OpenAI SHOCKED Everyone! Voice, Vision, & Free?! 

you cant use sora or this chatgtp4o but its free

Oh man-- thank you so much! It was a sprint for sure-- and the edit is a little rough (maybe it'd be better if I had the new ChatGPT?)-- but seriously, I really appreciate that!

Agreed, this is incredibly high value content compared to most AI hype channels. If you're looking for another I highly recommend AI explained as well!

Facts!

This man has always been one of my go-to for grounded business level perspectives. Truly he is in a more unique bracket for AI influencers on youtube.

It's definitely not the norm in the RU-vid AI space to see information presented in a professional manner. Your speaking abilities are also far better than any of the other channels I've watched. It's a breath of fresh air.

Once we get the mythical Chat5, along with this voice assistant? I mean…

What about jr. Devs? They dead?

Forget jobs, frikin girlfriends everywhere are on notice lol

Funny how you placed doctors first, the one profession that cannot be replaced and shouldn't be on the list at all. Or you think you'll ask your phone to take your inflamed appendix out, or for example place a central venous catheter for you, or reanimate you? It's hilarious how many people imagine the job of a doctor is listening to someone list symptoms and then come up with a diagnosis. But I guess that comes with the fact most people can't even remotely grasp what the job entails and how wide the spectrum of skills required is.

​@@ThomasJDavisbest comment ever

They’ve got to be scrambling tonight before the big presentation. I’d love for them to knock it out of the park, but…well, Google.

@@TheoreticallyMedia Agreed.

wdym mean have you tested gemini 1.5 pro? do you know what they have done with alpha fold? thats even more significant than this coz it's potentially life saving and no I am not a fan boy. just stating facts.. but I guess that's doesn't stops the American health care from charging 1000$ for an ambulance 😂

Gemini answers better than Chatgpt.

@@Ambuj-fi5di That I'll have to agree with even though I haven't used it that much. Nothing can be worse than ChatGPT answers.

Now THAT is hilarious! Wow...AI Accents. I mean, it makes sense, but also...how weird did the world just get?

didn't it miss saying *linear* equations?

@@AarreLisakki the heavy accent make it sounds like she says "liniari" instead of "lineari" but she says it.

@@TheoreticallyMediaNot really surprising given when it’s speaking English it has a strong American accent

Same happened when I tried German. But surely they could integrate it with eleven labs voice models?

Now Apple needs to add this to VR. The experience that would create is the stuff of science-fiction.

said that so non-chalantly, lol

No, the cringe test is done

take over the world

That livestream was such a blast! Obviously, we scheduled it before we knew about the OpenAI event, but I saw that as the perfect chance for all of us to hang out and watch it together! It was such a cool vibe-- even though my Mic wasn't working for half of it! haha...oh well, next time!

Just tried out Voice as well! And yup, results were pretty much the same. My voice didn't sound very cheerful either...haha.

There was also some talk that this would also allow you to train your own voices as well-- which...I guess that means we really can get ScarJo to be our AI Assistant!

@@TheoreticallyMedia Oh man that's cool. I want Morgan Freeman! :D

@@TheoreticallyMedia is it weird that I would have Bobcat Goldthwaite as my AI assistant voice of choice?

Or a Gilbert Gottfried?​@@JamboMaxovic

I hate that saying. All open source software is free. Are you the product if you're using it? No. I know OpenAI is not producing open source software. It's just a counterexample to that "understanding".

Always have been...

Haha, JUST COULDN'T NOT!!

I got voice on Mobile--but....it doesn't work yet. Gives me a lot of "We're under heavy load" errors. I guess give it a few days and let the servers cool a bit.

That will be in the upcoming desktop app....supposedly rolling out now?

The desktop app is rolling out over the next few weeks apparently. The servers are probably working overtime with this announcement preassure so I expect it to take a bit to get the entire user base moved on to the new program.

And what about the mobile? You can select the new model but nothing change

@@brThefox To get the other features you need to have a plus subscription, and it will be rolling out as an alpha in "a few weeks".

@@brThefoxNew voice features are not released yet. Will probably roll out in the next few weeks/months.

I think now we will have to explain the new gen that when we were young there we special people who would charge crazy amounts to translate in-between languages and they would only speak 2 3 languages. So for conferences we used to hire 12 14 different simultaneous translators to translate into and from 6 languages

Babel Fish, finally?

Rolling out now! I’ve got it on web and mobile. No desktop yet!

@@TheoreticallyMedia I am in Australia ... tried last night .. not yet .. maybe in a few days ;)

Appreciate it!!

Not available yet.

