AI is a Double-Edged Sword
Artificial Intelligence (AI) is an increasingly trending buzzword these days. Discussion boards and articles are full of outlooks and predictions about the impact of AI on businesses and our personal lives. Some experts argue that AI is the next industrial revolution.
More so, AI will likely redefine current sociocultural patterns and the human role in an AI-driven world. Sci-fi movies such as 2001: A Space Odyssey, Ex Machina, and I Am Mother provide a speculative glimpse of what the long-term future might hold for us once AI acquires distinct human characteristics.
In general, we cannot deny that AI offers tremendous potential. Early successes promise a bright future for process automation, customer service, and content management. But AI has also a “dark side,” to borrow a famous phrase from Star Wars. Without ethics and regulation, AI could overwrite established concepts of morals and evolve into an overreaching force.
Stephen Hawking, Bill Gates, and Elon Musk are among several minds who expressed concerns about the potential impact of AI on mankind. In 2016, Microsoft’s Tay chatbot experiment lasted less than 24 hours before Microsoft shut it down because Tay’s Twitter conversations turned ugly.
In Tay’s defense, it was trolled by human users. Essentially, Tay gained extreme views due to unrestricted learning. The experiment illustrated that AI has no limits in an uncontrolled setting. Unless we supervise AI learning and provide a moral compass, AI could quickly embrace humanity’s worst traits and biases. Additionally, big data has the potential for greater abuse and breach of privacy.
Regardless of our personal views, AI is here to stay for practical reasons alone. The scale and speed at which we generate and utilize data will require more intelligent mechanisms. For now, at least, AI focuses primarily on augmenting traditional business practices and technologies.
A Fast-Growing Digital World
I recall when the Internet was the new hype. It quickly transformed how we communicate and interact with each other. Today, we can stream videos and music on the go, subscribe to channels, and shop online with ease. “There’s an app for that” has become the catchphrase of our digital world.
According to an infographic prepared by MerlinOne, 50% of the world’s population in 2017 were Internet users and 37% were active Social Media users. Furthermore, 90% of worldwide digital data in 2017 was created in the previous two years alone. In addition, 93% of organizations were using various forms of cloud solutions. More importantly though, our digital space is growing at a staggering 40% a year. These numbers highlight the pace and scale at which we are expanding our digital footprint.
It’s hard to imagine a world without the Internet. Within a short time, social media has blurred the lines between professional and personal users. B2B and B2C were the original models for e-commerce. Social media, digital marketing, and digital marketplaces brought businesses and consumers closer together. C2C and C2B have unleashed new possibilities to share and deliver content. And it is these interactions that generate, consciously or unconsciously, big data destined for AI.
AI Examples in A Digital World
Speech recognition offers some of the most tangible evidence of AI capability. Amazon’s Alexa, Microsoft’s Cortana, and Apple’s Siri are voice-enabled virtual assistants, to name a few. These devices combine AI and Natural Language Processing (NLP) to interact with users via voice command. They are part of the generation of Artificial Narrow Intelligence (ANI) devices that perform basic tasks and problem solving.
If we thought a few years ago that touch screens were a big step in revolutionizing device interaction, then voice-enabled AI is surely a disruptive technology. Just think of the many software-enabled devices and systems we use daily, both as professionals and consumers. Adding voice capabilities would significantly change how we use a wide range of current technologies.
IBM’s Watson virtual assistant is a functioning example of AI-enabled support for internal and external customers. These types of chatbots are redefining the automated delivery of customer service. Chatbots are extremely flexible, patient, and scalable. This is especially helpful to users with speaking and typing disabilities, or when providing global support. Plus, you never have to wait in line compared to a live human customer representative.
Overall, chatbots are proving to be a viable option that can supplement and even expand conventional human customer service. Many organizations are already utilizing AI-enabled chatbots to communicate with their customers.
Although marketers use segmentation for content marketing campaigns, most campaigns still involve one-size-fits-all mass marketing. In addition, human marketers often carry out follow-up and make complex decisions to move leads through the sales funnel.
Marketers might use tools to automate some activities, but these tools only enable rule-based Robotic Process Automation (RPA). They are not intelligent enough to provide a uniquely personalized experience on a large scale and predict customer behavior. AI can provide the needed scale and optimize content search and content matching in real time. For example, Amazon adjusts pricing of items based on when and how you access its online store. The online store also creates customized lists of products and reading materials based on your shopping activity and past purchasing behavior.
Retail is just one industry that offers immediate opportunities for AI. The manufacturing industry, too, has great potential for AI to make production processes run smoother. GE’s Predix Manufacturing Execution Systems (Predix MES) is an example of how AI and smart data improve efficiencies and system downtime.
Perhaps you have heard of a tool named Grammarly. I am neither affiliated with the company nor am I promoting the tool but thought that it is a good working example of how AI can automate copyediting. The tool’s brain, as described by the company, uses a combination of advanced learning and AI to help improve written content. As you write content, Grammarly suggests grammatical changes and offers better word choices to improve readability. You can also use Grammarly for existing content and perform a copyediting check.
One can only imagine how such tools could easily accommodate multiple languages, specialized vocabularies (e.g., by industry or topic), and apply different writing conventions. New generations of authoring tools could leverage AI to help content creators write better content.
Machine Translation (MT) has been around for a while but has gained greater momentum in recent years. As a result, many Language Service Providers (LSPs) utilize MT as appropriate to automate the translation of selected content. Although Neural MT shows advantages over the older Statistical MT, both approaches usually require human post-editing of translated content.
