00:00 GPT-40, a natively multimodal model, can process and output visual, audio, video, and text inputs without conversion. 00:14 Despite no release of GPT-4.5 or GPT-5, the new model's significance lies in its multimodal capabilities, available for free to everyone. 00:43 GPT-40 is available in ChatGPT, but new voice and vision inputs, as well as the desktop app, are not yet available. 01:11 First use case: Marketing graphics with precise text generation, demonstrated through various visual examples. 02:06 Poster creation for movies using GPT-40, generating detailed and contextually accurate visuals and text. 03:02 Brand placement with GPT-40, mapping logos and words onto objects with high precision. 03:29 Consistent character generation for games, comics, and storytelling, showcasing the ability to depict the same character in multiple contexts and poses. 04:28 Tutoring: GPT-40's voice interaction assists with math problems, providing step-by-step guidance without giving direct answers. 06:20 Interview prep: Vision capabilities used for coaching, helping users present themselves better. 07:29 Customer service: AI acts as both personal assistant and customer service representative, demonstrating a two-sided service capability. 08:25 Meeting summarization and engagement: GPT-40 interacts in meetings, providing relevant information and enhancing discussions with real-time data. 09:50 The potential of GPT-40's use cases will be fully realized when the complete toolset is available to everyone.
So, I'm not sure if anything but the text functionality is working yet in 4o. That said, I tried to have it add some text to an image and it used the code analyser to edit the image and add text as a new file. lol.
It's definitely real because no woman can fake enthusiasm all day long 😅. No woman can speak for hours without sipping some water or coughing. Much less someone can speak from two different devices with such fluency.
00:00 GPT-40, a natively multimodal model, can process and output visual, audio, video, and text inputs without conversion.
00:14 Despite no release of GPT-4.5 or GPT-5, the new model's significance lies in its multimodal capabilities, available for free to everyone.
00:43 GPT-40 is available in ChatGPT, but new voice and vision inputs, as well as the desktop app, are not yet available.
01:11 First use case: Marketing graphics with precise text generation, demonstrated through various visual examples.
02:06 Poster creation for movies using GPT-40, generating detailed and contextually accurate visuals and text.
03:02 Brand placement with GPT-40, mapping logos and words onto objects with high precision.
03:29 Consistent character generation for games, comics, and storytelling, showcasing the ability to depict the same character in multiple contexts and poses.
04:28 Tutoring: GPT-40's voice interaction assists with math problems, providing step-by-step guidance without giving direct answers.
06:20 Interview prep: Vision capabilities used for coaching, helping users present themselves better.
07:29 Customer service: AI acts as both personal assistant and customer service representative, demonstrating a two-sided service capability.
08:25 Meeting summarization and engagement: GPT-40 interacts in meetings, providing relevant information and enhancing discussions with real-time data.
09:50 The potential of GPT-40's use cases will be fully realized when the complete toolset is available to everyone.
Thanks! 😊
Imagine lawyers having a discussion about an upcoming case and ChatGPT has every bit of knowledge in its database. 😮
Blade runner meets minority report
I'm surprised you didn't touch on the real-time interpreting. Do you plan on a similar video with Google's announcements from I/O?
well presented, concise and clear. too many of other channels tend to ramble forever.
So, I'm not sure if anything but the text functionality is working yet in 4o. That said, I tried to have it add some text to an image and it used the code analyser to edit the image and add text as a new file. lol.
Sleeping on the ”AI Girlfriend” use case I see! 😂
2:31
Thanks
Those videos are almost too dystopian for me.
It's definitely real because no woman can fake enthusiasm all day long 😅. No woman can speak for hours without sipping some water or coughing. Much less someone can speak from two different devices with such fluency.
I have not been able to accomplish consistent characters? Am I the only one ?