GPT-4o Advanced Voice is So Good it's Scary...

▼ Link(s) From Today’s Video:

-------------------------------------------------

▼ Extra Links of Interest:

Let's work together!

Thanks for watching Matt Video Productions! I make all sorts of videos here on Youtube! Technology, Tutorials, and Reviews! Enjoy Your stay here, and subscribe!

All Suggestions, Thoughts And Comments Are Greatly Appreciated… Because I Actually Read Them.

00:00 Breaking News: OpenAI's GPT-4.0 Voice Mode Launch
01:17 AI News Roundup: MidJourney and RunwayML Updates
02:24 ChatGPT Advanced Mode: Features and Access
05:42 Real-World Tests: ChatGPT's Voice Capabilities
08:51 Multilingual and Accent Capabilities
12:17 Fun and Practical Applications
21:30 Vision Mode: A Sneak Peek
23:50 Conclusion and Final Thoughts

[affecttheeffect]

I remember how difficult it was for the first DOS based personal computers to simply translate text into a robotic sounding voice, that was back in 1978 when I got my first Challenger C2-4P PC. I would never have predicted this voice capability 46 years later, it's so exciting. I'm 74 and wish that I have another 20 years ahead of me so that I can see where this is all heading. If I don't make it that long I'll leave knowing that you guys are in for a wild ride!

[skillerbg]

-Talk without breathing.
-How about you freaking do it

[Casey]

Youtube really needs to fix their bot problem. It's getting beyond a joke now.

[harlycorner]

"Call your friend" option in "Who wants to be a millionaire?" game, anyone?

[rebeccamiller]

The end of fall?? Crazy.. They definitely announced it too early.

[sportscommentaries]

I just got access to this, and wow, its amazing, worth the wait

[askalds]

Me being a psychopath : Count from 10 trillion to zero as Dracula.

[LydianMelody]

I have SO. MANY. music experiments I wanna do!! Starting with checking if it has perfect pitch (it must, right??). Then “Make me a chord chart for the song that’s playing.” “Check out this mp3 for a song I’m producing. Any tips for improving the mix?” “Listen to my piano playing. What should I focus on to improve?” “Does this arpeggio sound sweep-picked to you? Can you write the guitar tabs?” Just to start!

[설리-ow]

The crazy thing is even in the transcription the pauses and breaths are not captured at all when transcribed! I don't even think it's prompted to take pauses I think its just trained on so much human speech it just mimics it perfectly idk?

[duffthepsych]

I already have a great use case in mind for when I get this - I'm in a subfield of psychology that involves doing a lot of assessment of cognitive abilities like memory. It's always tough to find time to give new trainees the opportunity to practice the assessments on real people. Using this voice mode, we would be able to say, "This trainee is going to administer you a series of cognitive tests. Please perform these tasks in a realistic way, providing some correct and some incorrect responses so that the trainee will have a change to run through the entire administration and score your responses." (these can take 3+ hours, and it would be amazing to have a test subject that doesn't feel fatigue)

Alternatively, "Please play the role of a 75 year-old man who is coming in for an evaluation for Alzheimer's disease. You will hear a trainee asking you some questions as part of a clinical interview. Please remain in character and provide realistic responses to each query."

[jupreindeer]

I do love the ability for this to teach foreign languages with an encouraging touch and that need to breath. Fantastic. Reminds me of that one piano piece I asked that Udio to make and it insisted on putting in some background prep that sells it being a studio live recording.
I'm also reminded of that running joke when time traveling Beavis got his hands on a (then) modern smart phone and had 'interactivity' with that variant of an interactive voice. Imagine that joke working, now, with this technology in place. That is how fast our world is moving. After all, that B&B movie came out over a year ago.

[drowzy]

Yeaahhhh im not paying $20 a month for the next 3 months for a low chance of getting access. Ill just wait til its fully out.

[gaggix]

I believe that the model was trained on any type of sound and then finetuned to only produce a voice (depending on which one you select).

[carlosamado]

As a portuguese, i can tell you both the portuguese from portugal and brazillian were terrile. It kept doing some spanish accent and say spanish words.

[BionicAnimations]

I'm going to go to sleep. Can you wake me up when the end of fall is here?😁

[Henrique_Moura]

18:20 as a Brazilian, this really sound like a Spanish speaker talking in Portuguese, I think he got that he needed to stay with the Spanish accent, but nice to see how it deals with my language!

[beofonemind]

i feel u my dude.... some of these milestones are mind blowing. I'm 45 and i am so excited about it.

[BionicAnimations]

I just checked my phone just now, and I have it. 😁Anyone else?

[MerlinDerMagier]

The implications are huge, but I‘m also very worried about what it will do to society… Most „helpful“ inventions like the internet, smartphones, cars, etc. were all pretty devastating to society.

[elro_katz]

14:53 That's when I realized we're not talking to AI we're talking to an OpenAI service representative