To try everything Brilliant has to offer-free-for a full 30 days, visit brilliant.org/TheAIAdvantage/ . You’ll also get 20% off an annual premium subscription.
Unfortunately this might mean payment processors are gonna become even more strict, which sucks for businesses that are legitimate. And PayPal is already a shit show with banning legitimate businesses and keeping their money for ridiculous periods of time, often indefinitely.
They are easy to spot. There are too many stitches in the composite images and the edges and pieces don't align or agree. All of it creates a perceptual dissonance. The videos and music are even worse.
I would say that the definition of a photo stands. What we're generating are "images". a digital photo is an image, but any given image is not necessarily a photo. kind of like a square is a rectangle but a rectangle is not necessarily a square.
I agree with your summary of uses for ChatGPT, Perplexity and Claude. I also find there are use cases where Gemini produces the best results. The tone is similar to ChatGPT, but the image recognition from an uploaded images is very good and consistently better than ChatGPT or Claude.
It's incredible to see how far technology has come and how AI is revolutionizing so many aspects of our lives. The future looks bright with AI leading the way!
@@davidniquot6423 nope. It’s irrelevant what people think they’re looking at. This has no bearing on what a photo is or isn’t. As explained already one is generated data based on a bit of code the other is a moment in time captured for posterity. One is, one isn’t.
Hi, thank you for providing a lot of great information over a long period, including through this TH-cam channel. I enjoy informing my colleagues weekly about developments in AI, and for this purpose, I create my own summaries. It has been useful to use GPTs for this. When I tried to run a summary on this video, I received a message that transcription is disabled. Why do you do this?
For your update on the prompt caching, can you compare it to googles prompt caching that has been out for a while? Also from what I have gathered it doesn't really replace rag as the context window is way too small, practical uses of rag have thousands of documents etc but it can potentially be used in conjunction with rag? Maybe it could be useful as a way to store the system prompt or a common prompt that will held constant across all of your prompts.
Oh nice I wasn’t aware of googles version. Will have a look. And yeah I see your point but often Rag is usef to just store a few docs and if you could cache something like 150k tokens and maintain speed it would be big already. I also thought of storing multiple pieces of context in a list and then changing out the cache dependent on the context of the input prompt? Dunno have to play a bit.
Medical Research: The OS’s ability to handle large datasets efficiently supports medical research, enabling the analysis of complex medical data and the development of advanced AI-driven diagnostics.
Indistinguishable until you look at the chain he is holding next to his perfect face. Get your point though, it was just funny how bad the chain looked
Someone described this like when the internet was first created and people were like wow, I can send someone an email. Nobody realized how intertwined it would become with our lives.
AI is becoming frighteningly real. Whoa! Though it is incredible how far we've gone, you must wonder what lies next. Wonderful moments ahead! AI is currently used in everything, from everyday chores to industry transformation. How it is transforming our lives is astounding.
it is not a photo, it is a generated image. Big difference. Photo comes from photon. Light is captured and the wavelength and intensity is either stored as an analog or digital copy of that moment. Just because it looks like a real life photo doesn't make it one.
@@aiadvantage No, but then the question is an epistemological one: How do I KNOW whether something is a photo or not. It's still not a photo. Nothing about the concept of a photo has chnged.
@@jonah-b7k Many games allow you to take a "photo" but clearly they're not photos in the traditional sense either, they're more a snapshot of what the character can see in game. You can suggest "photo-realism" to AI image generation models and it means something in terms of the expectation of the output without the expectation its an actual photo. Perhaps the usage of photo is morphing in the same way the phrase "put it on tape" no longer means anything to do with actual tape.
What about A.I. detecting A.I. ? Will we all have an icon pop up telling us that the content we are watching or listening to is A.I. generated or manipulated?
Step by step, you cheapen the value of any digital file. Future coin of the realm will be a photograph negative and a “glass” magnifying glass. Every digital image magnifies out to rectangles; but every real image stays true all the way down to the diameter of a silver nitrate molecule. Nothing like file size to be the giveaway.
Even that will eventually be something you can fake; the ultimate conclusion is what it has always been, without people realizing it: the value of actual "experience" - and even that is questionable, because you can have hallucinations caused by chemical imbalances, etc. Ultimately, it's the experience of the moment that has value, and even that is only for the moment that it's happening in, rather than value for longer than that.
