AI has an increased use in the working environment as it is seen as a means of leveraging production.
Smart machines and applications are steadily turning into a daily development, helping us to make faster, more accurate decisions and with more than 75 percent of businesses investing in Big Data, the role of AI and machine learning is ready to extend dramatically over the next five years. As of 2017, a quarter of organizations are spending 15 percent or more of their IT budget on machine learning capabilities. With cloud computing providing organizations, machine learning can hit the mainstream and drive innovation in each sector (Mitchell Feldman,2017).
AI enabled by cloud
In order to develop “intelligence”, Deep Learning algorithms require access to massive amounts of data. Cloud Computing is enabling access to “Big Data”. Centralized storage and processing of “Big Data” in data centers is powering the current generation of AI applications.
In 2020, the digital universe is expected to reach 44 Zettabytes (equal to 1 Billion Terabytes). Data valuable for enterprises, particularly unstructured knowledge from IoT (Internet of Things) devices and non- traditional sources, is projected to increase both in absolute and relative sizes (IPR Corporation). Worldwide public cloud services market revenue is projected to grow 18.5% in 2017 reaching $260.2 B. Cloud Computing market projected to reach $411B by 2020(Forbes,2017).
According to Sundar Pichai, CEO of Google Inc., computing is evolving from a mobile-first to an AI-first world. The cloud delivery models of PaaS and SaaS will benefit more due to Cloud –AI wave. After cloud being a significant disruption within the technology market, AI is on its path to prove a major face-changer or cloud market worldwide.
2.1 Areas of Artificial Intelligence
2.1.1 Expert System:
Expert system is the specialized branch of Artificial Intelligence refers to the mechanism which has the capability of collecting core data, process on it, analyze, make synthesis, perform operations, and provide the correct and accurate results and helps in taking best decisions.
Expert System has three major components: –
- Knowledge Base: It is the storage of IF-THEN rules which are designed as per the knowledge derived from the human experts.
- Inference Engine: It retrieves rules from knowledge base for the problem being solved
- User Interface: It is the mechanism by which the user interacts with the Expert System through dialog boxes, command prompts, forms and other output windows.
Expert system in Banking Sector:
Suppose if the conditions for the client to get loan is, that he/she has been working for a minimum of one year and receive a basic monthly financial gain of at least 2,500 euros and the minimum age ought to be 22 years. The Expert systems based on IF-THEN rules decides the sanction of loan.
IF A=OK AND B=OK AND YEAR_JOB >= 1 THEN
C=OK LOAN=OK FIND LOAN_N
ELSE DISPLAY “{NAME} YOU ARE WORKING IN THIS JOB LESS THAN ONE YEAR
YOU CAN NOT TAKE HOME LONE” LOAN=NO C=NO;
The above rule states that if the age and salary meet the requirements and the year of job is more than or equal one year then you can get loan and if it is less than one year display message that you cannot take loan.
Expert systems increase speed for completing complex task and provides quality customer services by reducing errors, cost and training time. The system is quicker than employee of the bank.
2.1.2 Machine Learning:
Machine learning is one of the major breakthrough in AI development since it permits software applications to become more accurate in predicting outcomes while not being expressly programmed. The fundamental premise of machine learning is to create algorithms from input file and use statistical analysis to predict an output worth among an acceptable range. Most of the applications we interact with everyday are based on machine learning processes.
Google introduced “Rank Brain” a machine learning algorithm identifies the intent behind a user’s search and offers them tailored information on that particular topic. Google’s Rank Brain is responsible for 15% of all online searches and now it is one of the most important signal out of the hundreds of signals used to determine search ranking.
Google Maps use deep learning algorithms that extract the street names and house numbers from photos taken by Street View cars and increase the accuracy of search results. With over 80 billion high resolution photos collected by Street view cars, Machine learning mechanically extracting data from geo-located image and achieving an accuracy rate of 84.2 percent for most convoluted street signs.
Google introduced a smart reply function to Gmail to assist users tackle their inbox, with 10 percent of mobile user’s emails sent using this tool the following year. The smart reply function is based on two recurrent neural networks: one used to encode incoming mail, the other used to predict possible responses. These networks work together to decode the meaning behind each incoming message and to automatically suggest three different responses for each.
2.1.3 Robotics:
Robotics is the third pillar of AI. Building on the advances made in mechatronics, electrical engineering and computing, robotics is developing increasingly sophisticated sensorimotor functions. Robots can perform specialized autonomous tasks, such as driving a vehicle, flying in natural and man-made environments, swimming, carrying boxes and material in different terrains etc.
By 2020 the IFR (International Federation of Robotics) estimates that more than 1.7 million new industrial robots will be installed in factories worldwide (Frank Tobe, 2017). For every new industrial robot introduced into the workforce, six jobs were eliminated, a study from the National Economic Research Bureau.
Major robotics trends that are set to present opportunities to organizations:
- Robotics for Ecommerce, 45 percent of the 200 leading global ecommerce companies will deploy robotics systems in their order fulfillment warehousing and delivery operations by 2018.
- 30 % of commercial service robotic applications will be in the form of a robot-as-a-service (RaaS) business model by 2019. This will help cut costs for robot deployment.
- By 2020, 60 percent of robots will rely on cloud-based software to outline new skills, cognitive capabilities, and application programs, resulting in the formation of a robotics cloud marketplace.
- By 2020, organizations will have a larger selection of vendors as new players enter the $80-billion information and technology market to support robotics deployment. (Bob Violino,2016)
Example: Amelia – Virtual Agent for Enterprise Customer Service
Amelia reduced 60 percent of cost in IT solutions by automating repetitive tasks currently tackled by humans. Amelia control tasks that are currently performed by 250 million knowledge workers worldwide.
She has held more than 80,000 conversations to date, and resolved 69% of these queries end to end.
In the U.S., Amelia assists customers of a leading insurer by handling queries i.e. more than 3,000 users per week and achieves 93% accuracy.
In Sweden, Amelia is providing access to its 4 million retail banking customers for a leading Nordic bank and resolving more than 85% of the queries addressed to Amelia. (Business Wire, 2017)
Artificial intelligence is growing up fast, as are robots whose facial expressions can elicit empathy and make your mirror neurons quiver.” —Diane Ackerman
2018-1-7-1515298337