In recent years, Generative AI has become widely used by companies big and small, most commonly in the form of chatbots. AI chatbots have attracted attention all around the world in the technological evolution of customer service, marketing, and user engagement. In fact, it is predicted by Gartner, Inc. that chatbots will become the primary customer service channel for 1 in 4 organizations by 2027.
From small startups to global giants like Google and Meta, tech companies are embracing AI-driven conversational agents to automate responses, reduce costs, and improve efficiency. However, despite their widespread adoption, AI chatbots are far from perfect. There have proven to be many flaws within these systems, however it does not stop companies from pushing forward with their usage.
Flaws in GenAI chatbots causes backlash
AI has become a very powerful tool for companies to utilize, especially when it comes to interacting with their users and customers. However, it is important for companies to note that flaws in their systems may cause consumer backlash. A survey conducted by YouGov of over 18,000 consumers across 17 markets showed that 54% of people feel that the company using the chatbot is to blame for mistakes and inaccurate information. Even so, tech giants such as Meta and Google have continued to push forward.
Tech giants using chatbots despite errors
Hallucinations are a common flaw in AI chatbot systems. This is the term used to describe false information that a chatbot outputs as if it were true. Meta’s chatbot, known as Meta AI, has experiences with hallucinations which can be very harmful to individuals’ reputations. One example is false information produced about tech entrepreneur Vivek Wadhwa, whom the chatbot had stated was accused of plagiarism and cited a blog post that did not exist. Meta’s response to errors like these is simply that new technology will not always be perfect and that they are trying to make them better.
In another attempt to keep up with the competition, Google released its own AI chatbot programmed directly into its search engine by Gemini, Google’s largest language AI model, called AI Overview. This AI uses data from all over the web, however is sometimes unable to distinguish between factual sources and ones of falsehood or satire. One instance of this flaw is when AI Overview suggested the mixing of nontoxic glue into pizza sauce to prevent the cheese from sliding off the pizza. This was taken by the AI from an old Reddit post that was meant to just be a joke and not a real recommendation. This is just one example of the errors and falsehoods put out by AI Overview.
Mistakes companies make when deploying chatbots
There may always be inherent flaws within AI chatbots, however there are many mistakes that tech companies make that increase the likelihood of these flaws coming to fruition.
1. Deploying chatbots too quickly
This is perhaps the most common mistake that tech giants make when it comes to AI chatbots. Companies like Meta and Google try to keep up with Microsoft and OpenAI when it comes to putting out the newest technologies. However, it is important to take the time and effort to ensure that the program is built well and responsibly and that it is tested vigorously. This way, companies can discover flaws before users do, ultimately reducing the trust with consumers.
2. Failing to conduct maintenance consistently
AI chatbots should be designed to learn from their interactions continuously and be regularly updated by human minds. It is crucial to be aware of when mistakes occur and take immediate action to optimize the data so that it does not happen again. Implementing a feedback loop where users can rate the chatbot’s responses can also help improve its accuracy over time.
3. Depending too much on the chatbot
While AI chatbots are very helpful when it comes to receiving quick answers and assistance, companies should be sure that responsibility is not solely placed on the bot. This is especially crucial in the early stages of the chatbot’s release, as expecting too much from it will set up the company and users with potential dissatisfaction. Having humans available for additional customer service can help satisfy customers and reduce their frustrations.
AI chatbots represent a powerful tool for tech companies, offering the potential to revolutionize customer service and streamline operations. However, to fully harness their benefits, businesses must recognize and address the inherent flaws in these systems.
Sources:
Comments