{"id":191624,"date":"2025-10-05T13:42:37","date_gmt":"2025-10-05T13:42:37","guid":{"rendered":"https:\/\/www.newsbeep.com\/au\/191624\/"},"modified":"2025-10-05T13:42:37","modified_gmt":"2025-10-05T13:42:37","slug":"openai-releases-list-of-work-tasks-chatgpt-can-already-replace","status":"publish","type":"post","link":"https:\/\/www.newsbeep.com\/au\/191624\/","title":{"rendered":"OpenAI Releases List of Work Tasks ChatGPT Can Already Replace"},"content":{"rendered":"

ChatGPT maker OpenAI has released a new evaluation<\/a>, dubbed GDPval, to measure how well its AIs perform on \u201ceconomically valuable, real-world tasks across 44 occupations.\u201d<\/p>\n

\u201cPeople often speculate about AI\u2019s broader impact on society, but the clearest way to understand its potential is by looking at what models are already capable of doing,\u201d the company wrote in an accompanying blog post<\/a>.<\/p>\n

\u201cEvaluations like GDPval help ground conversations about future AI improvements in evidence rather than guesswork, and can help us track model improvement over time,\u201d OpenAI added.<\/p>\n

It\u2019s one of the most straightforward attempts to justify its AI models\u2019 financial viability to date, following skepticism that the tech may prove to be a dead end<\/a>. Experts have often criticized the company\u2019s boastful marketing<\/a>, such as CEO Sam Altman claiming that its GPT-5 model had achieved \u201cPhD-level\u201d intelligence<\/a>.<\/p>\n

In \u201cearly results,\u201d GDPval found that \u201ctoday\u2019s best frontier models are already approaching the quality of work produced by industry experts\u201d \u2014 a clear shot across the bow at critics who say the tech isn\u2019t up to the demands of the workplace.<\/p>\n

The 44 occupations where \u201cAI could have the highest impact on real-world productivity\u201d included a litany of professions including real estate sales agents, social workers, industrial engineers, software developers, lawyers, registered nurses, customer service representatives, pharmacists, private detectives, and financial advisors.<\/p>\n

The specific tasks, as laid out in a paper<\/a>, range from creating a \u201ccompetitor landscape for last mile delivery\u201d for a financial analyst, assessing \u201cskin lesion images\u201d for a registered nurse, and designing a sales brochure for a real estate agent.<\/p>\n

Surprisingly, the company found that its competitor Anthropic\u2019s Claude Opus 4.1 was the \u201cbest performing model\u201d after being graded by industry experts across 220 tasks, followed by GPT-5, which \u201cexcelled in particular on accuracy.\u201d<\/p>\n

An extra powerful version of GPT-5, called GPT-5-high, was \u201crated as better than or on par with the deliverables from industry experts\u201d just over 40 percent of the time. GPT-4o, which was released more than a year ago, scored a mere 13.7 percent.<\/p>\n

To be clear, OpenAI is treading carefully around the subject of replacing human jobs altogether. Its language suggests that AI will \u201csupport people in the work they do every day\u201d instead of saying outright that anyone could soon be out of work because of AI. That\u2019s unsurprising, considering the negative optics of celebrating the loss of employment.<\/p>\n

At the same time, whether that\u2019s really an honest interpretation of the industry\u2019s motives and end goals remains dubious. AI executives have long boasted<\/a> about replacing human labor with AI \u2014 drastic cost-cutting measures that are already starting to backfire for some companies<\/a>.<\/p>\n

There\u2019s also good reason to take OpenAI\u2019s latest evaluation results with a massive grain of salt. We\u2019ve already seen the use of AI cause major headaches for software developers<\/a>, lawyers<\/a>, and even customer service representatives<\/a>, often requiring more human oversight, not less. <\/p>\n

Hallucinations, in particular, remain a major sticking point, undercutting the output of large language model-based tools, forcing users to spend more time combing over the output<\/a> of AIs for false information. <\/p>\n

And while AI often excels at generating bursts of text in a particular style, it\u2019s easy for it to go off the rails during longer and less predictable tasks.<\/p>\n

Real-world tasks are rarely \u201cclearly defined with a prompt and reference files,\u201d OpenAI admitted.<\/p>\n

\u201cEarly GDPval results show that models can already take on some repetitive, well-specified tasks faster and at lower cost than experts,\u201d the company wrote. \u201cHowever, most jobs are more than just a collection of tasks that can be written down.\u201d<\/p>\n

More on OpenAI: NBA Coach JJ Redick Says He Spends Hours Talking to His \u201cFriend\u201d ChatGPT<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"ChatGPT maker OpenAI has released a new evaluation, dubbed GDPval, to measure how well its AIs perform on…\n","protected":false},"author":2,"featured_media":191625,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[20],"tags":[256,254,255,64,63,105],"class_list":{"0":"post-191624","1":"post","2":"type-post","3":"status-publish","4":"format-standard","5":"has-post-thumbnail","7":"category-artificial-intelligence","8":"tag-ai","9":"tag-artificial-intelligence","10":"tag-artificialintelligence","11":"tag-au","12":"tag-australia","13":"tag-technology"},"_links":{"self":[{"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/posts\/191624","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/comments?post=191624"}],"version-history":[{"count":0,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/posts\/191624\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/media\/191625"}],"wp:attachment":[{"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/media?parent=191624"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/categories?post=191624"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/tags?post=191624"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}