OpenAI Faces Backlash Over Vibe-Graphing Technology

During its major GPT-5 livestream on Thursday, OpenAI presented several charts that appeared to showcase the model’s impressive capabilities—but a closer look reveals some inconsistencies in the data visualization.
Ironically, in one chart illustrating GPT-5’s performance in "deception evals across models," the scale seems misaligned. For instance, the onstage graphic claimed that GPT-5 with "thinking" achieved a 50.0 percent deception rate. Yet, OpenAI's smaller o3 model—with a reported 47.4 percent—appeared as a larger bar. The data published in OpenAI’s official GPT-5 blog post, however, shows GPT-5’s actual deception rate at 16.5 percent.
This visual mismatch meant that onstage, GPT-5’s bar was shown as taller than o3’s, even though its reported score was lower. Elsewhere in the same chart, o3 and GPT-4o received different scores but were represented by bars of identical size. The error was significant enough for CEO Sam Altman to publicly acknowledge it, labeling the incident a "mega chart screwup," while clarifying that the blog contained the correct version.
An OpenAI marketing staff member also apologized, stating, "We fixed the chart in the blog, everyone—sorry for the unintentional chart crime."
On Friday, responding to a Reddit user’s query about the graphs, Altman explained that “the numbers here were accurate, but we messed up the bar charts during the livestream. On another slide, we got the numbers wrong.” He confirmed that the blog post and system card data were accurate, adding that “team members were working late and exhausted, which led to human error. It all comes together last-minute for a livestream.”
Still, the errors presented an unfortunate look for the company on such an important launch day—especially as it promotes the new model's “significant advances in reducing hallucinations.”
Updated August 8th: Added Altman's Reddit comment.
Related article
Google Tests Remy AI Agent for Gemini as Focus Shifts to User Control
According to Business Insider, Google is testing Remy, a new AI personal agent for Gemini. This tool aims to execute tasks on behalf of users, streamlining both professional workflows and daily routines.Currently, Remy is undergoing testing in an int
Sam Altman Sparks Debate Over AI's Deceleration
Listen onApple PodcastsListen onSpotifyOpenAI CEO Sam Altman recently suggested that it may be time to “pace the rate of AI development” to allow society to “harden around some of these new capability levels.”On the latest episode of TechCrunch’s Equ
OpenAI fights Apple trade secret lawsuit
OpenAI rebutted Apple’s trade secret allegations on Tuesday, arguing the lawsuit is unfounded.“We take these claims seriously but see no evidence supporting them,” OpenAI stated, as reported by Bloomberg’s Ed Ludlow on X. “We support fair competition
Related Special Topic Recommendations
Comments (0)
0/500

During its major GPT-5 livestream on Thursday, OpenAI presented several charts that appeared to showcase the model’s impressive capabilities—but a closer look reveals some inconsistencies in the data visualization.
Ironically, in one chart illustrating GPT-5’s performance in "deception evals across models," the scale seems misaligned. For instance, the onstage graphic claimed that GPT-5 with "thinking" achieved a 50.0 percent deception rate. Yet, OpenAI's smaller o3 model—with a reported 47.4 percent—appeared as a larger bar. The data published in OpenAI’s official GPT-5 blog post, however, shows GPT-5’s actual deception rate at 16.5 percent.
This visual mismatch meant that onstage, GPT-5’s bar was shown as taller than o3’s, even though its reported score was lower. Elsewhere in the same chart, o3 and GPT-4o received different scores but were represented by bars of identical size. The error was significant enough for CEO Sam Altman to publicly acknowledge it, labeling the incident a "mega chart screwup," while clarifying that the blog contained the correct version.
An OpenAI marketing staff member also apologized, stating, "We fixed the chart in the blog, everyone—sorry for the unintentional chart crime."
On Friday, responding to a Reddit user’s query about the graphs, Altman explained that “the numbers here were accurate, but we messed up the bar charts during the livestream. On another slide, we got the numbers wrong.” He confirmed that the blog post and system card data were accurate, adding that “team members were working late and exhausted, which led to human error. It all comes together last-minute for a livestream.”
Still, the errors presented an unfortunate look for the company on such an important launch day—especially as it promotes the new model's “significant advances in reducing hallucinations.”
Updated August 8th: Added Altman's Reddit comment.
Sam Altman Sparks Debate Over AI's Deceleration
Listen onApple PodcastsListen onSpotifyOpenAI CEO Sam Altman recently suggested that it may be time to “pace the rate of AI development” to allow society to “harden around some of these new capability levels.”On the latest episode of TechCrunch’s Equ
OpenAI fights Apple trade secret lawsuit
OpenAI rebutted Apple’s trade secret allegations on Tuesday, arguing the lawsuit is unfounded.“We take these claims seriously but see no evidence supporting them,” OpenAI stated, as reported by Bloomberg’s Ed Ludlow on X. “We support fair competition





Home






