Top AI chatbots are often at risk of making the same decision-making mistakes as humans, says new study<\/i><\/p>\n
OpenAI\u2019s ChatGPT gets billions of monthly visits from people curious about its transformative influence. But does the world\u2019s most popular AI chatbot think more like us than we are aware? Can it be just as biased?<\/p>\n
A study led by a team of Canadian researchers set out to answer this question by involving ChatGPT in several tests. Published in the journal \u2018Manufacturing & Service Operations Management\u2019, the findings showed that it made mistakes that humans also make in certain situations. The AI chatbot doesn\u2019t just process massive amounts of data, it \u2018thinks\u2019 like we do.<\/p>\n
The researchers put ChatGPT through 18 different bias tests. While excelling in logic and maths, it exhibited many of the same biases as humans when making subjective decisions. The biases include overconfidence, risk aversion, and the gambler\u2019s fallacy \u2013 the false belief that if an event has occurred several times before in the past, it will occur less often in the future. These biases remained consistent across different business scenarios, but the research team believes they could change as AI keeps improving with new versions.<\/p>\n
\u201cAs AI learns from human data, it may also think like a human \u2013 biases and all,\u201d commented lead author Yang Chen, assistant professor at Western University, in a news release. \u201cOur research shows when AI is used to make judgment calls, it sometimes employs the same mental shortcuts as people.\u201d<\/p>\n
The researchers also found that ChatGPT liked to play it safe, overestimate itself, seek confirmation and avoid ambiguity. \u201cWhen a decision has a clear right answer, AI nails it \u2013 it is better at finding the right formula than most people are,\u201d stated Anton Ovchinnikov of Queen\u2019s University. \u201cBut when judgment is involved, AI may fall into the same cognitive traps as people.\u201d<\/p>\n
\u201cAI isn\u2019t a neutral referee,\u201d added Samuel Kirshner of the University of New South Wales Business School in Australia. \u201cIf left unchecked, it might not fix decision-making problems \u2013 it could actually make them worse.\u201d<\/p>\n
Because of all the biases, the study authors assert that businesses and policymakers should be cautious about counting on AI for key decision-making. It may not eliminate human error, but automate it instead. \u201cAI should be treated like an employee who makes important decisions \u2013 it needs oversight and ethical guidelines,\u201d elaborated Meena Andiappan of McMaster University. \u201cOtherwise, we risk automating flawed thinking instead of improving it.\u201d<\/p>\n
The researchers recommend systematic audits with programming and reviews so as to help cut down on the biases that today\u2019s chatbots exhibit.<\/p>\n
\u201cThe evolution from GPT-3.5 to 4.0 suggests the latest models are becoming more human in some areas, yet less human but more accurate in others,\u201d explained Tracy Jenkin of Queen\u2019s University. \u201cManagers must evaluate how different models perform on their decision-making use cases and regularly re-evaluate to avoid surprises. Some use cases will need significant model refinement.\u201d<\/p>\n","protected":false},"excerpt":{"rendered":"
Top AI chatbots are often at risk of making the same decision-making mistakes as humans, says new study OpenAI\u2019s ChatGPT gets billions of monthly visits from people curious about its transformative influence. But does the world\u2019s most popular AI chatbot think more like us than we are aware? Can it be just as biased? A\u2026<\/p>\n","protected":false},"author":2,"featured_media":19665,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_scj_primary_category_id":130,"_scj_featured":true,"_scj_featured_from":"","_scj_featured_until":"","_scj_featured_order":153,"_scj_visibility_class":"current","_scj_layout_family":"standard","_scj_article_style":"","_scj_display_overrides":[],"_scj_intro_image_id":19665,"_scj_intro_alt":"ChatGPT replicates human decision biases in AI behavior study","_scj_intro_caption":"","_scj_intro_class":"","_scj_intro_float":"","_scj_full_image_id":19665,"_scj_full_alt":"ChatGPT replicates human decision biases in AI behavior study","_scj_full_caption":"","_scj_full_class":"","_scj_full_float":"","_scj_media_type":"","_scj_media_provider":"","_scj_media_external_id":"","_scj_media_url":"","_scj_media_poster_id":0,"_scj_media_width":0,"_scj_media_height":0,"_scj_media_aspect_ratio":"","_scj_media_description":"","_scj_gallery_items":[],"_scj_related_post_ids":[],"_scj_additional_authors":[],"scj_layout_family":"standard","scj_media_provider":"","scj_media_url":"","scj_full_caption":"","footnotes":""},"categories":[130],"tags":[2410,173,4523,2411,2412,2413,299],"class_list":["post-19666","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-articles","tag-ai-bias","tag-artificial-intelligence","tag-chatgpt","tag-decision-making","tag-ethics-in-technology","tag-human-behavior","tag-machine-learning"],"yoast_head":"\n