Mindful automation: these might not be the bots you're looking forNews / Distribution | Technology | OnlineBy Marisa Garcia | April 9, 2018Share This article was originally published on Sabre has published its latest Emerging Technology in Travel report, which looks at the top technology trends driving travel and social change over then next few decades, and considers both the applications and implications of this technology.Here, tnooz looks at some of the highlights that resonate most in Mindful Automation but the study is comprehensive with other valuable insights on authenticity in digital, and blockchain. Human factors Don’t be distracted by AI, no matter how sexy it has become. As Sabre says, AI is more a fuzzy concept which helps us express the evolving intelligence of the digital systems that will support automation.Sabre likens the advancement of mindful automation to the industrial revolution, finding parallels in the socio-economic impacts of each. There is no doubt that automation will re-shape the workplace and our lifestyles. As our digital systems begin managing both logical and physical tasks, jobs will be re-defined and the skillsets required for work will change. “As a paradigm shift, it may be helpful to stop thinking about business strictly in terms of jobs and roles, focusing instead on tasks and skills. Every job is made up of tasks; some of those tasks are going to be automated. Yet every worker has a set of skills, and those skills can be applied to a wide array of tasks. Valuing employees for their skillsets instead of simply filling particular roles is critical to maintaining and training—and retraining—a workforce.”Share this quote AI is a tool, not an employee In automation, the travel sector will find solutions to many of the infrastructure and staffing challenges that will arise with an increased number of people traveling in future, Sabre finds. Just as automation helped accommodate the growth of passengers flying by helping the airline industry move from manual reservations and flight management systems to today’s advanced PSS and GDS systems, so too will future AI-backed systems support growth. “On the front end, travel requires a tremendous physical infrastructure—cars and planes, roads and runways, hotels and high-rises, restaurants and resorts, ports and parking lots—as well as the people to staff and maintain everything. On the back end, travel has a similarly complex infrastructure handling inventory, reservations, fulfillment, staffing, identity, security, compliance, transactions (in hundreds of currencies and point systems), translation (in hundreds of languages), and countless other details often invisible to travelers.”Share this quote It is most likely that, as Sabre suggests, automation and AI will become the new tools that workers use to perform jobs, rather than replacing workers altogether. “Travel agents in particular may spend far less time searching through rates and optimizing itineraries and can shift their time into building and utilizing the skills to engage customers in providing a higher level of customization and service. The rise of boutique travel agencies over the past few years is demonstrating the demand for better service and more nuanced trip planning. In many cases, boutique agencies are embracing automation (through services like chatbots) to help with rote tasks and travel reminders, freeing agents to have more time to engage with travelers at any point in their journey.”Share this quote AI, Sabre suggests, will require human intervention for the foreseeable future. “The more complex an algorithm becomes, the more likely it is to be asked to operate with imperfect information. A calculator knows it has perfect information which leads to predictable, replicable correct results. In contrast, a weather algorithm is never exactly right; it can be approximate and useful, but perfection is impossible (it would require, at a minimum, modeling every atom in the system). The problem with more complex predictions and intelligence is that it’s hard to know for sure which data is important to make the right prediction. And because we don’t know which data to account for how to prioritize the data, we can’t expect a perfect outcome. Trusting AI with increasingly complex real-world tasks—increasingly human tasks—requires accepting imperfect outcomes. Adopting higher-intelligence AI algorithms will require humans to reevaluate acceptable levels of risk and continually reassess our tolerance for imperfections.”Share this quote Machine learn optimus cosi tutti-frutti Any suggestion that machine learning is a panacea to advancing AI misses the importance of human oversight to validate the data which is fed into the algorithm and also to filter through what the algorithm learns from the data.This is a process which those currently engaging in the process will testify can include surprising misinterpretations and odd extrapolations by the ghost in the machine.A good example of this is found in the post by AI Weirdness on the unusual Christmas carol compositions of a developing neural network. Practical AI applications will require significant and time-consuming human scrutiny. But AI has its uses in processing massive volumes of data. “One yet-to-be-realized promise of AI and automation is truly personal preferential filtering. This is a key opportunity for the travel industry to use AI: to work toward personalizing every aspect of the journey for the traveler, making every trip a bespoke experience.”Share this quote AWESOM-O, your robot friend Despite helpful bots and physical robots, human concierges and agents will play an important role in building trust in travel. Sabre predicts that automated service agents will become tools that enhance human service. “A hotel chatbot or in-room digital assistant can serve as a frontline for guests, answering basic questions, making spa and dinner reservations and recommending local attractions, etc. The digital concierge can alert a human concierge of opportunities to improve a guest’s stay. Similarly, a concierge bot should have the capacity to monitor social media channels to flag opportunities for a human concierge to both solve potential problems and affirm positive guest experiences. Augmentation of human capacity and human skills is a key role for future automation.”Share this quote Sabre also offers an important cautionary note of the brand image and legal risks of relying too much on automation for personal services. “One under appreciated value of the human concierge (and similar service roles throughout travel) is the ability to act with discretion and without complete data support. All digital assistants are powered by data—the more specific data they have about a person, the better their algorithms can optimize recommendations and predict user needs and behavior. Digital assistants need data to be effective. "However, data privacy laws prominently in place in Europe and emerging in other parts of the world, along with increased concerns over data breaches and cybersecurity, could complicate the efficacy of digital assistants. Many people are skeptical of how widely they share their personal data, and how widely their personal data may be used without their knowledge.”Share this quote Baby, you can’t drive my car No discussion of automation is complete without considering its future impact on transport. Sabre suggests that it will grow in proportion to how quickly cities build the new infrastructure required. “In the real world, adoption will be very uneven across the globe, at least for the next few decades.”Share this quote Recent incidents with automated transport trials show that there is a risk to blending automation and humans—often as a result of the unpredictability of human behavior. For automated transport to work at its best, cities will have to be smart first, with better signage and better pathways.