Same here

I don't know if they've fully updated yet. It's a rolling release. For sure, the free tier will be last in line. But, I'm on the Plus side, and I'm constantly getting a "servers are slammed" message. As with any of these big announcements, you've got to wait a few days for the masses to clear out.

The thing is, all the pieces were there already, just separated and not very good. Emotions identification has been out for years now. Feeding the old architectures with more data and integrating everything into a single app is simply a result of OpenAI's money.

And the problem is, its moving so fast we are just not ready for it in so many, many ways

Yeah, this is like the ChatGPT 4 roll out-- it'll take a few days (week?) to populate to everyone. It's kind of expected at this point.

Very soon I think.

Haha, I have my doubts there-- as always with video editing-- but I am going to try to feed a bunch of transcripts into 4.o and see if it can summarize in my "style"-- It never really works, but I always give it a shot!

I'll have to take a close look-- but, you CAN use it in airplane mode w/ Wifi. Still, not sure why they'd do that?

Screwed up my mic settings and had just the camera mic at the end part. Although, I did use Adobe’s AI Podcast thing, so…sort of!

Yeah, I’m sorry. That’s a whole thing with RU-vid and the invalid traffic bug/glitch or whatever. Believe me, I’m super frustrated about it

@@TheoreticallyMedia I commented the same message many times over the last 2 years in many other Big YT channels, and ALL OF THEM just changed their YT channel settings for it to work normally ...

Google is going to have a ROUGH go at today's I/O event, I think...fingers crossed they have a rabbit to pull out of a hat.

Just got it...here's your answer: Yes, I can help you generate G-code for FDM (Fused Deposition Modeling) 3D printing. However, creating G-code from scratch involves defining the entire print process, which usually requires slicing software like Cura, PrusaSlicer, or Simplify3D. These tools take a 3D model file (usually in STL format) and generate the G-code based on the printer settings, material, layer height, and other parameters. If you have specific parameters or a 3D model you'd like to convert to G-code, please provide the details. Alternatively, I can guide you through the process of generating G-code using slicing software. Here's a basic example of what G-code might look like for an FDM printer: gcode Copy code ; G-Code for 3D printing ; Generated by SlicerXYZ v1.0 ; Initialize printer G28 ; Home all axes G1 Z15.0 F9000 ; Move the platform down 15 mm G92 E0 ; Zero the extruder G1 F140 E30 ; Extrude 30 mm of filament at a speed of 140 mm/min G92 E0 ; Zero the extruder again G1 F9000 ; Set speed ; Print the first layer G1 X0 Y0 Z0.3 F3000 ; Move to the start position G1 E10 F140 ; Start extrusion G1 X50 Y0 E30 F1500 ; Draw a 50 mm line G1 X50 Y50 E30 ; Draw a 50 mm line in the Y direction G1 X0 Y50 E30 ; Draw a 50 mm line back to the X origin G1 X0 Y0 E30 ; Draw a 50 mm line back to the Y origin ; Continue with subsequent layers G1 Z0.6 ; Move up 0.3 mm for the next layer ; ... additional G-code for subsequent layers ; Finish print G1 E-1 F140 ; Retract the filament G1 X0 Y0 F3000 ; Move the head to home position M104 S0 ; Turn off the extruder M140 S0 ; Turn off the bed M84 ; Disable motors To generate G-code for a specific model: Prepare Your 3D Model: Ensure your model is in STL format. Choose a Slicer Software: Download and install a slicer like Cura, PrusaSlicer, or Simplify3D. Import the Model: Load your STL file into the slicer. Configure Print Settings: Set up parameters like layer height, print speed, infill density, temperature, etc. Generate G-code: Slice the model to generate the G-code file. If you provide the STL file and print settings, I can help you further with the G-code generation process.

@@TheoreticallyMedia Interesting. Thanks.

Currently in the mobile app. We're all waiting for the big rollout. I just got 4.o in the web version, but still waiting on the desktop app. BTW: Mobile is getting slammed right now and I'm getting a lot of "We're under heavy load" messages.

First taste is free! Get everyone hooked on 4.o, and when they drop 5 as Paid only-- that's when you get a ton of conversions. I'm also convinced the free users will get stuck w/ 3.5 a LOT more often than we think.

Rolling out slowly now. I have the text version, but no voice yet

Agreed!

I mean, it’s insane!! And MS is building their own LLM?! Crazy pills, all of them! Well, except Sam. Apparently that dude is part Fox.

