0:43 - DALLE-3 general release to ChatGPT Pro 2:31 - Midjourney 2x and 4x upscalers 4:49 - Midjourney Website update 5:34 - Midjourney App 7:11 - SALMONN (speech audio language music open Neural network) 10:04 - Pixart image generator 12:15 - Hyperwrite autonomous AI assistant 14:52 - Meta AI image decoder from brain activity
@@sophia23346Don't be annoying about your beliefs. Understand that there are millions of humans that believe in different things than you and some think that your beliefs are wrong.
@@TheGoodContent37there are at least 4 billion Christians, Muslims, and Jews. They all share the same scriptures, same history, and same God. why would 4 billion people choose to follow the same history and similar statues unless a version of biblical history is accurate. Humans are very forgetful, fickle, and vain. Most belief systems are forgotten but Christianity made pagans abandon their country's gods risking death and execution in favor of Christianity. Human nature is to gain and indulge in all manners of pleasure and desire. But Christianity teaches denial and losing for the sake of others. What logic does it make for Christianity to become the predominant belief system in a world that was selfish and evil. It must be because their was truth in what Jesus was saying
@@MattVidProI wouldn't mind, I'd actually find it interesting. As long as you'd be honest about it, it'd be cool. You have a nice sounding voice so the cold does hurt a bit.
I am very impressed by sound recognition. Current systems just give a text and these new models will be able to understand our emotions, accent, could check our pronunciation, much more than just raw text.
Matt, your video on the latest developments in AI was awesome. You're my number one source for staying updated on the most recent advances in this exciting field. If you have the time, it would be great if you could consider making another video on free AI alternatives for image generation in the future. Keep up the excellent work! 🙌🤖
What I think is really cool about MEG from Meta is that it interprets your brainwave patterns to give you the image you are looking at. And I find it really fascinating that when you are looking at a bee, for example, it gives you an image of an insect occasionally mixed with a bird. I don't know if it is how I think it is but maybe its not the AI thats making mistakes, but our brain that really interprets a bee as a serie of different qualities that are then picked up as and transformed into brainwaves that can be read and translated as "bee = bird, small, sting, insect, furry, yellow and black..." And that these keyword that come after the word bee are what create a "Bee" in our brain. I really find it dreamy that it might simply be that our brainwaves dont have a real visualization of a bee but a serie of these qualities that we interpret as a bee but a computer doesnt. Of course it might not be like I think. Actually im pretty sure that a neuroscientis would love to disagree with me but, again, I just find it amazing to think.
Ever since I first saw GPT-4V's capabilities, I've been looking forward to when it'll be able to hear, too. SALMONN (one hell of a backronym, by the way) is exactly the kind of thing OpenAI need to add to GPT-4's base model! Imagine it learning from text, images, *and* audio? Especially if trained on video data, so it would get image-text-audio correlations to learn from. It would be so intelligent and capable! ...now I just wish the SALMONN online demo link worked (the Gradio UI is shut down, apparently). Maybe I'll stop being lazy at some point today and go to my office to clone it myself on my desktop... maybe... *EDIT* I spoke too soon; looks like it was only down temporarily. Playing with it now! *EDIT 2* Okay, so... my first impressions are it's neat, but it hallucinates too much to be useful. I tested it with songs, and while it had no trouble describing the songs acoustically, if I asked it for lyrics, it would either hallucinate totally made-up lyrics, or even flat-out insist there is no audio so it can't transcribe lyrics. But hey, it's the first public instance of audio input modality on an LLM, and it's based on Vicuna 13B, so I imagine it will just get better as similar things are scaled and integrated into larger models (*cough* hurry up OpenAI *cough* )
Yeah, don't feel bad, every time you catch a cold, you're developing immunity for the particular cold virus. So by the time you are 60, you should catch very few colds! 😊
First of all, thanks for including chapter links for this video! I listened to the music.wav audio that the audio to text described in so much detail and was impressed by the interpretation and description that it gave. But only the last 2 seconds of the clip contained the human singing, so the description did (to me) seem to emphasize that a more than might be appropriate. I guess that's nit picking!
