Digging Up Recent Overlooked AI News!
Vložit
- čas přidán 10. 05. 2024
- ▼ Link(s) From Today’s Video:
Adobe Firefly 3: / 1782711515537174576
firefly.adobe.com/
PAID SD3 :( clipdrop.co/stable-diffusion
Llama 3 16k: / 1782576964118675565
Adobe's new Upscaler: / 1782376538429649067
LLama 3 on groq app: / 1781420183099466237
TLDraw Twitter: / tldraw
► MattVidPro Discord: / discord
► Follow Me on Twitter: / mattvidpro
-------------------------------------------------
▼ Extra Links of Interest:
✩ AI LINKS MASTER LIST: www.futurepedia.io/
✩ General AI Playlist: • General MattVidPro AI ...
✩ AI I use to edit videos: www.descript.com/?lmref=nA4fDg
✩ Instagram: mattvidpro
✩ Tiktok: tiktok.com/@mattvidpro
✩ Second Channel: / @matt_pie
-------------------------------------------------
Thanks for watching Matt Video Productions! I make all sorts of videos here on CZcams! Technology, Tutorials, and Reviews! Enjoy Your stay here, and subscribe!
All Suggestions, Thoughts And Comments Are Greatly Appreciated… Because I Actually Read Them.
-------------------------------------------------
► Business Contact: MattVidProSecond@gmail.com - Věda a technologie
Firefly is trained on public domain images which make its output 100% commercial safe
massive point of consideration, this channels talking head is full of gas.
However, their dataset was not seen by anyone. No one will, either.
@@bonecircuit9123 this CZcamsr's ego has gotten too high 😂
Absolutely no matter what the model is trained on.... No model uses parts of other people's images when generating new ones. So any model is 100% safe for commercial use. But firefly is terrible. SDXL is the best model right now and absolutely free besides.
They claim that,but we have no seen the dataset,so it;s entirely possible it's uses copyrighted images,personally I don't really care but they might have a problem in the future.
It's fascinating how if you would have generated these images a year ago, you'd have been all "Holy Cow, that's awesome!" But today they are "Meh."
its crazy how fast thing are progressing that it only took a year for us to get disillusioned with this
bro… at the very min it’s really good that Adobe keeps getting. better cause it forces Midjourney and DallE to keep getting better too to stay ahead… competition always drives innovation… open ai still hasn’t put gpt 5 cause people are barely starting to catch up to them now
Thanks for reporting on all AI; big and small. As always, great video.
After 4 days thought you were dead
Thanks bro good quality news are rare.
I also tried he SD3 on ClipDrop like you do in this video but I found it even more censored than the previous version. I asked it for women wearing specific outfits and it called them all NSFW. I didn't ask for NSFW, it just made them that way. I didn't ask for nudes or anything, so I found it useless in a short test due to censorship.
The entire concept of censoring these models is madness. In 10 years we'll look back at such prudish bs as ludicrous and borderline mad. It's like trying to hold back the tide with magic chanting and slogans.
Yes SD3 API is way too censored even for perfectly sfw prompts. This is why I’m hoping they do release the weights soon
Hey, thanks for giving your thoughts on the video! I was also a little underwhelmed by Adobe Firefly. I think you make some great points, especially about the current trajectory of SD3. I'll definitely be keeping an eye on Matt Schumer's work - that context window upgrade sounds awesome.
What is your name a reference to?
@@pattyprolapse Insightful overview of the latest AI developments! It's fascinating to see the rapid progress in areas like image generation and upscaling. Excited to explore the open-source possibilities mentioned here.
Matt just so you know, you are one of the most important people in my life, tech has become a huge influence on a daily basis and your content enlightens me up, I'm not even a native english speaker, but I love the content you do in english, and it inspires me to learn even more english.😊🎉
I love the NEW tech stuff that is like a new research paper but its some small obscure tech. It opens my mind to possibilities. Whereas some others show a bunch of apps that seem to be useful and somewhat polished and popular yet they are pretty average on the newness factor. I love what some of these imply about the future.
8:01 - That's very impressive, when are we likely to see this in Adobe Photoshop?
I need a outpain ai, or also inpaint included, anyone know a good one?
Phi 3!!!!, also hugging face seems to be down
The only person I get my Ai news from ! :-)
I've been debating getting gpt pro again. It's mostly the ability to upload just about any file to chat gpt and the built in code running thing.
Not to mention the quality of gpt 4 still hasn't been topped for free.
Why ideogram don't have inpainting feature, it is very important
Lykon claims that the API version of SD3 is an older model and that the good version of SD3 will be open-sourced soon. I hope he's right, but I'm not sure how much influence he really has at the company.
Id love to see Llama 3 on my phone, if you can show us how that would be awesome.
it would be fun your idea for videogame ai
Llama 3 hallucinates a lot.
We don't know whether the technological singularity is coming or not. But in the case it should ever come, then soon before it somebody that makes daily videos about everything AI would have to make a compilation video called _"Digging Up Recent Overlooked AI News"_ because they could not manage themselves to keep track of everything on time.
Show the install of llama pls
You should post affiliate-links to your equipment, especially this sick mic, in your video-notes - you could earn a good addon, I guess... 😉
can you plz try Hakko AI
Please! Make the vidoe of how to run Llama 3 locally.
things have really slowed down. just a ton of LLMs that are all about as good as each other. not the exponencial change we were expecting.
Thank you for pointing that out. Everyone is on the AI bandwagon and claiming that things are moving lightning fast. Excuse me. We are seeing proof of concept items beyond LLMs and image generation. We need the next AI product to actually DO SOMETHING and not just promise to do so in the near future!!
