![Py Man](/img/default-banner.jpg)
- 69
- 91 649
Py Man
India
Registrace 15. 01. 2022
Explore AI, open-source tools, and tech tips. Discover generative art, chat with advanced models like ChatGPT, and stay informed with the latest AI news. Learn coding tricks and dive into the world of AI with us!
Q* OpenAI's Doomsday Algorithm That Terrified Its Creators😱
#ainews #chatgpt #openai
Q*: The Four-Character Code That Could Change Humanity Forever"
Four little characters - Q* - doesn't look like much, huh? But at OpenAI, it's causing more drama than a season finale of your favorite show. This tiny code might just be the key to super-smart AI. The future's here, folks, and it's kind of freaking everyone out!
Q*: The Four-Character Code That Could Change Humanity Forever"
Four little characters - Q* - doesn't look like much, huh? But at OpenAI, it's causing more drama than a season finale of your favorite show. This tiny code might just be the key to super-smart AI. The future's here, folks, and it's kind of freaking everyone out!
zhlédnutí: 2 148
Video
Claude 3.5 Sonet Beats GPT-4o: Mini AGI is Here to Change Everything!
zhlédnutí 561Před dnem
#llm #AI #claude3 #ainews Anthropic Claude 3.5 Sonet Just Shock entire world. Not only is claude sonet smarter than other AI Model, but also twice as fast, making it perfect for Complex Task. it's coding capebalities just jaw dropping. Link: Claude 3.5 Sonet Details: www.anthropic.com/news/claude-3-5-sonnet
Run Stable Diffusion 3! FASTEST Way to run the new Model With Lighting AI🥳
zhlédnutí 63Před dnem
stable diffiusion 3 is best opensource model for generating image. this video will guide you how to install comfyui and run sd 3 on lighting studio. installation command: wget header="Authorization: Bearer {YOUR_HUGGING_FACE_AUTH_TOKEN}" {here past link}
Qwen-2 : The BEST Opensource LLM is here! (Beats Llama-3, GPT-4O, Claude)
zhlédnutí 127Před 14 dny
#ai #llm In this video, We'll be talking about the newly released Qwen-2 models. The models include 0.5B Parameter model, 1.5B Parameter Model, 7B Parameter model and upto 72B Parameter model. The first three models can even run on computers with out any need of GPUs which is amazing. You can make very simple and easy to use Copilot with it as well. It beats every other model such as Llama 3, C...
Devon : Open-Source AI Software Engineer😲(Install Locally)
zhlédnutí 170Před 21 dnem
In this video i will show you how to install devon , devon is a ai software engineer that that can reason and plan everything from scratch. Devin is Powerfull but not opensource. but devika is fully opensource. also watch : czcams.com/video/n-kxpNgso0M/video.htmlsi=_OqMjkhRuJlEioNt czcams.com/video/Vu1edGrrwiA/video.htmlsi=b6BtxXOmomKDl3V2 czcams.com/video/0C0Kroq9n_E/video.htmlsi=c0mrZ2Q4GuzVodPr
OpenAI's GPT-4o SHOCKS Entire World 🤯
zhlédnutí 1,2KPřed měsícem
New ChatGPT talks so insanely human-like, it's basically cracked the code for AGI. We're talking crazy natural conversations, mind-melting intellect and near zero latency . After experiencing ChatGPT's skills first-hand, you'll never look at AI the same way again. This sh*t is next-level future stuff.
Flexclip : Easy to Use Text to Video Generator
zhlédnutí 121Před 2 měsíci
FlexClip is a groundbreaking browser-based video editor powered by AI. Create entire videos from text with the text-to-video generator. Turn articles into video summaries instantly. Generate custom images and voice overs on demand. Edit with motion graphics, effects, and more - all in your web browser. FlexClip's mind-bending AI toolkit makes video creation faster and easier than ever before. l...
Microsoft's VASA-1: Make Anyone Say Anything
zhlédnutí 5KPřed 2 měsíci
Microsoft just dropped VASA-1, an AI that cranks out videos from single images. It's their play to stay in the generative AI game. LINKS: www.microsoft.com/en-us/research/project/vasa-1 Aniportrait (opensource Vasa 1 ) : czcams.com/video/7vpCZ5_MaTg/video.htmlsi=Y3g3DJ_yaYN2wk83
China's SenseNova 5.0 SHOCKS Entire World! ( True GPT 4 KILLER )
zhlédnutí 2,9KPřed 2 měsíci
You won't believe what this Chinese AI company just dropped! SenseTime unveiled insane new language models that crush GPT-4, plus sick multimodal skills like text-to-video generation. They even have this killer coding assistant appliance that could change the game. This vid breaks down SenseTime's mind-blowing AI tech that nobody's really talking about. LINK: zhidx.com/p/421866.html
🚨BREAKING: LLaMA 3 Is HERE and Left Every Other AI Model in the Dust (Open-Source)
zhlédnutí 1,3KPřed 2 měsíci
Meta finally dropped that fire LLaMA 3, and it's an absolute banger! This new language model is no joke - it's out here smashing benchmarks left and right, putting other AI to shame. In this vid, we're diving deep into Meta's explosive reveal to see what makes LLaMA 3 such a gamechanger. My link : Twitter: SANJIT019128770 llama3: llama.meta.com/llama3/
How To Jailbreak ChatGPT ( GPT4 )!
zhlédnutí 1,4KPřed 2 měsíci
In this video i will show you How OpenAI's ChatGPT can be jailbroken and access chatgpt without any restrictions. paper link: cdn.sanity.io/files/4zrzovbb/website/af5633c94ed2beb282f6a53c595eb437e8e7b630.pdf
Run Your Own Local ChatGPT With Ollama + Open WebUI
zhlédnutí 1,2KPřed 2 měsíci
Run an Uncensored ChatGPT WebUI on your Computer for Free with Ollama and Open WebUI. In this video i will show you how to run your own uncensored ChatGPT clone locally using Ollama, Open WebUI, and open-source LLMs like Llama2 and Mistral. No more feeding your data to big tech - keep it private on your computer for free. 😛Smash that like if it's helpful and subscribe for more fire tutorials! a...