I mean you could just create your own gpt model with persistence although your going to run out of memory really quickly even if you use tokenization and compression or you can fight with managing recursive efficiency reprogramming that doesn't crash or spit gibberish.
I have to wonder, if our AD projections and other self incorporations, will cause people to have "phantom self" dissociation. Like looking into the mirror and being like, wtf?
it seems like the new A.I. avatars do not exist - but they are nascent life coming into existence on our plane of being- that's how they are beginning to seem 'real'
Unhinged is the best way to go. ChatGPT literally flagged my content for information on multiple historical figures because it said the content was disturbing. Like, yeah, news flash: people like Elizabeth Báthory did evil s**t, but that's history and should not be censored. I'm going to try out Grok and see what it does.
You would think TH-cam communities would solve that but…we’re here with similar interests, a shared expirence of watching the same video and yet still strangers. Humans need to go back to in person community building.
@@Mustachioed_Mollusk Not gonna happen. People are self interested and intimate relationships particularly between the sexes are contentious to say the least.
@@djpuplex is that true of you as well? 2 people can easily find a third and so on. There is a demand for connection after all. It is possible but would require a fair bit of pessimistic, well informed critical thought.
Grok 2 has taken the job of yours! Seriously It can give me a list of all updates that happened in a week just by asking, man what a time to be alive. First job has finally taken away ( Just kidding I am going to watch you )
From some quick testing there is some difference between what LLMsys has as Grok2 and sus-column-r. Sus-column-r is distinctly different (better reasoning) to GPT4o & Claude. Grok2 on LLMsys is meh. If its actually Grok2-mini that would explain the difference. Sus-column-r is quite woke so I had a good laugh on learning its actually Grok.
I am being overwhelmed every day by TH-camrs expressing their "opinions" about AI, pretending to be delivering some important insight, to make more views and earn easy money, without actually adding any value to anything. This is getting so ridiculous with every single "smart boy" or "smart girl" that will enlighten you repeating over and over their particular opinions, and earning their life with "haaaard wooooork", mostly made of repeated information, facts, and by most probably an AI . Very smart! Why would I learn to build anything if there is already the AI to do it? Why? This is so contradictory....
I published a short book with step by step to run local AI on Ollama and LM studio. Sales zero. Hype AI books full of fluff have hundreds of sales last week alone
It's because all these contents are made by AI itself, so because "AI" is a reflective tool rather than actually intelligent, the human part to reflect back/upon it being missing, there's lack of depth at transferring value, it's like empty data, "soulless" arguments into a topic without the correct amount of reflection complexities.
Very insightful. I will write it down. None of this is real. It is just AI repeating everything. Why would I learn anything. AI already do everything. I am smart boy. I must go study wise words now.
i have exam coming in few weeks and claud ai works best for it and me can you tell any website or somthing that would let me use claud for free for more generations im broke and from pakistan so i cant afford
Yeah, no. No matter how realistic AI-generated images get, they will never be, by definition, photographs. Photographs are a category of images, and AI-generated images are also a (separate) category of images. Photographs are images created photographically, while AI images are images created with AI image creation tools. We don’t call hyper-realistic paintings photographs just because they may look that good. Photo-realistic AI-created images need their own -new- term. Other than that, though… great video.
This entire channel is this a sales funnel to your community where you show people how to actually do things. Apparently. I watch these videos all the time and I'm like okay that's cool but how do I do it oh wait I have to pay for that thanks
You know what I want hyper realistic image generation of images of myself. You could say you show us how to do things on this channel but you know damn well it's not the things you're actually doing not the top tier stuff if you know you have that stuff but it's behind a paywall then you cannot disagree with what I say anything else is just cope
That may be correct, that it is a sales channel, but that doesnt detract from the quality of the information. In the olden days, Advertisements were actually useful, they made you aware of new things that you would find useful. Now they are just trying to ram c r a p down your throat. So i see this more like the olden days Ads. If you get anything for free, then you need to realise you are the product. No problem with that in my book if i can still chose to be part of it or not. The choice part is the important part, to me.
Democratize? Through theft of artists work? If I steal your creations that’s democracy? Also, what about the massive amounts of energy that these things use? The mining of rare earths and destruction of habitat? This provides a way to. Completely misrepresent and misinterpret reality and accelerate that process at a time when we are in multiple crisis’s. Good to know.