Thank you for once again informing us all to recent inovations that have emerged within the field. Speaking of which, have you yet been alerted to The University Of Chicago's proprietary "Nightshade" project? Perhaps that ought to be a subject discussed by you.
Love how were splitting hairs about the imagine generation quality when just a couple months this never existed. Crazy how fun and spoiled oner feels with all this rappid news and advanced. Best time to be alive right now to witness this all happen. Thanks Matt!
FYI - The Niji - Journey app only lets you create regular MJ images using 5.0 and nothing higher than that. So you can't generate 5.1 or 5.2 images...which kind of stinks. Not to mention the overall UI is VERY crammed and not very intuitive or user friendly at all. However as a veteran user and member of Midjourney I'm grateful that they've even released this and also understand it's still a very new app.
In ChatGPT 4 can I take an image I painted and ask for it to reimagined using Dalle-3 in there? I know you can do it in MJ but seems you can't in normal Dalle-3.
Strongly disagree. As graphic designer I work with DALL-E and Midjourney on a daily basis and midjourney generates way more diverse range of styles and better photography quality. Also I never understood why people so strongly want to get rid of Discord. Discord opens way more possibilities than a normal interface on a website. I can organize different groups, add my buddies so they can have a peak of what I am generating, I can quickly send people pictures or videos of my creating progress when I combine AI with photography, rendering etc. I can install plugins which interact with the MJ bot. And much more.
OpenAI terminated my account, as well as others, without giving an understandable reason. Is DALL-E 3 with GPT Plus better than what I still have with Bing? It sucks being kicked to the sidelines when I didn't do anything. I appealed the termination about a week ago to no avail. Hopefully Gemini will be a good replacement.
I find the skepticism with Hyperwrite to be extremely interesting, but not surprising especially in the US. I feel like at this point, there's a collective hesitation against what big tech (or big companies in general) publicly say, since everything a business says effects their bottom line. That being said, when added to skepticism behind AI and it's threat/safety in general, it makes sense why people wouldn't be willing to give an AI access to their information, even if its free. Since, then, at the end of the day, in the mind of someone not fully invested into AI and how it works, their perspective is that now the AI AND a big company have access to their personal account. Ultimately, I don't think its a distrust in the credibility of the service, it's more of a distrust in a system in the wake of potentially dangerous technology that is growing rapidly.
I hope Apple learns from these awesome innovations in speech recognition by other companies because theirs is truly horrendous. I'm disabled so I use it a lot and it's pretty bad.
Matt, I love your videos. You do a great job covering AI. However, your negative bias against MJ is obvious and unfair in my opinion. I'm not sure what MJ did to you other than maybe charging for any real access but you seem to have a hard time saying nice things about MJ. Is Dalle 3 good? Yes. Is it better than MJ? Yes on some things like text. However, MJ in my opinion (from someone who has made substantial money selling AI design-based products) is still superior in many ways. There's plenty of room for two great AI graphic generators. It's like the difference between Photoshop and Illustrator. They're different tools that do different things but they both make great graphics. Try to balance your reporting and not make Dalle-3 sound so much better than MJ. It's not, just different.
Dude when you've got the most nasal voice on the internet the upside is nobody knows if you have a cold or not. You just sound like normal you today i.e. super nasal bros wonder 🥳
• I believe we are meant to be like Jesus in our hearts and not in our flesh. But be careful of AI, for it is just our flesh and that is it. It knows only things of the flesh (our fleshly desires) and cannot comprehend things of the spirit such as peace of heart (which comes from obeying God's Word). Whereas we are a spirit and we have a soul but live in the body (in the flesh). When you go to bed it is your flesh that sleeps but your spirit never sleeps (otherwise you have died physically) that is why you have dreams. More so, true love that endures and last is a thing of the heart (when I say 'heart', I mean 'spirit'). But fake love, pretentious love, love with expectations, love for classic reasons, love for material reasons and love for selfish reasons that is a thing of our flesh. In the beginning God said let us make man in our own image, according to our likeness. Take note, God is Spirit and God is Love. As Love He is the source of it. We also know that God is Omnipotent, for He creates out of nothing and He has no beginning and has no end. That means, our love is but a shadow of God's Love. True love looks around to see who is in need of your help, your smile, your possessions, your money, your strength, your quality time. Love forgives and forgets. Love wants for others what it wants for itself. Take note, love works in conjunction with other spiritual forces such as faith and patience. We should let the Word of God be the standard of our lives not AI. If not, God will let us face AI on our own and it will cast the truth down to the ground, it will be the cause of so much destruction like never seen before, it will deceive many and take many captive in order to enslave them into worshipping it and abiding in lawlessness. We can only destroy ourselves but with God all things are possible. God knows us better because He is our Creater and He knows our beginning and our end. Our prove text is taken from the book of John 5:31-44, 2 Thessalonians 2:1-12, Daniel 7-9, Revelation 13-15, Matthew 24-25 and Luke 21. Let us watch and pray... God bless you as you share this message to others.
