Googles GEMINI ULTRA Just SHOCKED The ENTIRE INDUSTRY! (GPT-4 Beaten) Finally RELEASED!
Google’s Gemini Ultra has just been released and it is causing shockwaves in the AI industry. The capabilities of this AI system are immense and the video takes a deep dive into its features and quirks. Gemini Ultra is the most capable AI model offered by Google and it is priced at around £18 to £19 in the UK and $20 in the USA. However, there is a free trial period for new users to try out the system before committing to a subscription.
One interesting aspect of Gemini Advanced is that it occasionally routes certain prompts to other, smaller models in order to save on computing power and deliver faster responses. Additionally, conversations with the system are processed by human reviewers to improve the technology, so users are cautioned not to input any sensitive or personal information.
The video also compares the performance of Google’s GPT-4 with Gemini Ultra, showing that they are very similar in terms of capabilities and responses, with the exception of formatting and sometimes length. Gemini Ultra is also noted for its speed and quality, with a demonstration showing its ability to deliver responses in just 21 seconds.
Overall, the release of Gemini Ultra is truly a game-changer in the AI industry, setting a new standard for capabilities and performance.
Watch the video by TheAIGRID
Video Transcript
Google’s Gemini Advanced Gemini Ultra is finally here and what an exciting day because there is so much to cover on how capable this AI system is I’m going to be showing you guys all the quirks and features so without further Ado let me show you guys the promo video and then
I’m going to get into every single feature you need to know About Go So that was the Magnificent trailer for Google’s Gemini and myomi I’m really excited because my preliminary findings on this AI system are really really good and this definitely does look in its current state to be better or at least on par with most of the capabilities
That exist in GPT 4 so you can see right here if you are trying to access Gemini Advanced the first thing you need to understand is that it is currently in the UK around £18 or £19 for Gemini advanc you can see that it says with
Ultra 1 .0 our most capable AI model this is of course Gemini Ultra so when you access Gemini Advanced you are going to have to pay £20 a month or £19 a month or in the USA $20 a month but of course they do actually give you two
Months free at no charge so essentially you do actually get to try it out for free because when I signed into Gemini I didn’t even pay anything and then of course I actually was able to access it and of course something that you do need to know is that this is really really
Good for what you are getting and of course price you know may vary based on C currency conversions around the world so I would just check when you log in to see how much it does cost and of course you know Bard is now gone so essentially
What that means is that there will be no more Bard it’s simply just Gemini and Gemini Advanced now of course essentially when you get Gemini Advanced you’re going to see here you know Gemini Advanced you’ve got our Gemini Advanced plan of course I I did sign up to the
Plan on another account and it says with our most capable AI model Ultra 1.0 and for any confusion for people that might be a little bit confused this is is the ultra model okay so as in when we got the previous demo where we saw Gemini Ultra and all of the crazy crazy
Benchmarks like for example when we saw these crazy benchmarks for Gemini Ultra this is the current system that we are going to be using so when you use Gemini Advance that’s what you can expect now there is one thing that I do want to talk about there are two in fact two
Things I do want to talk about before we get into the capabilities of this model because I you know people should be aware of these things okay number one is that of course there is occasional rooting so essentially here they state that Advanced it gives you access to
Google’s most capable AI model Ultra 1.0 and with ultra 1.0 it’s far more capable at highly complex tasks like coding logical reasoning following nuanced instructions and creative calibration and plus y y y now essentially they say Gemini Advance give you access to ultra 1.0 though we might occasionally root
Certain prompts to other models if you aren’t familiar with what they essentially mean by this is that for example if you ask Gemini Advance a basic question I’m guessing what they might do for people who haven’t paid and for those who are just on the free trial
Is they will root other simple questions to other models like Palm 2 and Gemini Pro that’s not Gemini Ultra so essentially in layman’s terms if someone asks a really basic question that doesn’t require extra compute and doesn’t require some Advanced you know reasoning engine they are going to root
It to a smaller model that can simply give you a very very simple answer and that’s just to save money on compute and of course to deliver faster responses so that is something you need to understand because sometimes if your response isn’t as Advanced as you might think it is
Sometimes the the responses have been rooted to a different model now before you start interacting with this model here’s one more disclaimer okay it will say um that your conversations are being processed by human reviewers to improve the Technology’s powering Bard so don’t enter anything you wouldn’t want
Reviewed or used so essentially every time you input a message into The Bard system human reviewers are going to be looking at that data and reviewing it to make sure the responses that you get are either good or either bad that way they can improve the system with the human
