Cloud is Making AI more democractic

I was really wondering whether AI will be available for small enterprises though I felt that it is becoming available for end consumers. End consumers like us have the google search and Google Maps in our hands all the time where extensive AI with machine learning capability is being used. The smaller businesses are at a disadvantage from AI usage perspective and obviously the businesses are an early investor in AI technologies. Let us look at three important things that enables an enterprise to take advantage of Artificial Intelligence. Three important factors are

(i) Algorithms : Smart Algorithms make the life easier for enterprises. As an example when Google Bought Deep Mind they did not buy it for gaming but for optimization. They optimized the power consumption in their data centers by around 40% by using the Deep Mind technology. Even though Deep Mind was a small company the fact that Google bought it means technology is available for only large companies from optimization perspective.

(ii) Data : The most necessary ingredient for AI to be successful is the data. Google - the most successful AI company has traded our data with the search engine service. Search Engine gave us empowerment and Google got the data. This resulted in huge amount of data getting aggregated , researched and presented to us in some form or other improving our experience. Data brought in experience value to the table apart from the usage value

(iii) Computational Power : AI requires huge computational power. Moore's law has been true until now and therefore computational power had increased dramatically with miniaturization. Huge amount of data for Machine algorithms has to be processed by computational systems. It is not uncommon to see AI developers using GPU systems instead of CPU powered systems. Thanks to the cloud. Enterprise access to bigger computational power had been very democratic. Both the large enterprises and small enterprises had an access to computational power. With the introduction of AWS lambda and the seconds based billing the computational power is getting cheaper with less cost of ownership and everything is based on consumption based mode.

Smaller enterprises still suffered from the lack of access to Algorithms as they cannot afford to hire expert AI developers as well as lack of data. Thanks to the Cloud power. All the four major cloud players are down with competing to give access to excellent algorithms. Let me start with IBM Watson which is delivered from cloud and though it may be expensive for a small enterprise it makes the access to AI more democratic. AWS released the Machine Learning SDK called the AWS Machine Learning. This is becoming available to one and all. AWS delivers AWS machine learning as a PAAS (Platform as a service) Service. Azure introduced the Azure Machine Learning in the PAAS format. Google provides the machine learning service with Google Machine Learning module namely Google MI.

While two major issues have been solved for the small enterprises the biggest problem remains the access to the data. Large enterprises have been using large data because of the traditionally they have invested in large storage. This puts the small enterprise at a disadvantage in AI. However there will be soon be data service providers delivering the data for small enterprises to take decisions. At some point in time data marketplaces will become true and cloud has enabled the same already. I hope you found this useful. Let me know what you think whether cloud is making AI democratic or not.