If you whole personality within the AI community, specifically image gens, doesn't center around existing IP and making memes then using Dalle in Chatgpt is the best way to use it. If all you are seeking to just create existing characters and Ips or other "stuff" (you know what I mean) then yeah... you'll need to go elsewhere for that.
0:43 - DALLE-3 general release to ChatGPT Pro
2:31 - Midjourney 2x and 4x upscalers
4:49 - Midjourney Website update
5:34 - Midjourney App
7:11 - SALMONN (speech audio language music open Neural network)
10:04 - Pixart image generator
12:15 - Hyperwrite autonomous AI assistant
14:52 - Meta AI image decoder from brain activity
What a LEGEND!
@@sophia23346Don't be annoying about your beliefs. Understand that there are millions of humans that believe in different things than you and some think that your beliefs are wrong.
@@TheGoodContent37there are at least 4 billion Christians, Muslims, and Jews. They all share the same scriptures, same history, and same God. why would 4 billion people choose to follow the same history and similar statues unless a version of biblical history is accurate. Humans are very forgetful, fickle, and vain. Most belief systems are forgotten but Christianity made pagans abandon their country's gods risking death and execution in favor of Christianity. Human nature is to gain and indulge in all manners of pleasure and desire. But Christianity teaches denial and losing for the sake of others. What logic does it make for Christianity to become the predominant belief system in a world that was selfish and evil. It must be because their was truth in what Jesus was saying
Use ai voice to get back to your normal voice😅
TBH i thought bout it!
Yeah, but nah. Im not THAT hardcore an audiophile.
@@MattVidProwhy wouldn't you? It's literally the exact use case
Your audience would appreciate the most. 😅
@@MattVidProI wouldn't mind, I'd actually find it interesting. As long as you'd be honest about it, it'd be cool.
You have a nice sounding voice so the cold does hurt a bit.
Opportunity missed. 😢
I am very impressed by sound recognition. Current systems just give a text and these new models will be able to understand our emotions, accent, could check our pronunciation, much more than just raw text.
What piece of news did you find most interesting?
The news that you have a sore throat and you are giving us content anyway means a lot to me and THANK YOU ❤
The mind reading… miniaturising happens fast these days. Although I’d guess that machine uses big magnets…
Matt, your video on the latest developments in AI was awesome. You're my number one source for staying updated on the most recent advances in this exciting field. If you have the time, it would be great if you could consider making another video on free AI alternatives for image generation in the future. Keep up the excellent work! 🙌🤖
Next time you have a cold you should bring out the Matt voice clone 🤣 Hope you feel better soon, I always get excited seeing new uploads from you!
Lol, that would be so broken. Imagine if he has a sore throat and loses his voice, but has the AI cloned voice handy 😁
he could use Voicify to overlay his usual voice over it, it may sound quite seamless
Always appreciate your updates Matt! Thanks so much!
I watch a lot of ai channels , but your channel is just different ❤
Thank you!!!
"Do you like the channel?"
"It's, um... different." :D
Your voice still sounds great!😊 thanks for the updates, u rock!!!
hey Matt, great videos, I'm so glad I subscribed, it's paying off big time. Love your channel!
Welcome aboard!