Feedback so don’t enter any you know passwords don’t enter any personal information because there are humans out there that are going to be looking at the conversations to improve the technology that is a disclaimer and now of course we can get into some of the actual features of this amazing
Technology so now let’s take a look at some of the comparisons that I did between gb4 and Gemini to show you just how good Gemini is so you can see right here I essentially just asked GPT 4 and Gemini some questions that I could ask
The llms SL AI models to see how good they were and they both gave me some very interesting responses chat GPT gave me a complete list you can see it says complex question and answering and of course I did actually you know prompt it again to ask it you know additional
Questions but with Gemini I did just have to prompt it once and you can see that says factual knowledge direct questions followup contextual queries I don’t have the full responses here but I will be showing you them um in the next bit in a moment but one thing I do want
To say is that if I’m being completely honest you know as I’ve used gb4 and and Gemini one thing that I’ve really noticed is that I can’t really tell the difference between the two um other than formatting and sometimes length because these models are really really similar
Okay like really similar and that’s a good thing now one thing that you do need to know about these models is that because Gemini’s reasoning is now so Advanced and definitely on par with gbt 4 one of the things that you do want to
Know is what sets it apart so one of the things I wanted to test out was of course the speed so in this video essentially what I did here was I wanted to test out how quick they could both answer the same question and one of the
Questions I wanted to ask them was what are the mental Frameworks for generating good questions because sometimes when you’re doing an interview or you wanted to talk to someone you want to have a mental framework for answering questions now at the bottom left you’re going to
See a timer and on the left hand side of the screen you can see Gemini Pro and on the right hand side of the screen not Gemini Pro Gemini Ultra 1.0 just to make that clear and of course on the right hand side you’re going to see gbt 4 so
When I set the timer I actually you know set the timer off and I recorded both signs of my screen at the same time and essentially we can see that that um you know one thing that Gemini Ultra 1.0 does have and Gemini has had for quite
Some time is the fact that it is blazingly fast like really really really really fast and um the speed doesn’t compromise on the quality because we can see here that this finishes at around 21 seconds and of course um GPT 4 I’m not sure what’s going on with the inference
Time or what’s going on with the you know compute on the back end but um I’m guessing that it’s a little bit slower because as I was continuing to time in this you can see that it’s on 308 seconds now uh gbt 4 is still going it’s
Still going it’s still going and then of course eventually we do get to 44 seconds around where that you know it it does give us a response so I think that was pretty fascinating because I think that whilst gp4 was always touted as you know the smartest one but a little bit
Slower um I think when we look at certain responses that aren’t too crazy and too different in nature if you can access something that’s nearly as twice as fast in many different use cases I would argue that um you know Gemini could be you know utilized more for
Other people because one thing people are trying to do inherently with these AI systems is they’re trying to save time so the fact that this is you know a lot quicker um in terms of you know one takes 21 seconds 44 seconds to answer the same question and you could you know
I guess you could say you could go to the responses and say that this one talked about uh you know first principles five y inversion critical thinking um then Socratic questioning blooms taxim money um then decision making y y y and then this one here um
Did uh you know eight points you could say that this one had more TCH so it had um you know more to do um I mean it’s really up to you okay but I think overall we do know that GT4 is a little bit slower and somehow somewhere I think
Google did actually prioritize speed and that’s why we do get these results because I’m pretty sure that Google does want this to be a very fast and Powerful system and it seems like they’ve managed to deliver on that so that was something that I wasn’t you know surprised by when
I did see this I was actually a little bit surprised that they was able to do this now of course this question isn’t something where I guess you could say you want to be you know asking and looking at into to much detail because of course it’s like it’s it’s I guess
You could say it’s a you know subjective answer because it’s not like you know a math question where the answer is either right or either wrong so I guess you could you know measure the compu in different ways on how accurate they are for you know answering different coding
Questions which I actually did and they were you know pretty similar but Gemini was still faster once again and it’s pretty crazy how faster it was now if you aren’t familiar with the benchmarks like I talked about before if you want to screenshot this or just share this
These are the benchmarks from Google official page on how the systems actually perform on the recent benchmarks and if you don’t know what the benchmarks are these are essentially you know tests like standardized tests that they have across all different you know categories so for example you know
