{"id":4370,"date":"2025-08-08T20:20:37","date_gmt":"2025-08-08T20:20:37","guid":{"rendered":"https:\/\/musictechohio.online\/site\/gpt-5-demo-dumb-errors\/"},"modified":"2025-08-08T20:20:37","modified_gmt":"2025-08-08T20:20:37","slug":"gpt-5-demo-dumb-errors","status":"publish","type":"post","link":"https:\/\/musictechohio.online\/site\/gpt-5-demo-dumb-errors\/","title":{"rendered":"GPT-5 Launch Demo Plagued With Catastrophically Dumb Errors"},"content":{"rendered":"
\n
\"OpenAI's<\/div>\n

OpenAI’s GPT-5 is finally here and already powering ChatGPT, but it hasn’t made a great first impression<\/a>.<\/span><\/p>\n

In a livestream dedicated to the release, OpenAI tried to show off its newest large language model which CEO Sam Altman called a “significant step along the path to AGI”\u2014 but instead turned heads with some catastrophically dumb errors.<\/span><\/p>\n

Across several examples, bar graphs intended to show off GPT-5’s awesome performance benchmarks, while appearing professional-looking, turned out to be horribly inaccurate nonsense upon closer inspection.<\/span><\/p>\n

The gaffes were flagged on social media<\/a> and <\/span>highlighted by <\/span>The Verge<\/span><\/i><\/a>. The <\/span>most egregious example<\/span><\/a> is a bar graph comparing coding benchmark scores <\/span>for GPT-5 compared to older models. Somehow, the bar for GPT-5’s score of 52.8 percent accuracy is nearly twice as tall as the bar for a score of 69.1 percent for the o3 model. Even more bafflingly, the 69.1 percent bar is the exact same size as another bar representing 30.8 percent for GPT-4o. Make it make sense!<\/span><\/p>\n

\n

this screenshot from GPT-5 livestream has to be among the worst chart crimes of the century pic.twitter.com\/HXsK2CWCon<\/a><\/p>\n

\u2014 Ege Erdil (@EgeErdil2) August 7, 2025<\/a><\/p>\n<\/blockquote>\n