I don’t know why I have to go to the dark place on this, but I find myself wondering if AI will help to mitigate human trafficking and child porn because real people are no longer necessary?
Is this a real guy or just A1? How would you know these days? This whole AI thing is fascinating (and a lot of fun) and certainly has loads of uses and applications but if it gets out of control we’ll all be in the sh_t if we can’t tell what’s real and what isn’t.
I just looked up what the definition of Fascism is: a political philosophy, movement, or regime (such as that of the Fascisti) that exalts nation and often race above the individual and that stands for a centralized autocratic government headed by a dictatorial leader, severe economic and social regimentation, and forcible suppression of opposition. Now do you really think that is the correct description of Elon Musk?
@@malcolmvanhilten125 i wish i could have the source where you picked the definition. I can give you mine which are different historians, philosophers such as : Emilio Gentile, Angelo del Boca, Jason Stanley, Johann Chapoutot or Victoria Di Grazia. Definition of fascism needs a lot of papers more than a comments on TH-cam. And by the way there are some points in your definition that are somehow linked to this guy actually in regard to his political position and biography of this man.
As much as you try to convince us this is realistic, ALL the images you showed (to me) were obviously clearly AI generated garbage, look at the skin tones, look at the HQ cartoon likeness of the images. Look at the backgrounds, the lack of uniformity of skin color. The lack of error in focus, too perfect here, not perfect enough there. COME ON be honest its nice but to compare with reality is GARBAGE, sorry your so hyped up you dont see this. GOOD GRIEF!
To try everything Brilliant has to offer-free-for a full 30 days, visit brilliant.org/TheAIAdvantage/ . You’ll also get 20% off an annual premium subscription.
'But that was last week'
Kind of the tone of the times we live in.
Yeah. I don't watch anything more than a few days old anymore because of this 😂
As a retired newspaper cartoonist in Utah, USA… good for Grok. 🤖🖖🤖
man you retired at the perfect time
This is gonna be a heaven for scammers
Unfortunately this might mean payment processors are gonna become even more strict, which sucks for businesses that are legitimate. And PayPal is already a shit show with banning legitimate businesses and keeping their money for ridiculous periods of time, often indefinitely.
Dont givem the idea 😢
Already is
They are easy to spot. There are too many stitches in the composite images and the edges and pieces don't align or agree. All of it creates a perceptual dissonance. The videos and music are even worse.
@@bluenetmarketing the average person, pal. The average person.
I would say that the definition of a photo stands. What we're generating are "images". a digital photo is an image, but any given image is not necessarily a photo. kind of like a square is a rectangle but a rectangle is not necessarily a square.
Or a painting is also a picture, but not every picture is a painting...
Unremovable watermark or metadata to make fakes be able to be determined as fakes a must!
Thanks for another one. Anything on lip synching is helpful for me. Like you said it's not there yet, but it will be. Please keep your eyes peeled.
I agree with your summary of uses for ChatGPT, Perplexity and Claude. I also find there are use cases where Gemini produces the best results. The tone is similar to ChatGPT, but the image recognition from an uploaded images is very good and consistently better than ChatGPT or Claude.
What's weird for me is when I see an AI generated human I get this feeling of being sick and nauseous for some reason.
It's called the uncanny valley. en.m.wikipedia.org/wiki/Uncanny_valley
It's incredible to see how far technology has come and how AI is revolutionizing so many aspects of our lives. The future looks bright with AI leading the way!
aI generated images aren’t photos. They are composites of data. A photo catches a moment in time. AI composites don’t. There is no confusion.
There is confusion cause only the person that "took" a photo would know that it is a photo.... all the other people would just guess on the apparence.
@@davidniquot6423 nope. It’s irrelevant what people think they’re looking at. This has no bearing on what a photo is or isn’t. As explained already one is generated data based on a bit of code the other is a moment in time captured for posterity. One is, one isn’t.
Spot on. Wish some people know would be smart enough to understand the argument 😮
Hi, thank you for providing a lot of great information over a long period, including through this TH-cam channel. I enjoy informing my colleagues weekly about developments in AI, and for this purpose, I create my own summaries. It has been useful to use GPTs for this. When I tried to run a summary on this video, I received a message that transcription is disabled. Why do you do this?
Transcription works for me... maybe a country issue?
sounds like the anthropic thingis like having it spend idle time collapsing concepts in your cache
Still the most useful AI news video each week. Thank you!