Math reasoning you know in coding these are the different you know tests that they take and whatever percentages they get on these tests that you know all these llm systems or or new AI systems do um this is basically how we grade and see how powerful the system is and on
The left hand side um we can can see that Gemini Ultra surpasses GPT 4 in nearly everything other than the hell swag Common Sense reasoning for everyday tasks so um that’s why I said um you know even though it supposes it the thing is the percentage gains like 2%
Here then like 7% here then 2% here like 1% here you know it it’s not that noticeable when you’re using both these AI systems side by side and one thing I will want to see is how this does perform in of course the arena chatboard not actually chatboard but of course
Actually the leaderboard and essentially the leaderboard is an arena where you can see how llms are performing in the real world based on users because one thing that I didn’t know about tests like the you know MML U is that they actually do have some small um problems
With them so that will be interesting to see so another thing that we do need to check out and something that I know people will want to see is of course the you know responses okay and that is when we’re doing summarization now I do want
To do a quick caveat here because gbt 4 unfortunately recently has become very inconsistent with its summarization capabilities and I know it’s got some good ones because sometimes I’ll message gbt 4 and I’ll ask it a question and it will do pretty pretty decently and then
Other times I’ll ask it to you know summarize a question or paper and it just won’t do well and you can see right here that it summarizes this paper and this was a paper that I covered in yesterday’s video and it was about self-discover a framework used to
Enhance the you know abilities of llms through a different reasoning approach and um it’s a different way to get out improved capabilities from the model but unfortunately and I even tried chat gbd classic which is like the raw version of GPT 4 and it just didn’t perform as well
As Gemini did in terms of you know performing this kind of summarization with the key points like Gemini actually did a lot better and it actually did it a lot quicker now of course like I said this is just one this is just my experience in terms of what I would be
Using this for and I know that summarizing the text is something that other people are going to be using this for now I wasn’t able to upload it I was able to just you know paste the text into Gemini because it doesn’t actually suit that at the moment but I’m sure
That that’s going to be a feature that’s there or if it is maybe I I just don’t know how to use it but like I said I think this goes to show that Gemini is definitely on par with gbt 4 despite some rumors saying that it’s not that
Good because I think the problem with Gemini is that it has had a PR problem and gb4 does actually have the ability to do this as well but you know in some cases for some reason it just didn’t and of course there’s something to remember is that Gemini 1.0 is of course the
Model in which they have uniquely skilled at uncovering knowledge that can be difficult to discern amid vast amounts of data and it’s remarkable ability to extract insight from hundreds of thousands of documents through Reading filtering and understanding information will help through deliver new breakthroughs at digital speeds in
Many fields from science to finance now one of the things that I actually did like about Gemini was this feature right here um and I know most people won’t realize that this feature is there because it’s kind of hidden but um essentially if you click the button
Right here after getting a response from Gemini Advanced I’m not sure if this is in Gemini Pro but essentially if you click this button right here you can modify your response so you can have it simpler so if you don’t want it to be very long you can have it simpler you
Can have it longer maybe you want to you have it for a blog post if you want it shorter if you want it more casual if you want it more professional um and I think this is something that’s really cool and it’s just ease of use because a
Lot of times in chat gbt you have to click you know can I get this more casual can I get this simpler can I get this longer can I get shorter can I get this more professional so I think that uh this is something that’s really cool
And I’m going to show you guys how it actually works right now so here I’ve just asked it for some good mental models for improving my life and essentially if you know what a mental model is a mental model is just a way of thinking so instead of just thinking how
Usually think you apply that thinking to you know your your problem I guess you could say it’s kind of like prompting for an LM so uh right here we got inversion which is essentially instead of trying to you know solve a problem you flip the problem and ask what would
Be a bad outcome and then take the actions to avoid those pitfalls um and you can see right here that this is a really really good response and like I said you can click shorter longer simpler more casual more professional and I really do like this like I really
Really do like this as a feature and also as well you can click show drafts here um this is something as well that uh you know G GT4 doesn’t currently have and if you’re wondering why I keep comparing this to GPT 4 and I think the answer is obvious because um this is
What other people are going to do as well when they are looking at this because the large majority of people when talking about AI systems would regard tb4 as the very best system so you see right here um the other drafts I think this is really cool because it
