{"id":86574,"date":"2024-11-19T08:53:39","date_gmt":"2024-11-19T15:53:39","guid":{"rendered":"https:\/\/inmoment.com\/?p=86574"},"modified":"2024-11-20T13:36:58","modified_gmt":"2024-11-20T20:36:58","slug":"ai-hallucination","status":"publish","type":"post","link":"https:\/\/inmoment.com\/blog\/ai-hallucination\/","title":{"rendered":"Addressing AI Hallucinations for Improved Business Performance"},"content":{"rendered":"\n
Think about the last time you asked ChatGPT a fairly simple question but got an unexpected response. Perhaps it provided a factually incorrect statement or just misunderstood your prompt. The result is described as a \u201challucination\u201d, a growing concern for businesses using AI systems.<\/p>\n\n\n\n
An AI hallucination occurs when an AI system produces false or misleading results as facts. A popular example is a large language model (LLM) giving a fabricated answer to a prompt it fails to understand.<\/p>\n\n\n\n
Humans hallucinate when they see something that isn\u2019t there. While AI models don\u2019t \u201csee\u201d anything, the concept works well to describe their output when it\u2019s inconsistent with reality. These hallucinations are mainly the result of issues with the training data. If the model is trained on insufficient or biased data, it\u2019s likely to generate incorrect outputs.<\/p>\n\n\n\n
An AI system is only as good as the data you feed it. It doesn\u2019t \u201cknow\u201d anything beyond its training data and has no concept of fact or fiction. An AI model like ChatGPT has one goal: predict the most appropriate response to a prompt. The problem is that its prediction can sometimes be well off the mark!<\/p>\n\n\n\n
There are various types of hallucinations, based on what a model contradicts:<\/p>\n\n\n\n
Generative AI has made impressive progress in content generation. However, it\u2019s still capable of generating incorrect or misleading information. These hallucinations are a concern for AI in customer experience<\/a>, affecting individuals and businesses alike. Here are some common examples of AI hallucinations in real-world systems.<\/p>\n\n\n\n AI models sometimes generate text that is inconsistent with factual information. A famous example of this hallucination is Gemini\u2019s incorrect response in a promotional video. The chatbot, formerly Bard, was asked, \u201cWhat new discoveries from the James Webb Space Telescope can I tell my 9-year-old about?<\/em>\u201d <\/p>\n\n\n\n Gemini claimed that the JWST took the first image of a planet outside our solar system. This information is false since it was the European Southern Observatory’s Very Large Telescope (VLT) that took the first photos of an exoplanet back in 2004!<\/p>\n\n\n\n AI models may invent details or references that don\u2019t exist. For example, Google\u2019s AI Overview generated this response to a prompt asking how long one should stare at the sun for best health:<\/p>\n\n\n\n \u201cAccording to WebMD, scientists say that staring at the sun for 5-15 minutes, or up to 30 minutes if you have darker skin, is generally safe and provides the most health benefits.<\/em>\u201d<\/p>\n\n\n\n AI Overview states incorrect information here and wrongly attributes it to WebMD.<\/p>\n\n\n\n Similarly, speech-to-text AI tools that transcribe audio recordings are prone to hallucinations. For example, transcription tools tend to insert random phrases from their training data when they encounter a pause in the audio. <\/p>\n\n\n\n A concerning fact is that these phrases can be inaccurate and misleading, or even worse offensive and potentially harmful such as incorrect treatments in the case of medical transcriptions. Therefore, the inability of traditional AI tools to handle breaks in audio can have negative consequences for organizations.<\/p>\n\n\n\n A generative AI system may respond appropriately but still misunderstand your prompt. An example of this hallucination is asking ChatGPT to solve a Wordle puzzle. <\/p>\n\n\n\n While the system generates a coherent response, its solutions tend to be well off the mark. For instance, it may suggest a word that doesn\u2019t match the pattern of letters you provide as input.