Artificial Intelligence - The More We Know IT, The Lesser We Seem To Know About IT
Sodexo Benefits and Rewards Services builds distinctive Employee Experience and seamless Mobility and Expense Management solutions both for within and beyond the workplace, proven to boost engagement and drive business success.
Does life mirror fiction? Or is it the other way around?
Artificial Intelligence is breaking all myths and suspending beliefs! Actions that were once upon a time considered dependent on human intelligence are today performed by machines with inbuilt and self-learning intelligence. AI has now become an inevitable part of our daily life, right from switching on lights to suggesting routes based on traffic conditions and even recommending best food options based on our dietary needs, effectively conquering every part of our life from personal space to work space and beyond.
So how do we differentiate between computer programs and AI? What is the line that differentiates both? Is AI a slightly sophisticated computer program? Not really, because a computer program tells the computer exactly what to do whereas AI teaches the computer how to learn and make its own decisions.
A majority of AI applications that we see today are Artificial Narrow Intelligence (ANI) where the machine is able to mimic human behaviour considering a defined or narrow set of parameters in a context. Narrow or weak AI systems, commonly used nowadays, are usually limited to specific tasks, such as voice and speech recognition, search recommendation engines etc and include virtual assistants such as SIRI.
The next level of evolution for AI is to move to Artificial General Intelligence (AGI) which is more advanced where an AI machine’s action is indistinguishable from that of a human. Strong or general AI is one able to find solutions without human intervention if presented with multiple and unfamiliar tasks. AGI is more a fiction today.
Optimistic experts believe in a third level of maturity - Artificial Super Intelligence (ASI) where AI no longer mimics but instead surpasses human behaviour and actions.
Whatever the type of AI level of implementation, there are 6 simple technologies adopted to create an AI application.
1. Expert Systems - Emulate the decision making ability of a human expert
2. Machine Learning - Automatically learn and
improve from experience without being explicitly programmed
3. Natural Language Processing- Enable computers to understand and process human languages
4. Computer Vision - An interdisciplinary field concerning how computers can see and understand digital images and videos
5. Automated Speech Recognition - Use of computer hardware and software based techniques to identify and process human voice
6. AI Planning - Related to strategies and action sequences.
AI requires paradigm shifts in thought and action as we propel to move in the era of digital transformation
While the race to implement and develop applications harnessing the power of AI continues, Sodexo is currently in the centre of this digital innovation. Sodexo Benefits and Rewards India aims to enrich employee experience by providing quality of life services to its clients, consumers and merchants. The brand has implemented AI to strengthen its digital vision in quite a few ways;
1. A dynamic chatbot provides all relevant information about Sodexo’s solutions and services
2. An ability to provide the right food choices and offers to our consumers based on the learning and analytics derived from food habits is in the evolution
3. 100 percent digital employee benefit solution that harnesses the power of AI to simplify the process of employees’ spends and claims
Numerous use cases to provide an employee experience by new and existing AI applications are evolving rapidly in Sodexo India and globally.
While AI has seen a lot of success in the recent few years, there have also been many solutions that have failed. And even though many experts including Stephen Hawking have expressed their cautionary view on evolution of this technology, it is evident that this technology is going to remain and evolve more to reach its high level of maturity.
AI systems are designed to do what you tell them to do, in theory at least, and more often than not, in practice too. Yet, the critical aspect of its control and safety cannot and must not be ignored. What if, as is shown in movies, control of weapons is taken over by AI? Or an automated power-grid is hijacked? Detractors of AI point out various pitfalls such as:
1. Control of the AI system being hijacked resulting in cyber-crime or crashes
2. Malicious programming of AI systems to cause mass damage
3. AI enabled weapons of mass-destruction in wrong hands. Or an AI powered arms-race.
4. Even AI systems programmed to be helpful can cause damages if goals are not aligned.
AI is a developing science, with breakthroughs and innovations happening faster than envisaged. The quest to design and build smarter AI systems, capable of doing all that a human can and more is still a grey area, fraught with risk. Theoretically, it is possible that a super-intelligence can be developed which can benefit humans by helping eradicate disease etc. But what if the goals of such super-intelligence and humans are not common and in tune? What if AI outpaced human intelligence? Beyond the basic concern that AI systems are designed and programmed by humans and therefore subject to bias, prejudice, insensitivity etc, and that they perpetuate unemployment, our inability to fully comprehend what can be possible is perhaps the biggest ground for caution.
While the capability, potential and prowess of AI cannot be denied, implementing the same responsibly is in our hands. Whether we perceive it as an excellent tool capable of improving lives and bring about positive changes, or as a threat to humankind, especially when it is left unchecked and uncontrolled is up to us. AI requires paradigm shifts in thought and action as we propel to move in the era of digital transformation.