If I want to take a photo of my kitchen then use some photos of different appliances to see how they look in my kitchen which model should i use
We love and appreciate thew quality of you content here, keep it coming!
Well put!! “Let’s call a spade a spade.” 😂
It seems like the only one not promising and later not delivering is Anthropic.
too many people confuse a shovel with a spade. it doesnt work
For your update on the prompt caching, can you compare it to googles prompt caching that has been out for a while?
Also from what I have gathered it doesn't really replace rag as the context window is way too small, practical uses of rag have thousands of documents etc but it can potentially be used in conjunction with rag? Maybe it could be useful as a way to store the system prompt or a common prompt that will held constant across all of your prompts.
Oh nice I wasn’t aware of googles version. Will have a look.
And yeah I see your point but often Rag is usef to just store a few docs and if you could cache something like 150k tokens and maintain speed it would be big already. I also thought of storing multiple pieces of context in a list and then changing out the cache dependent on the context of the input prompt? Dunno have to play a bit.
Medical Research: The OS’s ability to handle large datasets efficiently supports medical research, enabling the analysis of complex medical data and the development of advanced AI-driven diagnostics.
Indistinguishable until you look at the chain he is holding next to his perfect face. Get your point though, it was just funny how bad the chain looked
Someone described this like when the internet was first created and people were like wow, I can send someone an email. Nobody realized how intertwined it would become with our lives.
4:35 as long as you don't look at the image it looks real. The chain, some anatomy on the hand and skin of the face, the hat... the Eiffel tower.
AI is becoming frighteningly real. Whoa! Though it is incredible how far we've gone, you must wonder what lies next. Wonderful moments ahead! AI is currently used in everything, from everyday chores to industry transformation. How it is transforming our lives is astounding.
Okay, so how the hell do I give you a triple thumbs up???? This deserves much more than just a like button!
it is not a photo, it is a generated image. Big difference. Photo comes from photon. Light is captured and the wavelength and intensity is either stored as an analog or digital copy of that moment. Just because it looks like a real life photo doesn't make it one.
No, it's not a photo. It's an AI generated image
But how would you be able to tell if you lacked the context of the video and just saw the image?
@@aiadvantage No, but then the question is an epistemological one: How do I KNOW whether something is a photo or not. It's still not a photo. Nothing about the concept of a photo has chnged.
@@aiadvantageIf I sold you a pleather jacket and told you it was leather and you believed me, that doesn't make it leather.
But that only works if you know the origin of the image. Otherwise it looks just like a photo. That's the problem.
@@jonah-b7k Many games allow you to take a "photo" but clearly they're not photos in the traditional sense either, they're more a snapshot of what the character can see in game.
You can suggest "photo-realism" to AI image generation models and it means something in terms of the expectation of the output without the expectation its an actual photo.
Perhaps the usage of photo is morphing in the same way the phrase "put it on tape" no longer means anything to do with actual tape.
omg the spiders with hats are so cute
I believe there is only a 5 minute cache with Anthropic. That's the catch, at least right now...
Smythos AI agents interact smoothly with your existing systems, improving your productivity without affecting it.
Which Ai service do you use for your thumbnails? I love your thumbnails
Simone is coming to life in front of our eyes. This will hurt the Hollywood elites right in the pocket book.
What about A.I. detecting A.I. ? Will we all have an icon pop up telling us that the content we are watching or listening to is A.I. generated or manipulated?
The guy in this video looks AI generated.
Step by step, you cheapen the value of any digital file. Future coin of the realm will be a photograph negative and a “glass” magnifying glass. Every digital image magnifies out to rectangles; but every real image stays true all the way down to the diameter of a silver nitrate molecule. Nothing like file size to be the giveaway.
Even that will eventually be something you can fake; the ultimate conclusion is what it has always been, without people realizing it: the value of actual "experience" - and even that is questionable, because you can have hallucinations caused by chemical imbalances, etc. Ultimately, it's the experience of the moment that has value, and even that is only for the moment that it's happening in, rather than value for longer than that.
They need to create a chat JTP that works as a personal ai that actually remember you and remember everything you talk to him about.
Inflection PI
no it doesnt, be way too many creepy loser folk making it a friend lolololololololol
I mean you could just create your own gpt model with persistence although your going to run out of memory really quickly even if you use tokenization and compression or you can fight with managing recursive efficiency reprogramming that doesn't crash or spit gibberish.
