This just gave me idea of OpenAI releasing a ‘Touch’ module.
Multi-modality is quite addictive!
10K people waiting already for the stream to start in 2 hours. But there will be one hour of corporate messaging before the show starts, in case anybody is wondering about the starting time of the video feed.
Does anyone know the music being played right now before the event starts? It’s so good! Would love to study to it
I hope its new modalities but some improvements for models are good too.
What time does it start? How much more time to wait, with this amazing music. Already feeling sleepy .(watching from china)
Waymo pile-up near office - Altman delayed?
It is supposed to start at 01:00 China standard time.
It sounds like poor quality, but cool if it was realtime, AI music.
AI music can be done much better. Will likely be the next thing to take over. Imagine music stream companies give their own royalty-free AI music when you go searching for anything by genre.
I have finished the Godzilla vs Kong, and the event hasn’t started yet, watching form Dundee, Scotland.
Will join the Zoom call with a delay while preoccupied and acting like I am present…
This is amazing! Do you know if the new App for Desktop can help us Handle Assistants too? Where do i find the app?
wow, the voice input output modality was so freaking spontaneous, mind blown.
Thanks for your update @sashirestela
yeah, just tested it. The speed is just wow…
More info about GPT-4o
Gpt-4o is fucking wild. Wow. The latency is incredible. Just tried it compared to GPT-4v. Looks like it’s already available for Assistants as well.
I think this is the first time OpenAI has said “available now” and it’s ACTUALLY available now.
Don’t like the “free” fact. Where will the money come from then? I still feel like this is a stepping stone towards a search engine.
Hey guys, I imagine the use coupled to other apps like WhatsApp, electronics and also as emergency support. Amazing Openai team!!!
I just tested the API ‘gpt-4o’ and it immediately returned a response, with console confirming both gpt-4o call and automatic normalizing to the current ‘gpt-4o-2024-05-13’ model too.