<\/p>\n\n\n\n Sometimes, AI models fail to respond comprehensively, leading to dangerous results. Once again, Google\u2019s AI Overview provides an example of this occurrence. It generated largely correct information when asked which wild mushrooms are safe to eat.<\/p>\n\n\n\n However, it failed to specify how to identify fatal mushrooms. It suggested that mushrooms with \u201csolid white flesh\u201d are safe to eat, but it didn\u2019t mention that some dangerous variants have the same feature.<\/p>\n\n\n\n AI hallucinations create challenges across various industries. Its inaccurate predictions and information hurt the customer experience<\/a>, impacting the business\u2019s reputation. Here are some of the problems these hallucinations cause in key sectors:<\/p>\n\n\n\n AI has become a significant part of healthcare workflows. Its ability to summarize patient information and even help with diagnoses is impactful. One of its most notable applications is transcribing medical visits. AI-powered transcriptions help doctors record and review patient interactions to make informed decisions.<\/p>\n\n\n\n It is vital to maintain accuracy and completeness in these transcriptions. A hallucination in the text would make it difficult to provide effective treatment and diagnoses. <\/p>\n\n\n\n For example, OpenAI\u2019s Whisper, an AI-powered transcription tool, raised concerns by inventing phrases during moments of silence in medical conversations. Researchers found that Whisper was hallucinating in 1.4% of its transcriptions. This is a significant figure given that the tool had been used to transcribe around 7 million patient visits.<\/p>\n\n\n\n Some hallucinations were in the form of irrelevant text like \u201cThank you for watching!<\/em>\u201d during a conversation break in the transcription. Other instances were far more concerning, including fake medication like \u201chyperactivated antibiotics\u201d and racial remarks. These hallucinations can have harmful consequences as they misinterpret the patient\u2019s intent, leading to misdiagnoses and irrelevant treatments.<\/p>\n\n\n\n In customer service, contact center AI<\/a> hallucinations can damage brand credibility. Customers won\u2019t be able to trust a business after getting an inappropriate response to their queries. <\/p>\n\n\n\n For example, a chatbot might give incorrect information about a product, policy, or support steps. Similarly, transcription tools often hallucinate phrases during pauses in agent-customer conversations. These hallucinations can provide an inaccurate view of the customer\u2019s experience, resulting in poor analysis that fails to solve actual pain points.<\/p>\n\n\n\n Therefore, your CX program will suffer if it\u2019s relying on inaccurate call center transcriptions. Despite your best intentions, a hallucination could be enough to cause customer dissatisfaction.<\/p>\n\n\n\n Unlike traditional tools, Pearl-Plaza\u2019s advanced AI-powered solution addresses this specific problem to ensure your CX team accurately records customer interactions. As a result, you can be ensured you\u2019re taking the right steps towards improving the customer experience.<\/p>\n\n\n <\/div>\n\n <\/div>\n\n<\/section>\n\n\n AI enables legal professionals to save time on research and brief generation. Generative AI models can help produce drafts and summarize key points. However, due to hallucinations, relying on these models for crucial information like legal references can be tricky.<\/p>\n\n\n\n A law firm was fined $5,000 after its lawyers submitted fake citations hallucinated by ChatGPT in a court filing. The model invented six cases, which the lawyers used to support their arguments without verifying their accuracy. These cases were either not real, misidentified judges, or featured non-existent airlines.<\/p>\n\n\n\nStating obvious errors or false information as fact<\/h3>\n\n\n\n
Making up information and references<\/h3>\n\n\n\n
Misunderstanding the prompt<\/h3>\n\n\n\n
Providing incomplete information or context<\/h3>\n\n\n\n
What Problems Does AI Hallucination Cause?<\/h2>\n\n\n\n
Healthcare<\/h3>\n\n\n\n
Contact Centers<\/h3>\n\n\n\n
Legal<\/h3>\n\n\n\n
Finance<\/h3>\n\n\n\n