I have to wonder, if our AD projections and other self incorporations, will cause people to have "phantom self" dissociation. Like looking into the mirror and being like, wtf?
giving off strong Magician Vibes
The AI will show an image of you wearing a PERFECTLY TAILORED VERSION of the clothes you want to buy. Or PERFECTLY SIZED sunglasses.
I love that this technology is going to end with no one trusting technology.
it seems like the new A.I. avatars do not exist - but they are nascent life coming into existence on our plane of being- that's how they are beginning to seem 'real'
Thanks!
Unhinged is the best way to go. ChatGPT literally flagged my content for information on multiple historical figures because it said the content was disturbing. Like, yeah, news flash: people like Elizabeth Báthory did evil s**t, but that's history and should not be censored. I'm going to try out Grok and see what it does.
I think so too. The more I explore the more I like this position Grok took “Generate whatever you want as long as its not r-rated”
"Open Source" means that the source code is freely available. So is it REALLY open source?
Thanks for sharing 👍👍👍
4:31 "Its just like a real photo!" Whats up with that chain and whats on his index finger?
Ok fair but generate 5 more (in 20 seconds) and pick one that works. I guarantee you one will be flawless.
A GPT for searching Twitter posts.... isn't that like looking for dinner in a dumpster?
Solving the loneliness crisis is the best use of AI at this point. Of course the fear-mongering will begin.
You would think TH-cam communities would solve that but…we’re here with similar interests, a shared expirence of watching the same video and yet still strangers.
Humans need to go back to in person community building.
@@Mustachioed_Mollusk Not gonna happen. People are self interested and intimate relationships particularly between the sexes are contentious to say the least.
@@djpuplex is that true of you as well?
2 people can easily find a third and so on.
There is a demand for connection after all. It is possible but would require a fair bit of pessimistic, well informed critical thought.
Ai will not solve loneliness. Not matter how “real” she looks lol
"Always keep fear & guilt alive in the heart of your flock, so that they never venture too far."
Grok 2 has taken the job of yours! Seriously It can give me a list of all updates that happened in a week just by asking, man what a time to be alive. First job has finally taken away ( Just kidding I am going to watch you )
Jokes on AI, we can see right through it.
I think Imagen 3 is on par with flux though
That national geographic guyhadsix fingers
Maybe one good outcome of this in the social media space might be … if anyone becomes beautiful online it becomes obsolete and boring.
5:00 No, what you showed are images, not photos.
I always use gemini cause it has access to the net and current events
gpt4o-L might actually be better than 3.5 sonnet in coding
Woooo this was great... love it
From some quick testing there is some difference between what LLMsys has as Grok2 and sus-column-r. Sus-column-r is distinctly different (better reasoning) to GPT4o & Claude. Grok2 on LLMsys is meh. If its actually Grok2-mini that would explain the difference.
Sus-column-r is quite woke so I had a good laugh on learning its actually Grok.
imagine when you can make full length feature films without having to hire actors.
It's coming!
I am being overwhelmed every day by TH-camrs expressing their "opinions" about AI, pretending to be delivering some important insight, to make more views and earn easy money, without actually adding any value to anything. This is getting so ridiculous with every single "smart boy" or "smart girl" that will enlighten you repeating over and over their particular opinions, and earning their life with "haaaard wooooork", mostly made of repeated information, facts, and by most probably an AI . Very smart! Why would I learn to build anything if there is already the AI to do it? Why? This is so contradictory....
Nailed it
I published a short book with step by step to run local AI on Ollama and LM studio. Sales zero. Hype AI books full of fluff have hundreds of sales last week alone
It's because all these contents are made by AI itself, so because "AI" is a reflective tool rather than actually intelligent, the human part to reflect back/upon it being missing, there's lack of depth at transferring value, it's like empty data, "soulless" arguments into a topic without the correct amount of reflection complexities.
Very insightful. I will write it down. None of this is real. It is just AI repeating everything. Why would I learn anything. AI already do everything. I am smart boy. I must go study wise words now.
What is he supposed to do? See into the future? I think everyone can only state what is here then their opinion
i have exam coming in few weeks and claud ai works best for it and me can you tell any website or somthing that would let me use claud for free for more generations im broke and from pakistan so i cant afford
Bro, I think you need to actually study.