In 2018 Microsoft demonstrated success with AI-enabled MT by achieving near-human translation quality. The specific demonstration involved the translation of news content from Chinese into English. This was a significant breakthrough that highlighted that we might be closer to AI-enabled MT as a widespread technology.
Historically, though, one of the main challenges for traditional MT is to decipher context and content complexity. Moreover, traditional MT tends to produce better results for texts that use shorter sentences and simple grammatical constructs.
Nevertheless, we should see improved translation quality for diverse subject matters and content types as AI makes its impact on current MT engines. In addition, eliminating human post-editing should be a major milestone in the evolution of MT and real-time translations.
AI remains a somewhat broad term that people use to refer to different concepts and capabilities. Not all applications that we might associate with AI are enabled by AI. Therefore, it can be confusing when the term “AI” gets mixed up with rule-based and data-driven solutions.
Likewise, AI is changing the meaning and value of data. Although rule-based data processing has been around for a while (e.g., business analytics), truly intelligent data processing is still in its early stages for most businesses. The idea of a dedicated AI data strategy continues to evolve as businesses learn to recognize AI opportunities and how to apply AI.
For now, AI focuses primarily on enabling businesses to achieve greater efficiencies and providing a more personalized customer experience. As with any disruptive or revolutionary technology, adoption of AI will not occur simultaneously across businesses and industries. There seems to be a sense of urgency to give AI a try in one form or another. However, understanding AI and how to implement it might take time.
AI has also the potential to diminish the perceived value of a company’s current offerings. Therefore, being complacent could prove detrimental if industry peers embrace AI to gain a competitive advantage. Turning a threat into an opportunity is usually a smarter approach to dealing with disruptive change.
- Panchal, S. (August 26, 2019). What are the possible ways of achieving Artificial General Intelligence?
Retrieved from https://towardsdatascience.com/what-are-the-possible-ways-of-achieving-artificial-general-intelligence-91b08e51f270
- Marr, B. (March 19, 2019). How To Develop An Artificial Intelligence Strategy: 9 Things Every Business Must Include. Retrieved from https://www.forbes.com/sites/bernardmarr/2019/03/19/how-to-develop-an-artificial-intelligence-strategy-9-things-every-business-must-include/#2d06d0968360
- Grammarly, Blog (2019). 4 Ways Writing Technology Will Change in 2020, According to Experts
Retrieved from https://www.grammarly.com/blog/2020-predictions/
- Panchal, S. (August 24, 2018). Types Of Artificial Intelligence And Examples.
Retrieved from https://medium.com/predict/types-of-artificial-intelligence-and-examples-4f586489c5de
- Reavie, V. (August 1, 2018). Do You Know The Difference Between Data Analytics And AI Machine Learning? Retrieved from https://www.forbes.com/sites/forbesagencycouncil/2018/08/01/do-you-know-the-difference-between-data-analytics-and-ai-machine-learning/#62e96af35878
- The Florida High Tech Corridor (July 1, 2018). Artificial Intelligence: The Next Industrial Revolution.
Retrieved from https://floridahightech.com/magazine/artificial-intelligence-the-next-industrial-revolution/
- Peng, T., Sarazen, M. (March 14, 2018). Microsoft AI Achieves Milestone in Machine Translation.
Retrieved from https://medium.com/syncedreview/microsoft-ai-achieves-milestone-in-machine-translation-ba0cf8f26408
- Microsoft AI & Research (March 2018). Achieving Human Parity on Automatic Chinese to English News Translation.
Retrieved from https://www.microsoft.com/en-us/research/uploads/prod/2018/03/final-achieving-human.pdf
- UBS (March 2, 2018). The evolution of artificial intelligence.
Retrieved from https://www.ubs.com/microsites/artificial-intelligence/en/new-dawn.html/
- Wilson, L. (February 9, 2018). AI just at the early stages of showing real results.
Retrieved from https://www.healthdatamanagement.com/news/artificial-intelligence-just-at-the-early-stages-of-showing-real-results
- insideHPC, Blog (February 5, 2018). Artificial Intelligence: The Next Industrial Revolution.
Retrieved from https://insidehpc.com/2018/02/artificial-intelligence-industrial-revolution/
- Ashwini, A. (January 19, 2018). How To Create A Successful Artificial Intelligence Strategy. Retrieved from https://medium.com/swlh/how-to-create-a-successful-artificial-intelligence-strategy-44705c588e62
- Boyle, A. (December 27, 2016). Microsoft’s chatbot gone bad, Tay, makes MIT’s annual list of biggest technology fails.
Retrieved from https://www.geekwire.com/2016/microsoft-chatbot-tay-mit-technology-fails/
- Kelnar, D. (December 2, 2016). The fourth industrial revolution: a primer on Artificial Intelligence (AI). Retrieved from https://medium.com/mmc-writes/the-fourth-industrial-revolution-a-primer-on-artificial-intelligence-ai-ff5e7fffcae1
- Reese, H. (March 24, 2016). Why Microsoft’s ‘Tay’ AI bot went wrong.
Retrieved from https://www.techrepublic.com/article/why-microsofts-tay-ai-bot-went-wrong/
- Hsu, S. (September 3, 2015). Don’t Worry, Smart Machines Will Take Us With Them.
Retrieved from http://nautil.us/issue/28/2050/dont-worry-smart-machines-will-take-us-with-them