Allows you to see different versions um and then you can think um maybe I like this one maybe I like that one um and one thing that I do want to say about this as well um one thing that I really do think and I don’t know what I think
This is going to change in the next model and I’m not sure if you guys have noticed this as well okay um compared to every other model gbt 4 for some reason always usually just generates list whereas I think other models generate for example they just generate X that
Seems a bit more human likee so for example um right here when I asked it a simple question it just generates 1 2 3 4 5 6 um of course you could always ask it to generate in a different format but it was something that I did know about
Gbt 4 that was always just 1 2 3 4 5 um and it was just something that was just constantly fr frustrating but you know I’m not hating rbt4 it is still you know a very very capable and Powerful system this was another example of an environmental question um and there
There’s not really much Nuance here in the details in you know in in these environmental questions this one says prevention at the source clean up and removal you know improving Waste Management clean up efforts it was just a simple question I asked about plastic pollution there was another one um about
Explaining Concepts now this was kind of fascinating because this is where we get into how explaining stuff is I guess you could say translated and I guess you could say that this this essentially means that you know we can see from this how well an AI system really understands
The problem because they always said that you know if you can’t explain a problem to someone who is you know less well versed in that area than you are then you don’t really understand it as well and essentially what I’ve asked them to do is explain the Monty Hall
Problem so if you don’t know what it is it’s basically a problem about probability and I’ve asked to explain them and I’ve read both of them because I understand the problem and why it is what it is and I think they’re both pretty pretty similar but I think chat
Gbt slightly slightly does it a little bit better in terms of the conversational because um you know he talks about a toy this one talks about a bike but like I said these answers are very subjective because they’re not based in you know a a definitive answer
So one of the things I did want to do was test the coding capabilities cuz on the benchmarks they actually said that you know on a human eval and the natural to code um it was really really better and I expect coding capabilities to take
A huge jump next year based on some of the things that we’ve seen with things like Alpha codium so I did my own test so I did my own test and I’m going to show you guys how that worked and after this I’m going to be talking to you guys
About the native image capabilities because when we do take a look at the image capabilities in image and 2 which is Gemini and darly 3 which is chat TBT there’s some very interesting things even though I did an entire video on it that you’re going to want to know so
Right now I’m going to be looking at the coding okay so let’s take a look at how good the coding is and I’m going to show you guys an example that shows you why I think most people are going to be switching to Gemini based on these
Thoughts but let me know your thoughts cuz I’m going to show you guys the coding now so what I’ve done here is I’ve asked Gemini Advanced Ultra 1.0 and gp4 the same exact question and I said can you create a trading bot that uses the moving average to have entry and
Exit points for maximum profit do it for trading view for pin script please and within a couple of seconds like literally 10 seconds it gave me this and I tried this and it didn’t work and before you like whoa it didn’t work I literally just put in the error code and
Then I got this back and then this worked and this entire process took around 20 seconds okay um I know you guys are going to see cuz um I show you guys how it worked and then of course let me show you guys the other version
So we have chat GPT right here it output this code it did take around you know 30 or 40 seconds and this one I’m going to show you guys what happened so this is the code I got right here so I come over to my trading board I put this into the
Pine editor then I click save and you can see it says compiling then it says saved and updated on chart and then of course we can see um on the strategy tester it shows us I think that’s actually different but um the point I wanted to prove is that it does code you
Know perfectly so gbt 4 you know then of course with Gemini advanced I’m going to take the first one cuz I’m going to show you guys the first one right here so we get the first one and of course I tested the first one I click save and of course
It didn’t work straight away but of course after updating it and then clicking save it did work and it did do it pretty pretty quickly now I think like I said this is just one example of stating that you know um sometimes you know different things work on different
Things and maybe that GPT 4 has experienced you know those kinds of problems more nuanced but um definitely before B’s coding was completely awful like when it was was released and the fact that it now is able to get the coding stuff really really well and of
Course based on what we saw before this is something that is very very surprising considering the fact that before um you know this just wasn’t possible and something that I do like about bod is that it is really really really really fast and maybe I should
