by Richard Lynch, on Mar 19, 2019 11:57:27 AM
By Seth Earley with Richard Lynch
Many people consider Artificial Intelligence (AI) to be a threat to jobs – something to replace humans. Automation has always made manual processes more efficient but there is a human in the equation someplace. There may be fewer people required for a given task or process but automation tools free people up for more productive tasks. The problem with the so called “jobless recovery” in the US is that while efficiency has increased, more can be done with fewer workers. This trend will accelerate as AI and robots become more capable.
In the short term, robots will not be able to handle complex business tasks that have unpredictable elements, require creative problem solving or entail judgement based on significant real world experience. Knowledge can be codified but use cases still need to be considered for most bot approaches. It is costly to account for edge use cases – they need to be considered through explicit examples or through large data sets for machine learning.
Hybrid applications where a human handles complex tasks and a bot responds to clearly defined tasks and requests seems to be the more cost effective approach with today’s state of technology. As the system gains more experience by “observing” the human in complex conditions, increasingly difficult processes can be offloaded. This will be transformational in many industries.
Training Bots
Training data is always an important factor when it comes to AI systems. The type of training varies with the target problem being addressed. If the problem is anomaly detection the algorithm can just look at the data. But even then, the developer must have some idea of what type of anomaly the AI is looking for. Training for cognitive applications – the ones that people will interact with – is different than AI that looks for patterns in data or tries to identify images, move through complex environments or solve other problems that cannot be explicitly programmed. A chat bot helping a banking customer needs the correct answer. It cannot learn from making mistakes or decide over time which content is most appropriate for the banking customer. There would be a lot of unhappy customers and regulators if that were the case.
In fact, the training for large systems like Watson is the costliest aspect of deployment. The head of IBM marketing suggested when referring to cognitive computing and AI systems that “these are not application development projects. These are information projects. That means you need to define the problem you are trying to solve, find the information to solve that problem, get that information into the system, then develop the tool”. When speaking with vendors of chat bots or support applications, they typically expect that the client has data and knowledge to support the target application or process. In many cases, that is the fundamental problem – that the knowledge is not codified or represented in a format that is easily consumable by the system. Here’s were transformation teams can help: training content must be developed. The interesting insight is that good training content for bots is also good training content for humans. Therefore, making the content consumable for humans is a step toward training content for cognitive applications.
AI, Cognitive Computing & Public Sector/ Economy
AI/Machine Learning is creeping into our lives impacting the economy, our work life, consumer products and services and public safety and government services. The impact will be felt in changing job roles and requirements, shifting skill sets and further continued economic disruption as more sectors are automated.
Better Healthcare, Improved Education, High Quality Services
The upside to AI will be in a variety of areas. In healthcare, machine intelligence will allow for faster and more efficient processing of research papers and clinical trial results with tools such as Watson ingesting more knowledge at a faster pace. This will lead to advances in areas such as personalized medicine, where an individual’s genetic makeup will be factored into treatments. Robotic surgery will continue to make significant advances and artificial limbs will enable the disabled to have new levels or freedom and capability.
AI also promises a safer world through reduction of accidents caused by human error as we shift to self-driving cars, improved security through threat identification and mitigation, more secure finances through “personal financial security guards” and better preparation for natural occurrences like extreme weather through more accurate forecasting.
Human welfare will be impacted in multiple areas as AI is embraced in behavioral health through cognitive therapy agents, personal health coaches, and psychiatric and social case management. Coaching, teaching and training offer exciting possibilities as skill and knowledge development approaches are customized to leverage personal learning styles, specific knowledge gaps, and optimized emotional interaction and engagement.
Though improved machine approaches will reduce dependency on trained lawyers, access to legal research and advice will be more democratized through legal services agents.
AI approaches will improve the read of the overall pulse of society allowing for rapid responses by the public and private sector to citizen and customer needs - bringing service to transformational levels of performance.
For a case study on how AI is transforming customer interaction centers
Read this blog to download the case study: Blog
About contributing author Seth Earley: Seth Earley is Editor, Data Analytics, IT Professional Magazine from the IEEE and CEO of Earley Information Science (EIS). His interests include Knowledge Strategy, Data and Information Architecture, Search-based Applications and Information Findability solutions. Seth has worked with a diverse roster of Fortune 1000 companies helping them to achieve higher levels of operating performance by making information more findable, usable and valuable through integrated enterprise architectures supporting analytics, e-commerce and customer experience applications.If your business model is changing, strategy becomes everyone’s job not just the C-Suite.