ChatGPT-4o Explained | Features, Benefits, and Use Cases

open AI strikes again and is anyone surprised they just launched a brand new model called Chat GPT 4 0 and that O stands for Omni and what this means is the model is much more capable of handling all three inputs of audio visual and text and outputting any of those three as well announcement is about me well color me intrigue are you about to reveal something about AI yes it looks like the king is in residence right now the Royal Standard flag is flying above Buckingham Palace which is the signal that the monarch is present for all you Marvel fans out there who wish you had access to Jarvis just like Tony Stark you might have that now we are up to 80 hours a day to counteract the symptoms this new model is fast uh it offers an average response time of just 320 milliseconds compare this to GPT4 which was 5.4 seconds and GPT 3.5 which was 2.8 seconds so the reason they wanted to refine this so much is to make it feel human like when you're talking or interacting with the model the real question is what can this thing do open AI gave us a ton of examples on their website and today we're gonna go over the ones that I found most interesting next to each of the examples I have additional use cases and in future videos I plan to address some of these used cases specifically so stay tuned for those but let's go ahead and jump right into the first one hey my friend and I are learning Spanish and we're wondering if you could tell us names of these objects in Spanish Hola where's your voice though cool what about what are these objects in Spanish the objects you're showing are Una mansana and Un platano in Spanish nice for a choice cool in this first example you can see just how powerful the audio in video combined with this new model is you can see that they're just looking around the room pointing at objects and it's giving the Spanish translation immediately hello there I'm here with my son and I love you to tutor him on this math problem but don't give him the answer you can ask questions and nudge him in the right direction but I really wanna make sure he understands it himself and he's here in the room so you can talk to him directly which side do you think is the hypotenuse um remember I mean hypotenuse I'm not totally sure I think I think it might be this one but I really am not sure this side AC you're close actually side AC is called the adjacent side to the angle alpha the hypotenuse is the longest side of a right triangle and is directly opposite the right angle can you find which one is the hypotenuse oh okay I see so um I think the hypotenuse is this really long side from a to B will that be correct exactly well done with the mobile app already out there and a desktop application coming soon you can now share your screen and let GPT talk to you to do endless possibilities for example you know in this situation Chat GPT is tutoring but you can imagine sharing your screen for any number of other things as well shipping it back would be the best option could you please provide the shipping information and process to get a replacement absolutely I'll email a shipping label and return instructions could you provide Joe's email address for me sure it's Joe at example dot com awesome alright I've just sent the email here's an example of allowing GPT 4 o to be your customer service agent so this unlocks a lot of possibility for automation in the workplace hey chat GPT hey there how's it going yeah it's going really well I was wondering if you could help me with a coding problem today of course I'd love to help you out what's the coding problem you're dealing with okay I'm gonna share with you some code one second sure thing take your time okay so I just shared some code with you could you give me a really brief one sentence description of what's going on in the code this code fetches daily weather data for a specific location and time period smooths the temperature data using a rolling average annotates a significant weather event on the resulting plot and then displays the plot with the average minimum and maximum temperatures over the year here's another example of using the desktop app to share code and what chat GPT see what you're seeing and get things done based on it in Open AI's presentation they really focused on the Omni capabilities that come with the new update but there's a lot of other added capabilities with the model that they didn't really touch on coming back to the website um if you go to this exploration of capabilities portion it shows you all the other things that are able to be done with the update and these are extremely exciting as well the first thing is it's able to take text as input and output that well which is very hard for most image models to do so you can see that they wanted a robot typewriting the following journal entries and then they put exactly what they wanted on the typewriter and you can see the output shows exactly what they wanted so this is exciting this is something we can really do before and another similar example is what they're calling variable binding so you're able to give the model different things to put together so variables you know putting variables together and you can see in this example they wanted the model to create an image with three different blocks with three different letters on them and it does it does an amazing job the next big thing is consistent characters so this is another thing that with generative AI we've kind of struggled with it's gotten a lot better with other applications like mid journey uh but it looks like with the new model JGBT model it's it's gonna be pretty easy to do this in this example they made a cartoon um so Sally the mailwoman and then they were able to replicate her doing different things um and you can see they're they're able to make basically a whole story book based on this character moving along with that same example they have another one here of Geary the robot so you can see they're having him do different things consistently it's the same character the same hat just performing different actions which is really exciting for things like comic books coloring books and you could it doesn't have to be you know cartoon character you can make it whatever you want another big thing is you're able to upload your own images or images of real people and combine them together create uh different variations uh use this image in different situations so this does get a little bit scary right because uh deep fakes and and other things out there but in this example they took two images and asked them to be combined into a movie poster this one didn't turn out so hot but they had the model redo it and you've got these two people in a movie poster with the words that they asked for and the facial expressions that they were looking for on the two people in a similar example you're able to upload a photo of someone so we have the same person in this one and their example and they turned it into a cartoon character so as you can imagine so many use cases for something like this and in the past there were entire applications that were built just to do this one thing so this next one blew my mind a little bit um you're able to make a 3D render based on images that you create so it does say there's four hidden steps here and basically what that means is they created each view of this 3D render and then they had the model put them all together and create a 3D render output so they have open AI spinning here um and they also made the seal spinning as well the next thing I wanna touch on is audio so this example they're showing how you can put a logo created by Chat GBT on something an object like a coin right which is awesome but the part that I really took away from this is at the end they said play the sound of the coin clanging on metal and then it outputs the sound effect so creating things like sound effects that you want for videos or anything else right you're able to just do this within the model and in addition to that you can upload any audio you want so think of a meeting um you can upload the entire thing into here and ask it questions have it summarize the meeting take notes um in this example yeah they uploaded an entire 45 minute lecture and had the model summarize the entire lecture and on this portion of their website they go through uh the comparisons to their own models and other models in text audio um SR audio translation um the M3 exam in vision understanding and it really does lead the way in all of these of course this is on Open AI's website so take it with a grain of salt but it'll be interesting to see how it does perform out there on all these tasks Sam Altman also posted a couple of comparisons on Twitter with um some different charts so the overall Elo score you can see that uh I'm also a good GBT2 chat bot leads the way by far this was just a nickname they were using for GBT 4 0 um and then he also mentioned that encoding uh it does far better than before which is interesting cause on the website the comparison says you know the real jumps are in audio and vision uh but it looks like based on these numbers coding is greatly improved as well so I'm excited to see what the outcomes of that are what is the cost so the amazing thing about GBT 4 o is that it's completely free um the plus users are gonna get 5 times the amount of message limits so if you're using it a lot you know it's recommended to get the plus version but anybody can go out and try GPT 4 o for free they're doing a slow roll out so GPT 4 o should be out there on the mobile app and web but it does not have the entire feature set yet the audio and video features are coming soon the last added benefit with the new model is that it's much cheaper hopefully we'll be able to get more out of it for less to wrap things up while we're not quite at Jarvis level capability we are getting really close and I'm so excited for the entire feature set to come out overall it just seems so intuitive with the audio and visual capabilities coming with the model and like I said before I can think of so many used cases for this and I'll have follow up videos going over those specifically I can't wait to see what y'all come up with as well thank you so much for watching and stay tuned for more

Channel: TrevorsTake

Views: 2,959

Keywords: trevorstake, chatgpt, chatgpt-4o, gpt-4o, gpt, vision, use-cases, use cases, applications, gpt use cases, 4o

Id: y-3vt7BApmE

Length: 13min 50sec (830 seconds)

Published: Fri May 17 2024