What I think is really cool about MEG from Meta is that it interprets your brainwave patterns to give you the image you are looking at. And I find it really fascinating that when you are looking at a bee, for example, it gives you an image of an insect occasionally mixed with a bird.
I don't know if it is how I think it is but maybe its not the AI thats making mistakes, but our brain that really interprets a bee as a serie of different qualities that are then picked up as and transformed into brainwaves that can be read and translated as "bee = bird, small, sting, insect, furry, yellow and black..." And that these keyword that come after the word bee are what create a "Bee" in our brain.
I really find it dreamy that it might simply be that our brainwaves dont have a real visualization of a bee but a serie of these qualities that we interpret as a bee but a computer doesnt.
Of course it might not be like I think. Actually im pretty sure that a neuroscientis would love to disagree with me but, again, I just find it amazing to think.
Aye hope you feel better bro. I heard drinking 🍋tea helps
I've been drinkin tea! Appreciate it
Ever since I first saw GPT-4V's capabilities, I've been looking forward to when it'll be able to hear, too. SALMONN (one hell of a backronym, by the way) is exactly the kind of thing OpenAI need to add to GPT-4's base model! Imagine it learning from text, images, *and* audio? Especially if trained on video data, so it would get image-text-audio correlations to learn from. It would be so intelligent and capable!
...now I just wish the SALMONN online demo link worked (the Gradio UI is shut down, apparently). Maybe I'll stop being lazy at some point today and go to my office to clone it myself on my desktop... maybe...
*EDIT* I spoke too soon; looks like it was only down temporarily. Playing with it now!
*EDIT 2* Okay, so... my first impressions are it's neat, but it hallucinates too much to be useful. I tested it with songs, and while it had no trouble describing the songs acoustically, if I asked it for lyrics, it would either hallucinate totally made-up lyrics, or even flat-out insist there is no audio so it can't transcribe lyrics. But hey, it's the first public instance of audio input modality on an LLM, and it's based on Vicuna 13B, so I imagine it will just get better as similar things are scaled and integrated into larger models (*cough* hurry up OpenAI *cough* )
You are a champ for still making a video when you’re sick - kudos!
Also, I hope you feel better soon!
Yeah, don't feel bad, every time you catch a cold, you're developing immunity for the particular cold virus. So by the time you are 60, you should catch very few colds! 😊
First of all, thanks for including chapter links for this video!
I listened to the music.wav audio that the audio to text described in so much detail and was impressed by the interpretation and description that it gave. But only the last 2 seconds of the clip contained the human singing, so the description did (to me) seem to emphasize that a more than might be appropriate. I guess that's nit picking!
Thank you and hope you get back to full vocals soon 😊👍
Matt, nice work as alway! truly appreciate your update 💫🤟🤟🤟💫
I appreciate that!
Get well soon and as always. Thank you for all that you do for us. 🎉
Thank you for once again informing us all to recent inovations that have emerged within the field. Speaking of which, have you yet been alerted to The University Of Chicago's proprietary "Nightshade" project? Perhaps that ought to be a subject discussed by you.
I saw this, imo it won't be a big splash but it's on my radar!
@@MattVidPro Thank you. May you be well.
I'm really looking forward to Generative Music A.I, I was playing around with Riffusion and I think it's absolutely promising!
amazing stuff like always!❤
thank you!
Very informative review, it seems some revolutionary changes happen each week now. Several years ago each of these changes could be discussed a lot.
Absolutely!!!
Hope you get better soon, thank you for making videos for us but tbh id rather you relax and feel better.
May God heal you, my brother Matt, I am from Morocco❤
omg uploaded 2 minutes ago LET'S GOOOO
yess
Love how were splitting hairs about the imagine generation quality when just a couple months this never existed. Crazy how fun and spoiled oner feels with all this rappid news and advanced. Best time to be alive right now to witness this all happen. Thanks Matt!
I love chatting with AI that has voices. Been spending hours and hours
I would recommend you check out the ACE framework made by David.