Have just you know you know looked at the drafts or maybe I should have asked it in a different way because there are different ways that you can ask it but I think that b is and it’s actually not called B that was actually a mistake but
I think the Gemini here shows us that you know it’s still able to improve itself and that’s actually what they did say in the coding part of the section of Gemini so now what I want to take a look at as well is of course the native image
Capabilities because like based on you know most of the you know individual tests that I ran it looks like Gemini Ultra is really really good and I really can’t see a huge difference between these two models I mean one of them I would say if I’m just being honest is a
Little bit faster and although there are going to be some you know use cases where you know Gemini might fail because it hasn’t been tested as much um and I would say that gbt 4S had over a year to be red teamed to be tested to figure out
All the bunks and Kinks just because of the fact that it’s had like over 100 million users every single month it’s got a crazy user base so people are reporting issues I think that you know that that is probably going to be the only area where it falls I think right
Now both systems based on what I’ve seen are really really good now something different is happening in the image capabilities cuz it was something that I did another video on um but I’m going to show you guys a section from that now because if you don’t know okay
Essentially what im and 2 is IM and 2 is the image based system that is inherently in this Gemini Ultra system so essentially this AI model has a very very Advanced image generation system that I’m going to show you guys about from now so essentially I’m going to
Just have a part of that video in here and then I’m going to come back to this because there’s no point repeating everything so you can see right here that this is a really really nice photorealistic image um quite surprising actually just how the lighting is like I
Said I wouldn’t know that this is a generat it doesn’t look AI generated to me um but in fact this next image does look a bit more AI generated because it’s got that I’m not sure if this is like a bouquet effect or what kind of
Effect it is but um it definitely is like I guess you could say have that AI generated thing but Google’s photo realism it is definitely most 100% there now we do need to take a look at something else because this was one of the features that you may have not seen
So here we have out painting from Google Now in other Google products that they did talk about they have actually mentioned um out painting and in painting before essentially if you don’t know what that is that’s essentially where you zoom out and essentially where that image you can just essentially just
Increase the size of that now that is something that we did see in mid Journey but it is good that we are now getting this from Google as well and I think this will be better implemented in Google just because of how um they’ve made the software and of course like I
Said they also did include in painting which is of course if you want to add something like right here in this example they wanted to add a shelf and it looks pretty pretty cool um and you can see they just set a Shelf with a couple of books on and vases hanging on
A green wall and then of course they’ve just managed to add that right into the image so I feel like that is basically you know it’s essentially just photoshop’s generative fill if you do remember photoshop’s generative fill it was essentially just this which you can access natively in Photoshop and of
Course um that was something that definitely went down well with a lot of people and I’m sure that now that Google has it especially with its photo realism it could definitely compete because we know that generative fi while it’s okay it sometimes doesn’t always give you the
Best results so if Google manages to get this area right um and currently um you’re going to see in vertex that I don’t think they gave us this feature but you’re going to see that it actually is really cool now what we also did get was we also did get text rendering
Support So essentially you can see right here that there is text being able to be put into these images with a remarkable degree of accuracy so the prompt STS a tube of toothpaste with the word symbol written on it on a bathroom counter advertisement and you can see that it’s
Got the words all the way down to the blur really really perfectly then of course a cup of strawberry yogurt with the words delicious written on the side sitting on a wooden tabletop next to a cup of yogurt is a plate with a toast and a glass of orange juice and you can
See that the text here looks very very accurate and it even manages to do it with like the you know the blur going on there and of course with different fonts that look really really creative and stylistic so I think that text renderings support is definitely one of
The things that was really hard it was only recently achieved by I think it was Dar 3 and mid Journey so now that Google has added this as well it goes to show us that they are not messing around when it comes to this and I’m sure that this
Image release is going to be there because they want Google’s Gemini to be very very good when it does release in its complete model now Gemini Pro is actually pretty good but of course Gemini Ultra they want that to be absolutely astounding now here is where we talk about the intuitive editing so
This ladies and gentlemen is image effects so essentially if you don’t know what this is this is from Google’s Test Kitchen and this is why I said this is probably even better than mid Journey if you are using this so essentially what you can do you can see that when you
