{"id":19081,"date":"2023-12-09T19:17:07","date_gmt":"2023-12-09T19:17:07","guid":{"rendered":"https:\/\/afreeurl.com\/?p=19081"},"modified":"2023-12-09T19:17:10","modified_gmt":"2023-12-09T19:17:10","slug":"openai-investigates-complaints-of-lazy-gpt-4-in-google-reviews-x","status":"publish","type":"post","link":"https:\/\/afreeurl.com\/?p=19081","title":{"rendered":"OpenAI investigates complaints of “lazy” GPT-4 in Google reviews, X"},"content":{"rendered":"

<\/p>\n

OpenAI, the company that launched ChatGPT just over a year ago, recently took to social media to address concerns about GPT-4’s “lazy” performance on social media and Google Reviews.<\/p>\n

Screenshot from X, December 2023<\/span><\/p>\n

The move comes after growing feedback from online users, which even includes a one-star Google review on the company’s Google Business Profile.<\/p>\n

OpenAI provides insights into training chat models, performance evaluations, and A\/B testing<\/h2>\n

OpenAI, via its Twitter account @ChatGPTapp, detailed the complexities involved in training chat models.<\/p>\n

\"xatgptScreenshot from X, December 2023<\/span><\/p>\n

The organization stressed that the process is not a “clean industrial process” and that variations in training sessions can lead to noticeable differences in the AI’s personality, creative style and political bias.<\/p>\n

Comprehensive AI model testing includes offline evaluation metrics and online A\/B testing. The final decision to launch a new model is based on a data-driven approach to improving the “real” user experience.<\/p>\n

OpenAI Google review score affected by GPT-4 performance and billing issues<\/h2>\n

This explanation comes after weeks of user feedback about GPT-4 getting worse on social media like X.<\/p>\n

I don’t know if anyone else has noticed this, but the performance of GPT-4 Turbo is significantly worse than standard GPT-4. <\/p>\n

I know it’s in preview right now, but it’s much worse.<\/p>\n

\u2014 Max Weinbach (@MaxWinebach) November 8, 2023<\/a><\/p>\n

It has been debated whether GPT-4 has become “lazy” recently. My anecdotal evidence suggests that may be true.<\/p>\n

I repeated a sequence of old analyzes that I did with Code Interpreter. GPT-4 still knows what to do, but keeps telling me to do the work. One step now there are many and some are strange. pic.twitter.com\/OhGAMtd3Zq<\/a><\/p>\n

\u2014 Ethan Mollick (@emollick) November 28, 2023<\/a><\/p>\n

Complaints also appeared on the OpenAI community forums.<\/p>\n

\"userScreenshot from OpenAI, December 2023<\/span><\/p>\n

The experience led one user to leave a one-star rating for OpenAI via Google Reviews.<\/p>\n

In addition to GPT-4’s “degenerate” performance and lack of intelligence, some Google reviews noted billing and customer support issues.<\/p>\n

\"OpenaiScreenshot from Google Reviews, December 2023<\/span><\/p>\n

Interestingly, the most liked Google review from five years ago aligns with recent rumors<\/a> about a volatile workplace, claiming that OpenAI is a “discreet environment. Unfriendly. Toxic workers.”<\/p>\n

\"GoogleScreenshot from Google Reviews, December 2023<\/span><\/p>\n

Glassdoor’s most helpful reviews of OpenAI suggested that employee frustration and product development issues stem from the company’s shift in focus to profits.<\/p>\n

\"OpenaiGlassdoor screenshot, December 2023<\/span>
\n\"OpenaiGlassdoor screenshot, December 2023<\/span><\/p>\n

This incident provides a unique insight into how customer and employee experiences can impact any business using various review and rating platforms.<\/p>\n

\"openaiGoogle screenshot, December 2023<\/span><\/p>\n

Google SGE highlights Google’s positive reviews<\/h2>\n

In addition to occasional complaints, Google’s reviewers recognized the revolutionary impact of OpenAI’s technology in several fields.<\/p>\n

The most positive review mentions about the company appear in Google SGE (Search Generative Experience).<\/p>\n

\"GoogleScreenshot from Google SGE, December 2023<\/span><\/p>\n

conclusion<\/h2>\n

OpenAI’s recent insights into training chat models and the response to public feedback on GPT-4 performance illustrate the dynamic and evolving nature of AI technology and its impact on those who rely on AI platform.<\/p>\n

Especially people who just received an invitation to join ChatGPT Plus after being on the waiting list while OpenAI paused new subscriptions and updates. Or those developing GPT for the upcoming GPT Store release.<\/p>\n

As AI advances, professionals in these fields must remain nimble, informed, and responsive to technological developments and the public’s reception of these advances.<\/p>\n

Featured image: Tada Images\/Shutterstock<\/p>\n