Debunking OpenAI Fake GPT 4o Demos. This is Not OK
Can we talk about the new era of fake demos? How companies promise features to paying users, launch the thing, and those features are not there, not for everyone, not at the same time, or they simply don't work. A lot of can be forgiven to a startup, but I see these new practices coming from the leading AI companies which set an example what is okay, and responsible. As Content Creators we are almost expected to report on demos, and announcements. But often we report same marketing messages or down right lies. Can we go back to: Tell us what you have, not what you will have. Release then you have tested it, and saw that it works. Simple. The GPT-4o Use Cases I am showing are here: https://openai.com/index/hello-gpt-4o/ I am glad to see on OpenAI Dev Forum, that I am not the only one getting frustrated with this: https://community.openai.com/t/chatgpt-4-or-4o-is-not-doing-what-i-ask-for/759805 In this video, we're debunking 17 fake demo use cases of GPT-4o that just don't work. If you're curious about GPT-4o and its capabilities, this video is a must-watch! 00:00 OpenAI & Fake Examples of GPT-4o 02:16 Consistent text & Consistent image (NOPE) 05:35 Consistent characters (NOPE) 06:26 Insert yourself in a poster (NOPE) 07:56 Typography with iterative Editing (NOPE) 08:28 Audio file creation (NOPE) 09:30 Photo Caricature (MEH) 09:58 Text to Font (3x NOPE) 10:16 3d Objects (NOPE) 11:26 Live Testing Gone Rogue What do you think? Thank you for watching!