Generate a photorealistic image you can essentially break up the words into different sections and these different sections can be changed to your preference if you want to adjust how that image is so you can see that he changed it from you know adjust jungle to a city um and you can just simply
Break that down and it allows for greater creative freedom so I really like this so much because being able to switch from City to Jungle and just instantly get um a new generation I think that that is a feature that we have been asking for for quite some time
In many of these AI generators for example um and this is no hate to Mid Journey or whatever but it’s just a comparison like mid journey is absolutely incredible if you don’t know what it is it’s arguably the state-of-the-art model in terms of image Generation Um but you know just how it
Is on Discord it is a little bit harder to use than just a simple like drag and drop you know a simple you know uh drag and drop and just you know change things with you know just click of a button and I know that they’ve made it easier and I
Know that they are working on a website but if Google manages to roll this out you know globally I think this will become more adopted because of how easier it is to use so I think something like this goes to show us as well um and
I’ll show you guys the demo of this later because like I said you can actually use this with a link in the description so you can see right here there’s other um things at the bottom you know HDR 1990s all of these things you have to remember that the average
Person who’s using an image generation software they don’t know about every single you know photo style they don’t know you know what warm coloring is you know uh you know certain types of film style you know they know HDR 199s Renaissance style they don’t you know immediately have all of these in their
Head so it is very very useful that we get something that can actually you know give us that ability to you know brainstorm for us and then we can see okay we can change this part we can change that part we can change from l
Ape to this um and I think it’s just going to give us uh a greater degree of creativeness to you know completing whatever project or whatever you know outcome that we do desire so I think that this is really cool um and this is definitely something that is an
Improvement over other models that we haven’t seen before so Google actually have done a great job on that bit now in addition you can see that right here instead of photo real portrait you can see handdrawn abstract impressionist so if you want a handdrawn one you can
Simply change it and then of course you can see that is how it looks you can then of course as well you can share this and you can get the seed as well so it’s pretty much like mid journey in the sense that you can um get the same seed
Which allows you to generate really similar images you know which are borderline consistent so I think this is you know one of the things that like I said from Google coming from Google and their image effects so if you want to actually use this this is going to be
Different just to make the distinction clear what we’re seeing here is actually Google’s image effects and essentially what that means is that this is in Google’s Test Kitchen so if you don’t know what Google’s test kitchen is the test kitchen is the area in which you can test new Google releases before they
Are essentially put out to the mainstream so I guess you could say it’s like an alpha version of um you know the AI stuff before it is wildly rolled out and there’s a few different things there like a music generator and an image generator um and they’re actually still
Pretty good but I’m guessing that Google just you know wants to test them get some feedback um and then of course um roll them out um completely so this is image effects and this is not Google B but you can use this now um and I’m guessing that this is just a more
Advanced area so like I said I am a real real fan of this because like I said I think this helps out people more than you do realize so there was also this as well so logo generation and you could see that they actually talked about how Google’s you know new image generator
Image and 2 which is of course now powering pod essentially prompt a clean minimal emblem style logo for an ice cream shop cream background you can see that that looks really cool then an abstract logo representing intelligence for an Enterprise platform vertex AI written under the logo so I think that
This is also once again you know of course you can do this with chat gbt chat gbt’s D 3 is actually pretty good but like I said Google’s coming over to here and as far as I know when I was testing this out in Google’s bod there
Wasn’t really a limit on the amount of image Generations I could have so I think that this is really effective for those of you who don’t have a gbt 4 subscription and you don’t want to subscribe to anything you just want to have your Google account you want to go
Ahead you want to create an image you can go ahead and Bar um and you can just do this straight away so I think this is something that’s pretty effective and um yeah another thing um is that you know these are pretty clean as well because
The text looks pretty decent um and I would say that this is a win this is definitely a very good win now there’s also something that I wanted to include as well and that’s of course the seeds like mid Journey so if you don’t know uh mid what mid Journey has mid Journey
Essentially has seeds a mid Journey seed number is basically just the number that allows you to direct the consistent in the image generation process so essentially it’s just the starting point for the AI to generate the field of visual noise and basically it’s just a reference point that allows you to