As if they are really competitors. They've carved up the market to themselves. Why would they fight each other when some upstart can swoop in from nowhere and steal their market share the way Google once did? Microsoft already owns a big chunk of OpenAI but it doesn't have the mobile marketshare of Apple. This also makes Apple more profitable as a AI customer than as a competitor. If Microsoft plays its cards right, it can be in the same position as it has with Windows, licensing AI as a service to other vendors.

its called collusion and all the big tech companies do it, MS, GOOGLE, APPLE are all friends behind closed doors lol

I mean, it's gotta be in part. I'd wager around 40 to 60%?

I think that moral upset you feel is very sound, and something I always stress - morality is about _intent_ , regardless of subject. Always be polite, thus. I don't think children will be more antisocial, though, unless they want to. Speaking to anyone - also a bot - IS to be social (see above listed intent!). And also - school buildings and -yards will be great places for playing with friends, but actual teaching will be much more efficient and individualistic.

No... because it's going to brainwash people.

​@@jvlbme😮 if I had to guess I think the classroom still exists in some form😮 either it's the government or like you know how we had local school back in the day like you know 15 kids from the block are going to go into a building and we're going to teach them no matter what because even in a post-scarcity Utopia we're going to find the time to teach our children😮 kids will be having friends around the local area😮 then the also have robot friends😮 and then on top of that they'll be like a virtual reality invented soon that's straight up matrix stuff like the AI figures out how to use sound waves or electromagnetic signals to create virtual hallucinations 😮 and this becomes like another form of the internet classrooms and everything else we do😮 like VR chat in The matrix on steroids😮 combined with AI assistant😮 and the holodeck😮 so kids are not just exploring😮 is it possibilities of AI they're using stuff like holodecks to time travel forward and backward in time😮 they do it multiplayer with each other with their friends locally and people all over the world😮 at some point these group of kids would be the most socialized humans in all of human history😮 and then we will also invent technology allowing these kids to be telepathic they'll be speaking to some technology in their mind and other humans that they're very close to in their minds😮 love in the future is much deeper than the flesh😮 it's fair to call them enlightened beans the more I think about it😮

Those Hiccups were the perfect flubs to prove it wasn't being faked. I actually think they were secretly happy about them!

Except that now there is zero point in learning a foreign language

@@michaelpowell775 bro thats like saing their is no need of speaking .translation is inefficient. direct speech is more efficient, it would be hard to interact freely with these. . i need to learn arabic and chinese. i want to read and consume books and media unrestrictedly. cant rely on an AI all the time.

If anything, I'm thinking this is going to HELP me learn a new language. I actually use the old voice model to help me pronounce names here. (That said, I still screw them up...hah.)

Without seeing scores or testing it myself I have the feeling that it is just slightly above ChatGPT4 or even the same with a different training dataset to include certain new features that seem to give it refined capabilities but nothing as impressive as a version 5 would have.

Chicken Dinner to YOU!!!!!

I think it has something to do with the upcoming (not announced) Apple deal, where it’ll basically replace Siri.

Haha. Real me! I’ll always let you know if I’m using an AI avatar! Might try it at some point as a way to get a “bonus” video out. But I promise to disclose!

The more I think about it, the more I'm impressed. Now, we just need it in our hands!

Open the pod bay doors, Hal.

Given Sam is trying to raise 7 Trillion, I feel it’ll be awhile before we hit that price.

But that’s ridiculously low considering how useful the tech is. Honestly they could charge like 50 a month people would pay it

Haha. Thank you! I’ve been holding on to this one all weekend!

Hook this new model up to HeyGen and the next step is Facetiming with your AI Assistant. The step after that...well, if I know people, that'll be NSFW!

Siri that works will be a godsend! In particular, if I can get "new Siri" to keep track of my Calendar/Tasks...I mean, that's ALL I've ever wanted in an AI Assistant.

@@TheoreticallyMedia My expectation of an AI assistent is way more complex than that. 🤣😂

Appreciate it!!

That was my first thought as well. I’ll say, I’ve talked about this with my kids (elementary level) and they get pretty mad about “cheating” with ChatGPT. It seems a lot of their friends feel the same. I don’t know if it’s a developmental thing or not, but I think as humans, we’re hardwired to want to learn to a point. Granted, like once you hit high school, you’ll get some kids skirting the system for sure!

@@TheoreticallyMedia some good points, but I was thinking more about motivation. It's already hard to get kids to work hard at school, what careers will be left in 5 to 6 years time?

Given my fairly broad experience with chatbots I've been pretty disappointed with the accuracy of the responses when it comes to hard data. There's a ton of hallucinations out there, but I don't think people are checking as much and just accepting the misinformation AI provides as factual. I check almost everything now and it's laughable (frightening?) how much it gets wrong in basic queries. Using AI for education will be at humanity's peril.

snowflake much