulate a theory about computational patterns beyond Turing completeness, but nevertheless interesting patterns that resemble those in nature may be found by computational experiment. This is the essence of A New Kind of Science. Although Principle of Computational Equivalence is fallacious, simple programs and the complexity they generate are valuable to science.

Many cloud computing platforms require users to install a SDK. However, OS upgrades often break SDK. Cloud platforms without SDK, like web apps, are almost non-existent. Although virtual machines are good SDK-less solutions, there should be cloud computing platforms without SDK for ease of use.

Wide accessibility aside, the advantage of cloud computing is its capability to perform heavy computations that are too resource intensive for a local machine. Machine learning is a typical example. For concreteness, cloud-based medical diagnosis better than contemporary medical gadgets may one day help doctors achieve superior accuracy. Medical services are carried out with SOP, which is very suitable for cloud AI automation. Combined with home cloud and IoT, medicine at home may eventually reduce stress at hospitals.

Cloud computing offers great promises. It's up to people's imagination.

Comments

Popular posts from this blog

Limitations of Knowledge Engines

Maps and Yellow Pages

數位革命:行動裝置與雲端