From snail mail to email, DVDs to streaming, paper charts to electronic medical records, information creation, storage, communication and consumption are being digitized into the universal language of ones and zeros. As we digitize more information, businesses need better ways to crunch, analyse and feed data into decision-making and other activities. AI and machine learning provide us with the ability to more easily analyse and learn from the vast amounts of digital data that we create. Computers can crunch massive amounts of data quickly, and machine learning makes computer programs “smart” by enabling them to learn, predict patterns, spot anomalies, and recommend and implement new processes or activities based on data.
Machine Learning is sometimes brought up as a subspace of Artificial Intelligence, based on the concept that we can let the machines learn for themselves by providing them access to large amounts of data. On the other hand, Artificial Intelligence is the extended and wider perception of machines becoming capable of carrying out tasks in an intelligent way. Compared to the generalized Artificial Intelligence (a generalized AI system, in theory, can handle any task), applied AI is more suitable for next-generation communication systems as the applied AI system can be devised to adeptly controlling and optimizing the wireless networks. Unlike Machine Learning models, AI models reach out the world, accustomed to the changes and rebuild themselves. While machine learning is great for predictive analytics, AI goes beyond predictions and prescribe plans/suggestions with implications to realize a benefit.
For instance, you could teach a computer to differentiate fruits from vegetables by entering a series of photos with the right identification. Then, you could feed in a new set of photos for the computer to identify. The program gets smarter and better at completing this task as it analyses more data. If you use Netflix, you’re already using machine learning algorithms, which are analysing your viewing habits and comparing them to those of millions of other viewers to suggest what you might want to binge watch next. Likewise, smart cars use AI and machine learning to adjust settings, such as temperature and seat position, to driver preferences; provide advice about road conditions; report and fix vehicle problems, and drive themselves. AI and machine learning also power chatbots, which “converse” with users via a fast, friendly interface to get things done. For example:
• Pizza Hut’s chatbots let customers place orders and ask about dietary information, delivery areas and more via Facebook Messenger and Twitter.
• FedEx is using Amazon.com application program interfaces (APIs) to build an app that lets customers ship packages by saying, “Alexa, I want to ship a package.” On the back end, instead of searching through lists, tables and invoices, a FedEx employee might use a chatbot to tell Alexa how much a package weighs and where it should be shipped.
• Wynn Hotels plans to outfit nearly 5,000 hotel rooms with Amazon’s Echo device, which will enable guests to ask Alexa for the room, hotel and other information.
References:
- 04_artificial_intelligence.pdf. (sem data). Obtido de http://people.idsia.ch/~andrea/edss/lectures/04_artificial_intelligence.pdf
- Aprendizagem Automática | Coursera. (sem data). Obtido 22 de Maio de 2018, de https://www.coursera.org/learn/machine-learning
- Maini, V. (2017, Agosto 19). A Beginner’s Guide to AI/ML 🤖👶. Obtido 22 de Maio de 2018, de https://medium.com/machine-learning-for-humans/why-machine-learning-matters-6164faf1df12
- What Is Machine Learning? | How It Works, Techniques & Applications. (sem data). Obtido 22 de Maio de 2018, de https://www.mathworks.com/discovery/machine-learning.html
- What Is The Difference Between Artificial Intelligence And Machine Learning? (sem data). Obtido 22 de Maio de 2018, de https://www.forbes.com/sites/bernardmarr/2016/12/06/what-is-the-difference-between-artificial-intelligence-and-machine-learning/
theindustry4.wordpress.com 