In recent years, the term “innovation” has been a leader’s obsession. Digital transformation has practically become a need for all businesses, especially since the pandemic has wreaked havoc on the global economy. Artificial Intelligence investments would reach $341.8 billion in 2021, up 15.2 percent year over year, according to IDC.
With these numbers, do you think companies can fail? Well, it is part of the deal too. According to IDC, approximately 28% of artificial intelligence projects fail. Here’s a sobering reality: Artificial Intelligence is science and not magic! Let’s check out some of the top company failures. So, we can learn our lessons.
1. Microsoft Tay
Microsoft chose to enter this market about six years ago. Tay, their Twitter chatbot, made its debut on March 23, 2016. While it appeared to be a good idea at first, things went horribly wrong. Twitter users exploited the bot’s inadequate natural language processing (NLP).
Also, they found ways to exploit its design flaws, causing it to learn and repeat unacceptable thoughts. Tay didn’t take long to start imitating some of the comments and phrases on Twitter, eventually making sexist, racist, and insulting insults toward other Twitter users. Microsoft shut it off permanently in less than 24 hours.
2. Amazon’s AI Recruitment Tool
Indeed, organizations are increasingly turning to artificial intelligence (AI) to assist them to find competent job candidates. Amazon created custom technologies in 2014 to automate and streamline the talent recruiting process.
A year later, the company’s machine learning researchers decided to delve deeper into their AI-powered recruiting tool. They discovered that the algorithm frequently recommended men for technical positions while rejecting women’s resumes when it did so.
Machine learning algorithms appear to have mastered the hiring process after examining applications submitted to Amazon over a ten-year span.
The majority of those resumes were from men because men have typically performed those roles. As the algorithms became more aware of the pattern, they began to favor men. Amazon attempted to make the program more gender-neutral, but it was ultimately abandoned.
3. Hong Kong Real Estate
You’re a Hong Kong real estate billionaire who needs someone to manage a portion of your money so you may expand your business. Instead of employing a financial counselor, you decide to invest in an artificial intelligence system. Every day, the robot drains money from your account to the tune of USD 20 million. We are not talking about fiction here.
Samathur Li Kin-kan experienced exactly that between 2017 and 2018. He sued the fintech firm behind the corporation for USD 23 million in an attempt to collect some of his losses, claiming the firm exaggerated the bot’s capabilities.
Common Mistakes behind Artificial Intelligence Failure
We know that the industry is booming, but knowing where AI is failing is also important. Let’s see the common reasons for AI failures in healthcare, IT supply chain, accounting, etc.
In the haste to install AI at their company, executives will often grab any data they can get their hands on and try to incorporate it into their machine learning program. They then question why it isn’t producing insights from that data. Data must be clean and correct in order to be actionable and AI-ready.
Frequently, business issues arise as a result of a flaw in an existing workflow or process. While AI may be able to assist in the resolution of some of these issues, it is not a quick solution. To really identify and correct inefficiencies, business process reengineering is frequently required.
Sure, your data science team may be able to complete an AI project by themselves. But what if the system configuration isn’t in line with your company’s needs? Your operations team should be involved in the project from the beginning.
True, AI is a fascinating technology, but artificial intelligence failure is also real. However, it’s easy to become so engrossed in technology that you lose sight of the fact that it’s being used by actual people.
An AI project can elicit apprehension in your departments, and particularly among employees who think that the technology will reduce or even eliminate their function. This uneasy personnel will be hesitant to learn and adopt new technologies if organizational change management is not prioritized.
Bottom line
If you don’t want to fail with your latest AI investment, choose one of the best ERP vendors. We, at DFSM, can lay out the perfect plan for AI technology implementation and analyze your company and prevent any project management failures.
So, if you want to harness the true power of AI, contact us now for a one-to-one consultation. Our experts are also available to guide you on the implementation of AI in healthcare and AI in supply chain management.
