This demo feels a lot like GPT-V. Like they've gotten a lot of the latencies down, but it's doing the same thing GPT was doing previously with transcription after silence detection and TTS of the output.

Is there a reason why Open AI and Google have events so close to each other?

This feature has been in iOS for a while now, just really slow and without some of the new vision aspects. This seems like a version 2 for me.

That old feature uses Whisper to transcribe your voice to text, and then feeds the text into the GPT which generates a text response, and then some other model synthesizes audio from that text.

This new feature feeds your voice directly into the GPT and audio out of it. It’s amazing because now ChatGPT can truly communicate with you via audio instead of talking through transcripts.

New models should be able to understand and use tone, volume, and subtle cues when communicating.

I suppose to an end user it is just “version 2” but progress will become more apparent as the natural conversation abilities evolve.

Yes, per my other comment this is an improvement on what their app already does. The magnitude of that improvement remains to be seen, but it isn’t a “new” product launch like a search engine would be.

Does it feed your audio directly to gpt4?To test it I said in a very angry tone "WHAT EMOTION DOSE IT SOUND LIKE I FEEL RIGHT NOW?" and it said it didn't know because we are communicating over text

No it's not the same thing, the link for this submission even explains that. Anyone who comments should at least give the submission a cursory read.

I did and regardless of the underlying technology it is, in fact, an improvement to an existing product - not something new from whole cloth.

If they had released a search engine, which had been suggested, that would be a new product.

I don't doubt this is authentic, but if they really wanted to fake those demos, it would be pretty easy to do using pre-recorded lines and staged interactions.

For what it's worth, OpenAI also shared videos of failed demos:

https://vimeo.com/945591584

I really value how open they are being about its limitations.

Let OAI actually be released to the masses. Then we can compare.

I’m not a big fan of announcing something but it not being released.

They say available for api but it’s text only. Can’t send audio stream to get audio stream back.

Time will tell. I’m holding my emotions after I get my hands on it.