Create more realistic SLC consistent results across your work so that’s why I said that this is really cool because you know right now from Dar 3 we don’t actually get any seed numbers um and even through you know Discord and mid Journey the way how you get the seed is
You have to put slash type in you know a lot of people just don’t want to do that like I know it’s fairly simple to someone who’s using it all the time but someone who’s just getting started you’re really just going to want to be able to click a button just copy and
Paste seed um because that’s how majority of the things and apps that we use are now there was something uh that I think is good because image and 2 includes built-in safety precautions to help ensure that generated images align with Google’s responsible AI principles and essentially it’s you know
Watermarked with Google synth ID which allows you to generate invisible watermarks and verify images generated by this software so if you don’t know what synth ID is essentially it just generates a synthetic ID so it’s BAS basically a digital Watermark directly into the pixels of these air generated
Images that are imperceptible to the human eye so if you ever want to check the validity of an image you can do so with this software now essentially what’s good about this as well is that they’ve managed to make it so that even if you edit the image so even if you you
Know add some filters add some exposure The Watermark ID is still completely there so um you know there’s no way and you can see that we designed synth ID so it doesn’t compromise image quality and allows the waterm mark to remain detectable even after modifications like adding filters changing colors saving
With the various lossy compression schemes most commonly used for jpeg so I think this is really good because right now we are entering the era of you know images which we don’t know are real or not and if Google does manage to become the number one in terms of image
Generation it’s going to be good to be able to have something that we can actually back check and then go back and say okay that is AI generated and that isn’t AI generated um and I think that other companies should take a note from Google because this is something that we
Do need in the future and right now whilst it’s not that big of a problem I think in the future it will become a larger problem as more people start to realize so let’s go ahead now and take a look at some of these images as well
Because I wanted to look at some of the individual images so this is a really cool realistic one like I said the photo realism with Google is really really good so the prompt is generate a realistic photo of a person looking off camera during sunset portrait mode so
The background is faded then of course we have this one right here I really do like this one I don’t know why it just you know um ushers a sense of peace I guess you could say and it says generate a collage art with photorealistic images of oceans and plants with muted colors
And 3D shading that’s mixed media then we have this one I I find that this one is really cool as well um write a social media post and generate a mouthwatering image that I can use for a buffalo wing Festival that actually does look remarkably realistic um and I wouldn’t
Be able to tell that that is AI generated maybe my eyes are awful but um just the lighting the shading um definitely uh pretty pretty incredible stuff there and then of course image in 2 which is generating an image of a fashion show and a steampunk style
Digital art zoom in on their face looks very very effective I mean some of these uh right here don’t look that you know accurate but in the digital art style that is actually really cool so if you seen the digital art style you know why that actually looks really really cool
So um essentially what they’re trying to do with these these images is they’re trying to show how good the range is is on these images because essentially what they’re showing us is that you know there’s a diverse um uh you know style set that you can apply to this um and it
Just goes to show us how effective this is and generate an image of a futuristic car driving through the mountain and then of course it generate a vibrant and Lively image depicting an elephant parting in the heart of a lar vibrant jungle the elant should be in various
Colors and be adorned with fun now um I’m going to show you one more image here and then what I want to show you guys is the comparisons to darly 3 um and like I said before even though I’m doing these compar don’t forget that this is like I said
Google’s second iteration whereas uh you know this is I guess you could say Dolly 3’s third iteration because before they had doly 2 they had darly then they had darly 2 then they got doly 3 so I would say you know just just REM just remember
That because um I think that is you know the only fair way to compare it so um this is just Genera an image of a cluttered Alchemist Workshop filled with bubbling flasks glowing crystals and the tiny luminous swirling within the bottle so then of course we have imagin 2 and
Dar 3 on the right hand side um and I think this one here if a being honest this one D 3 does look pretty cool but I would say a tiny luminous World swirling within the bottle um this one here kind of does I mean the nuances here actually
Do look better from image and 2 I just think the understanding is better but um we never know when they’re going to drop darly 4 so it will be interesting to see what is done there we also do have this comparison so imag two um on the left