The answer is simple...its come as far as it can go and is fizzling out...!!🤣🤣🤣🤣🤣🤣
Big companies are slowing down and are not releasing because society isn't simply ready.
OpenAI has a audio model they haven't released since 2023 because it could create deepfakes. Sora is known to the world, but still not open to the public. Udio, the new music AI service, was created by former Google Deepmind employees, who made this type of technology since probably 2023. The AI Apps we have access on the internet are probably a fraction of what is possible with LLMs inside huge corporations like Google or Microsoft.
I think the only way forward is smaller companies trying to get ahead of the competition by making products that bigger ones were not ready to release (like Udio), and then bigger companies feels its time to release them to the public.
All in all, the AI revolution is being delayed because governments, the media and society in general because they are simply not ready or aren't aware of the life-changing potential.
This is gonna age poorly when GPT-5 is announced.
...and then the regulations will come, and we will have no more Innovation for a year.
You will be seeing more change once Nvidia Blackwell starts rolling out
SD3 will be better only after open source community efforts to refine it. The same as it was with SD 1.5 and XL.
Please, show us how to run on local machines!
You got a Video about deepfake video Faceswapp ?
Nice
yay
💯 Why I've always loved Warren G and always will!
Damn you were hella negative on adobe
Firefly is free for people with the adobe suite already no? so that's the reason
so not free at all, lol
@@ouroborostechnologies696 ye, unless you're still in school, so that's their target audience with this
❤
I heard if you said your favorite CZcamsr’s name three times, he’ll heart your comment!
MattvidPro
MattvidPro
MattvidPro
Edit: NO WAY HE ACTUALLY HEARTED IT 🤩🤩🤯🤯
Edit 2: NOOOO HE UNHEARTED 😭😭😭😭😭
if you edit your comment then it removes hearts by the creator...
@@hypercoder-gamingRookie’s mistake
@@hypercoder-gaming bro…..
Made me laugh
😂😂😂
👋
AI content is gonna stay meh until GPT-5. But for real Copilot is lobotomized and all the search bots are uninteresting. There's a stangnancy going on. Tiny LLM score wars until the end of 2024.
Locallfree!
So how many killers in the room? I guessed 3……. ❤️🇦🇺
Ai bubble is bursting. It's the end. A lot of pies in the sky (sora, sd3) but nothing I can touch or work with rn.
it may be like the 2000 web bubble that bursts fast but is part of everyday life 5-10 yrs later
Too many damn commercials man
I used SD3 and it sucks
SD 3 is superior superior to MJ 6 in a number of ways including text generation and prompt coherence
Yeah but for long prompts only Dalle-3 really works
@chanpasadopolska
Depends.........I access Dall-E 3 via Bing Image Creator (BIC) for 98% of my usage. In fact, I've used CoPilot to create shortened versions of prompts which are more than 480 characters (spaces and punctuations included) as this is the BIC limit. There's very good reasons to use BIC vs any other access point for Dall-E.
naw stable diffusion 3 sucks an its super censored
@@vegax5589
I suppose it's all subjective. I've gotten some amazing results with SD3.......better than MJ 6 is quite a number of goals.
Besides, once the model is available via forks like AUTOMATIC 1111 and Comfy & other access points, that will be big.
Tid bits and nuggets 💀🤮🤮🤮
First.
Experts predict Ai will never actually kick out humans and will fizzle out on its own. 🤣🤣🤣🤣
Firefly is dreadful - all the pictures look shiny and soulless and the engine never ever does what you tell it to. Stable Diffu$ion is absolutely crap.
whats wrong with using midjourney images to train AI?
Let’s face it. LLM were the big thing and they’re showing heavy sign of technology plateau. More parameters leads to more hallucination rate and models aren’t better so we won’t be able to train multimodal models with giga trillons tokens. So we may have that videos games feel = real time 3D never came close to real life photorealistic in 20 years. Yann Lecun is right even if I don’t like this conclusion = we’ll need a lot more that only computational power to achieve AGI. And so we’re not sure it could happen in our life time now. BTW I had the feel it was close and it was great. Now it’s deception. LLM aren’t reliable and just toys or can achieve non critical simple tasks.
Obviously hallucination is a problem, but nothing has given me the impression that more parameters = more hallucination. Where are you pulling that from?
@@justtiredthings papers I spent months to read… it’s showing how training can affect models, and more training can kill the performances, but also the number of parameters, the layers, de dilation, and so many more parameters ..
The transformer architecture by itself has not been improved since the first models and research is not very productive in a good way.
What I understand is that to get more sustainable evolution we could train billions of LLM models with slight changes and let it evolve with better benchmarks. But that would require more than six order of magnitude more GPU horsepower than B100 clusters….
When improving something in the models you have to run very expensive training to see how it affect your model evolution… that’s the main issue to go faster. So decades long plateau is coming. The famous AI winter scientists were talking about 2 years ago…
That’s not what I was expecting. That’s very disappointing as I KNOW humanity will achieve AGI in the next century with all the benefits (life expectancy, abundance, big scientific discoveries… maybe the true nature of the universe or the answer to Fermi paradox…)
Have to live longer to see that
@@notalkguitarampplug-insrev784 from everything I've seen, hallucination seems to be reduced a bit as models gain more world knowledge and parameters. Small parameter models are famously bad with factual accuracy.
But I wish you were right, because AGI is a horrifying outcome, and we should be doing everything we can as a species to avoid creating it.
Wait what. GPT 4 level performance with a 70B model and we're not even past Q2 2024
@@markonfilms i strongly encourage you to have fun with Llama 3 70b using groq.
It’s faster than any human could be but still less accurate and reasoning is not on par with GPT4 (and even less with current version)….
Definitely do a tutorial on how to get this going locally plz🫶