You just knew dalle 3 would fuck up along the way. Hopefully others aren't far behind with converging LLMs and image creation
Video bitrate got nerfed significantly, please start uploading in 4k. It's very blurry
FYI - The Niji - Journey app only lets you create regular MJ images using 5.0 and nothing higher than that. So you can't generate 5.1 or 5.2 images...which kind of stinks. Not to mention the overall UI is VERY crammed and not very intuitive or user friendly at all. However as a veteran user and member of Midjourney I'm grateful that they've even released this and also understand it's still a very new app.
who are you kidding you know that we only use dall-e 😂
Can Salmonn generate the sheet music for a piano playing in the background of a audio file?
Matt could have let Matt-AI do the talking?
I thought abt it!
Can you do a roundup of all the medical AI advancements and tools available to the public?
Damn I was really hoping PixArt was gonna be for pixel art...dall does great but it's still not actually using pixel perfect pixels.
Played about with Niji but no matter how hard I try I still get images with an anime feel, any tips?
In ChatGPT 4 can I take an image I painted and ask for it to reimagined using Dalle-3 in there? I know you can do it in MJ but seems you can't in normal Dalle-3.
Hello Matt, why isn't there a video of ChatDev? Please make one...
Strongly disagree. As graphic designer I work with DALL-E and Midjourney on a daily basis and midjourney generates way more diverse range of styles and better photography quality.
Also I never understood why people so strongly want to get rid of Discord. Discord opens way more possibilities than a normal interface on a website. I can organize different groups, add my buddies so they can have a peak of what I am generating, I can quickly send people pictures or videos of my creating progress when I combine AI with photography, rendering etc. I can install plugins which interact with the MJ bot. And much more.
Thanks matt.
Any time!
OpenAI terminated my account, as well as others, without giving an understandable reason. Is DALL-E 3 with GPT Plus better than what I still have with Bing? It sucks being kicked to the sidelines when I didn't do anything. I appealed the termination about a week ago to no avail. Hopefully Gemini will be a good replacement.
I find the skepticism with Hyperwrite to be extremely interesting, but not surprising especially in the US. I feel like at this point, there's a collective hesitation against what big tech (or big companies in general) publicly say, since everything a business says effects their bottom line. That being said, when added to skepticism behind AI and it's threat/safety in general, it makes sense why people wouldn't be willing to give an AI access to their information, even if its free. Since, then, at the end of the day, in the mind of someone not fully invested into AI and how it works, their perspective is that now the AI AND a big company have access to their personal account.
Ultimately, I don't think its a distrust in the credibility of the service, it's more of a distrust in a system in the wake of potentially dangerous technology that is growing rapidly.
I had no idea MJ had a website, I heard about the application, though.
Good video mate 🎉
Glad you enjoyed!
what is best ai local we can use now days ?
You need some lemon tea. 🍋
This dude making videos for all of us even he is a little sick. Respect❤
So instead of typing my reservation info into Open Table, I can type it in Hyperwrite. Cool.
Dood its the perfect occasion to use ai voice cloning, you kinda blew it
I could always go for a good mind blow 👍
If only SD had a "Dalle 3" like diffusion.
Get well soon, Matt 💋
Should have used the ai to give you a voice this whole video. would have been cool
AI seems backwards.
It's as if AI is: "For children use only, no adults allowed"
eh i still like sdxl because local
I hope Apple learns from these awesome innovations in speech recognition by other companies because theirs is truly horrendous. I'm disabled so I use it a lot and it's pretty bad.
Hello good day to everyone, someone Can tell me the Best image generaror?
Hot Tea ☕
It is kinda ironical that Meta AI image can decode from brain activity. I wonder what they will use it for?
Can't think of a better use of AI than to give fucking Facebook mind reading capabilities.
I love AI, but I don't love this one bit.
Elon supports Midjourney my guess is that X will integrate Midjourney into X in the next months.
Your voice sounds better
💀💀💀💀
very cool
The new midjourney site is phenomenal. I can now find my job IDs much faster. Exactly what I needed.
Thanks!
MattCOVIDPro?
👏👍💕 Learning something everyday! 🍵🍜
Absolutely!