And then darly 3 on the right so I guess they have different interpretations of you know what collage is um I guess that one’s open up for interpretation and of course we got darly 3 here as well um in terms of generate photo realistic person looking off camera during sunset um and
I think it’s it’s clear that you know um image in 2 in terms of photo realism just really really wins in that sense because I think you know the way that they train the system in order to generate um how that looks is particularly stunning so we’re going to
Take a look at some of these more and then once we take a look at these then I’m going to show you guys how it actually works and how you can actually use this so you can see um this one again futuristic car driving through an old mountain surrounded by Nature um I
Think definitely Dary 3 does look a little bit better here but that’s completely fine um and then of course we have here I think it’s just different interpretations of what we’re looking at because like I said here this one seems to be a little bit more playful like a
Like a kids thing and then this one here seems to be a bit more artistic in terms of what we’re seeing and of course this one right here as well so you’ve got this um this is you know like an image of uh some buffalo wings um and of
Course this one on the right here doesn’t really look as realistic and then this one on the left you know Google’s one does look really realistic so um right now what I’m going to show you guys is a quick de demo of The Bard which you can actually use right now
Then I’m going to show you the image effects which is essentially this one which is the intuitive one which is uh you know like I guess you could say a soup top version of what mid journey is kind of supposed to be so um I’ll take a
Look and I’ll show you guys how those are so here you can see we are now in Google’s Bard and I just simply asked it to make me an image of a car and these are both powered by the same system so when I go and show you guys the images
From image effects it’s all going to be powered by the same system so I said make me an image of a car um and it’s chosen to essentially make me an image of some BMWs now I didn’t really do any kind of you know prompting really good
Prompting there I just simply said make me an image of a car and what I find cool is that you can generate one image and gen generate three pretty quickly um and I think it does generate a lot of images pretty quickly like for example um this image of an egg that doesn’t
Look as realistic um but this definitely does like I would say that yeah that’s a real Burger the sauce and everything right there and for this second iteration I would argue that it’s pretty pretty realis so we click generate and then you guys are going to see how long
It does take I think it is pretty pretty quick I’m I’m not sure how how they managed to get it to be this fast but um the first time I generated these images it was really really quick like surprisingly quick and of course you can see here that um I said generate an
Image of photorealistic food and then of course we do have those images here so I think that um you know it it does really really really well um and I’m going to say um and so now that you’ve seen you know Google Gemini image and 2 you’ve seen everything the only thing that’s
Coming now is essentially an app which is going to be pretty cool and one thing that I am waiting for for Google Gemini like right now I’m probably going to use this every day because you know it’s pretty another point I did want to make before the video ends and I really want
You guys to you know hone this point in because this is going to be the future of AI systems and with Gemini being where it is now I think they are in probably one of the best positions ever to take over because of the fact that
It’s with Google Now one thing that we did see in the trailer right here you could say that this is the next chapter okay and of course it is because what Gemini has that most people are going to really really need and use is the personalization so everyone that signs
Into Gemini Advance what you can see is that it says your name it says hello Lisa chat doesn’t have that at the moment we know that Sam Alman is working on the personalization the AI agents all of that good stuff but one thing that it doesn’t have as well is that it can
Literally go into your emails and summarize your Gmail and all that stuff with your workspace and we know that Google has so many different users from all these kind of apps and stuff so it’s going to be really cool to see how Google you know implements the personalization and stuff with your
Account so that I think is going to be the next stage because you know just looking at the benchmarks okay like even if they manage to squeeze out another 2% on all these benchmarks it’s not going to improve that much in terms of your daily usage like yeah and the email
Might be 2% better but are you going to realize it you know what I mean like gbt 4 to gbt 3.5 there was a huge jump but if we go from 90% to 90.2% I mean you’re not going to really realize it between these systems so the usability the speed
The efficiency are going to be key determining factors when people do choose their AI system so with that being said if you’ve enjoyed this if you enjoyed you know looking at the AI you know image capabilities of this new model you know this image in two based
On exactly what it is if you’ve you know enjoyed looking at you know of course another thing like how fast the model is compared to of course chat GPT let me know and if you enjoyed the video of course leave a comment down below and
Video “Googles GEMINI ULTRA Just SHOCKED The ENTIRE INDUSTRY! (GPT-4 Beaten) Finally RELEASED!” was uploaded on 02/08/2024 to Youtube Channel TheAIGRID