Rick Collins, President, AI, Next IT, Inc.: To bot or not? Understand the role of artificial intelligence in the enterprise

In the past few months, robot programs have been a hot topic on the Internet.

Major companies in the technology industry such as Microsoft and Facebook are opening up their own artificial intelligence platforms to developers. These open tools and platforms mean that the industry has taken an important step toward the popularization of artificial intelligence applications and is also an exciting step.

However, for these companies, the various industry challenges faced by senior corporate executives in the past few years have been further exacerbated by this robot program boom. Many enterprising companies are aware that if companies need to implement an artificial intelligence strategy, they must use this technology throughout the business. However, the challenges they face are only just beginning.

Revolutionary technology

These companies understand that artificial intelligence is a major business integration and has revolutionary significance. It will eventually cover all aspects of business operations, from customer service, business intelligence and analysis, sales to customer relationship management (CRM), and even internal knowledge management. And Human Capital Management (HCM).

A lot of enterprising corporate executives are aware that the strategic decisions for deploying artificial intelligence will have an overall impact on the business in the coming years. Therefore, they will not take decisions lightly but adopt a wait-and-see attitude.

However, such decisions have now had to be put on the latest agenda. If, as Microsoft and Facebook announced, the robot program is a new application, it is reasonable to expect that soon we will witness the explosive growth of robotics programs and related e-shops, as if it were a Cambrian life without acquiescence. Big explosion. The new generation of applications will develop natural language processors, speech analyzers, and other intelligent functions that are collectively covered by artificial intelligence.

If you are just deciding which component of the future enterprise organization artificial intelligence will apply to, the many anxieties and confusion surrounding the robot program can now dampen your enthusiasm for the post-application era. Businesses are eager to adopt artificial intelligence in order to avoid being disconnected from this world, which is increasingly driven by machine learning and cognitive computing. Robot programs may provide them with an attractive channel to take the first step toward low risk.

Companies also need to be vigilant if they want to apply robot programs.

There is no doubt that bots will be used to process large amounts of data and will also apply and perform many tasks that support automation. For some small-scale work tasks that do not require humans or require a lot of cognitive work on the computer, the robot program is the perfect solution.

This is nothing new, but what we are faced with is a world of proliferation of robot programs. In fact, most of our initial interactions with the company may take place in some form of robot program. This means that companies need to develop a strategy and framework to understand when bots will be effective solutions within the enterprise and when they will need to integrate deeper into smart systems.

When many companies begin to study artificial intelligence and its role in the enterprise, robotics can be a powerful way. Unlike clumsy built-in applications and push notifications, robot programs and newsletter applications meet existing user needs. They are also a great way to start using the conversational user experience and form a business case in this process. Perhaps the most attractive thing is that they promise that they will be easily generated and maintained.

Robot programs and instant messaging applications may be what we call "hard-coded" intelligence, that is, the use cases for their services are pre-determined, and scripts are pre-written. For this reason, such programs and applications are extremely vulnerable. When you ask them to go out of their comfort zone, they become very picky.

Many companies are attracted to the rapid development of a robot program, but they will soon raise their demands. Developers like the concept of end-user experience, which ultimately requires a higher level of intelligence to implement the concept. For example, robot programs cannot tell themselves when they have problems. In particular, such procedures cannot be self-analyzed, which raises the question of whether they meet business objectives.

As companies delve into artificial intelligence and rely more heavily on automated tasks and communications performed by human-distribution robot programs, robotics programs also pose scale challenges. Unless one institution is preparing to increase the number of bots required for certain tasks (maintaining the number of escalations to meet the demand), the institution will soon discover that a more powerful intelligent system is needed in that type of system. In the middle, the robot program is just an endpoint for entering and communicating.

"More than one" guideline for corporate robotics

If you want to weigh, the robot program will not be a solution that effectively meets the company's needs. Consider the "more than one" criteria that I have inducted. It can help you. This thinking framework helps guide decision makers when robot programs are suitable for the automation of a task and when they need to improve intelligence.

Robot programs and instant messaging applications are call and answer mechanisms. They immediately received and responded in a timely manner. Most of them responded in a unified manner. Once your use cases begin to require real conversations at some point, it means that the discussion is going to continue, and naturally you want the user to explain it or make a choice later. This requires a further integrated solution. Robot programs and instant messaging applications are not good at real context conversations going back and forth. The kind of dialogue relies on domain expertise, and it requires deeper development of artificial intelligence than the study of natural language processing interactions.

Robotic programs and instant messaging applications do not run a sophisticated model of the underlying domain and language. With such models, they can perform complex tasks on behalf of users. Travel schedules are the most common examples of robot program applications. For more tasks such as insurance qualification enquiries, financial advice, and customer service inquiries, when your business process has more than one destination, or more than one acceptable result, the robot program needs help.

Robot programs and instant messaging applications may not be able to collate information from more than one recording system. They do not necessarily have this intelligence and can only work well on their own highly coordinated single database or subscription data. Faced with different data structures and/or different network services, robot programs and instant messaging applications will not perform well. Because different data structures or different network services are mixed together, they will inevitably be overwhelmed. For enterprises, because of the large number of recording systems, they must use artificial intelligence in order to achieve effective solutions and deliver the right information, so smart systems may need to be further integrated.

If cross-cutting is needed, robot programs and instant messaging applications cannot do anything. A vertical single-command model should be built, the number of commands is determined, and the common commands are from a statistical point of view. This method of modeling is much easier than teaching a computer how to cross multiple fields of knowledge. With this in mind, robot programs and instant messaging applications are best suited to meet the needs of a single service and a single environment.

Undoubtedly, the robot program will play an important role in the company in the next few years and become a more familiar form of interaction with customers. For companies seeking to apply artificial intelligence, robotics may be a good starting point because they begin to understand their own needs and begin to fully understand how much artificial intelligence can be used within the company. In this budding period of corporate artificial intelligence, it is equally important that policy makers think in terms of strategic long-term perspectives, how intelligent systems will integrate with all aspects of the enterprise, and allow companies to change their appearance.

VIA VB