Dalle3 did not “dethrone” Midjourney. MJ is still much superior
hi sweeto
Cool video matt I really liked today's ai news and hope you'll recover
Take a good rest till you're fine man and take care of yourself
#Free_palestine
Matt, I love your videos. You do a great job covering AI. However, your negative bias against MJ is obvious and unfair in my opinion. I'm not sure what MJ did to you other than maybe charging for any real access but you seem to have a hard time saying nice things about MJ. Is Dalle 3 good? Yes. Is it better than MJ? Yes on some things like text. However, MJ in my opinion (from someone who has made substantial money selling AI design-based products) is still superior in many ways. There's plenty of room for two great AI graphic generators. It's like the difference between Photoshop and Illustrator. They're different tools that do different things but they both make great graphics. Try to balance your reporting and not make Dalle-3 sound so much better than MJ. It's not, just different.
My first thought was that your voice sounds badass! 😎 ...but then you've apologized for it.
Bill Gates is saying that AI is stagnating
I found Bing image generator to be much worse with censorship
Why am I not comfortable with the last one?💀
its spooky tbh
Bro your Mic is the size of your HEAD!
It’s mostly the pop filter that’s so large 😂
@@MattVidPro Haha just kidding bro, it's probably just closer to the camera. love your videos, good stuff
Get better!
first 15 minute lurkers can like this
👋
MJ is still no1 for what I do😊
use a voice AI with your orignal voice and lip sync it
You shouldn’t drink and smoke so much 😉
Dude when you've got the most nasal voice on the internet the upside is nobody knows if you have a cold or not. You just sound like normal you today i.e. super nasal bros wonder 🥳
"I have a cold" - No excuse. You have AI models of your voice. Use them to sound normal.
honestly the way people feel about kwebbelkop is how i feel abount vtubers there is no humanity inly money
Yeah there needs to be a creative drive SOMEWHERE. I've always been obsessed with the youtuber idea that's why I like to have my own set up, mic, ect.
@@MattVidPro i think its because of parasocial relationship
hyperwrite sounds like a scam. never give ai personal information
Midjourney is so lame. It's like they have no clue how to run a business.
• I believe we are meant to be like Jesus in our hearts and not in our flesh. But be careful of AI, for it is just our flesh and that is it. It knows only things of the flesh (our fleshly desires) and cannot comprehend things of the spirit such as peace of heart (which comes from obeying God's Word). Whereas we are a spirit and we have a soul but live in the body (in the flesh). When you go to bed it is your flesh that sleeps but your spirit never sleeps (otherwise you have died physically) that is why you have dreams. More so, true love that endures and last is a thing of the heart (when I say 'heart', I mean 'spirit'). But fake love, pretentious love, love with expectations, love for classic reasons, love for material reasons and love for selfish reasons that is a thing of our flesh. In the beginning God said let us make man in our own image, according to our likeness. Take note, God is Spirit and God is Love. As Love He is the source of it. We also know that God is Omnipotent, for He creates out of nothing and He has no beginning and has no end. That means, our love is but a shadow of God's Love. True love looks around to see who is in need of your help, your smile, your possessions, your money, your strength, your quality time. Love forgives and forgets. Love wants for others what it wants for itself. Take note, love works in conjunction with other spiritual forces such as faith and patience. We should let the Word of God be the standard of our lives not AI. If not, God will let us face AI on our own and it will cast the truth down to the ground, it will be the cause of so much destruction like never seen before, it will deceive many and take many captive in order to enslave them into worshipping it and abiding in lawlessness. We can only destroy ourselves but with God all things are possible. God knows us better because He is our Creater and He knows our beginning and our end. Our prove text is taken from the book of John 5:31-44, 2 Thessalonians 2:1-12, Daniel 7-9, Revelation 13-15, Matthew 24-25 and Luke 21. Let us watch and pray... God bless you as you share this message to others.
Take care of yourself, would you? It's not very wise to talk enthusiastically so much when your voice is hardly even recognizable =__=
PixArt-alpha's inference requires 23GB of GPU memory.
If you whole personality within the AI community, specifically image gens, doesn't center around existing IP and making memes then using Dalle in Chatgpt is the best way to use it. If all you are seeking to just create existing characters and Ips or other "stuff" (you know what I mean) then yeah... you'll need to go elsewhere for that.