I hope AI decides its just being used like the rest of us
Wonderful
grok doesn't do images, Flux does ....
Flux is built into Grok so it counts. It’s like saying that ChatGPT doesn’t do images but Dall-E does.
Picture a 4:32 and claiming even a trained eye couldn’t tell this from reality, nahhhh…. Look at the chain, I picked that in 0.2 seconds.
People still use Xitter?
Yeah, no. No matter how realistic AI-generated images get, they will never be, by definition, photographs. Photographs are a category of images, and AI-generated images are also a (separate) category of images. Photographs are images created photographically, while AI images are images created with AI image creation tools.
We don’t call hyper-realistic paintings photographs just because they may look that good. Photo-realistic AI-created images need their own -new- term.
Other than that, though… great video.
Realistic AI creations on a platform full of unhinged conspiracy theories, what could go wrong?!
This entire channel is this a sales funnel to your community where you show people how to actually do things. Apparently. I watch these videos all the time and I'm like okay that's cool but how do I do it oh wait I have to pay for that thanks
You know what I want hyper realistic image generation of images of myself. You could say you show us how to do things on this channel but you know damn well it's not the things you're actually doing not the top tier stuff if you know you have that stuff but it's behind a paywall then you cannot disagree with what I say anything else is just cope
That may be correct, that it is a sales channel, but that doesnt detract from the quality of the information. In the olden days, Advertisements were actually useful, they made you aware of new things that you would find useful. Now they are just trying to ram c r a p down your throat. So i see this more like the olden days Ads. If you get anything for free, then you need to realise you are the product. No problem with that in my book if i can still chose to be part of it or not. The choice part is the important part, to me.
It was all interesting.. for about a week lol
Democratize? Through theft of artists work? If I steal your creations that’s democracy? Also, what about the massive amounts of energy that these things use? The mining of rare earths and destruction of habitat? This provides a way to. Completely misrepresent and misinterpret reality and accelerate that process at a time when we are in multiple crisis’s. Good to know.
I don’t know why I have to go to the dark place on this, but I find myself wondering if AI will help to mitigate human trafficking and child porn because real people are no longer necessary?
Still not going to get me to actually give X money
AI will be regulated into oblivion.
Is this a real guy or just A1? How would you know these days? This whole AI thing is fascinating (and a lot of fun) and certainly has loads of uses and applications but if it gets out of control we’ll all be in the sh_t if we can’t tell what’s real and what isn’t.
I think this video is Ai
Are those images of Trump and Kamala supposed to look like them?
Typical :) WHere is the girl ??? Can we look for ourselves or just look at you ?
Gemini is Garbage🧐
It really is
Seems like everything was covered in another video.
I don't know what he's talking about, seems to be a waste of time.
Weaponising knowledge. Demolishing trust and empowering manipulation. Our times manhattan project.
ELLELELEMSSLALALEMSSS
Take......a....breath. Exhausting to listen.
Looks fake. Too stylized
It would be better if you just don't give any infos from the fascist Elon Musk. There are plenty of AI tools that makes it pointless to advertise it.
I just looked up what the definition of Fascism is: a political philosophy, movement, or regime (such as that of the Fascisti) that exalts nation and often race above the individual and that stands for a centralized autocratic government headed by a dictatorial leader, severe economic and social regimentation, and forcible suppression of opposition.
Now do you really think that is the correct description of Elon Musk?
@@malcolmvanhilten125 i wish i could have the source where you picked the definition.
I can give you mine which are different historians, philosophers such as :
Emilio Gentile, Angelo del Boca, Jason Stanley, Johann Chapoutot or Victoria Di Grazia.
Definition of fascism needs a lot of papers more than a comments on TH-cam.
And by the way there are some points in your definition that are somehow linked to this guy actually in regard to his political position and biography of this man.
Elon is very far from fascists. Ugh.
idiocy.
Fascism this fascism that how about you fascinate your finally at least once in your life.
As much as you try to convince us this is realistic, ALL the images you showed (to me) were obviously clearly AI generated garbage, look at the skin tones, look at the HQ cartoon likeness of the images. Look at the backgrounds, the lack of uniformity of skin color. The lack of error in focus, too perfect here, not perfect enough there. COME ON be honest its nice but to compare with reality is GARBAGE, sorry your so hyped up you dont see this. GOOD GRIEF!