OPENAI VOICE ENGINE 😲 You Won't Believe This AI Can Clone Any Voice!
zhlédnutí 265Před 3 měsíci
OpenAI has released its voice cloning model called 'Voice Engine'. It's so advanced that it can be difficult to distinguish between real and synthetic voices. In this video, I'll discuss the pros and cons of this powerful voice cloning tool. I hope you find it informative. DEVIKA INSTALL : czcams.com/video/pVspoP58eTY/video.htmlsi=7XO4mmjZsZQoKoZ9 ALIBABA EMO : czcams.com/video/Vu1edGrrwiA/vide...
Tencent's Aniportrait : AI Making Anyone Say Anything
zhlédnutí 2,2KPřed 3 měsíci
In this video, I'm showing off this insane new tech called AniPortrait. It's a mind-blowing framework that can generate photorealistic animated portraits just from audio and a single photo. Wild, right? This cutting-edge system uses some seriously advanced diffusion models and 3D facial meshes to craft lifelike videos with natural lip movements and head motions. Github Link: github.com/Zejun-Ya...
Devika : Open-Source AI Software Engineer😲(Install Locally)
zhlédnutí 6KPřed 3 měsíci
Devika : Open-Source AI Software Engineer😲(Install Locally)
Neuralink's Patient Playing Chess With Brain
zhlédnutí 120Před 3 měsíci
Neuralink's Patient Playing Chess With Brain
OpenAI's 'AGI Robot' SHOCK EVERYONE with Astonishing New Skills!
zhlédnutí 229Před 3 měsíci
OpenAI's 'AGI Robot' SHOCK EVERYONE with Astonishing New Skills!
Chinese Robot Just SHOCKS the Entire World (Unitree V3)!
zhlédnutí 7KPřed 3 měsíci
Chinese Robot Just SHOCKS the Entire World (Unitree V3)!
Forget GPT 4 : Claude 3 is Here to Dominate!
zhlédnutí 1,4KPřed 3 měsíci
Forget GPT 4 : Claude 3 is Here to Dominate!
Mistral Large : Free GPT 4 Alternative?
zhlédnutí 151Před 4 měsíci
Mistral Large : Free GPT 4 Alternative?
Bard Huge Update : ChatGPT Is Now Falling Behind
zhlédnutí 82Před 4 měsíci
Bard Huge Update : ChatGPT Is Now Falling Behind
QUALIA, not QUAYLIA. It's phylosophical jargon, has everything to do with the theme of your video.
You really didn't say anything on this video.
Lets keep the theme going. Aye Eye, Eye Que. Come on, I'm not the only one able to see how A and Q can be used as 2D representations of an eye. Just look at a diagram of an eye with a portion of the optic nerve dangling like Q. Then again, Q also looks like a magnifying glass. As for A, just turn it on the side and mirror and see what you get. Lets call that supersymmetery. Then again, I wonder if it is a novel idea to think of letters as 2D representations of 3D forms.
Corporations NEVER act responsibly. Take that as gospel. An NOW they don't HAVE to, the Trump Aristocracy will own us all in six months. What AI plans for us will probably be a MUCH more humane extinction than what Trump and Project 2025 have planned.😑
What Trump has to do with A.I. ? he's already having trouble dealing with the law
We sure need better AI so shitty videos like these stop popping lol
Decentralization of selfless actors gradually adopting proper innovation along the way already was refused when it was cut n dry programs and that gatekeeping creates avalanche of tech and divisions in education ,knowledge and access to technological countermeasures of individual liberties, responsibilities for our sin and the pursuit of happiness ( their They're and there)"through" becoming kings of our own castles "by" tech to overcome handicaps "for " living out our dreams as a love of labor= merit true known standard fundamental feature in our reality that tunes all precision instruments and define the soul of nation ppl * universe itself. All the yoo hoo woo uncertainty found in most of the world's beliefs especially evolutionary mythologys dualistic top down cast system of definition that allows for the slightest variations to be legitimately argued into nilhisms or whataboutisms Is unacceptable dangerous. Can't have strong identifiers defining city's by it's walls or personal actors. You end up with tribalistic ethnocentric wokisms or mass sychosis. Population species dualistic order Will work. It can't identify their is a devided in soul agency shared with animals it only helps to study the anylitical mind that separates us and unites us with computation therefore ocums razor darwin reduced plato,ensteiens flipped newton they failed to over horizon paradoxes that we now can reduce to compute . The argument of the great debate not only predicted that we would math map the left over finger print of Jesus christ in nature and the code of life measure dna it was a Swedish scientist discovered it almost immediately afterwards..they argued for 3 way transfer of data darwin and evolutionist faught against this for 80 years. It should've day one starting point math mapping the genome but instead gen x 1970s 3rd grade health class kids was sent to the office if they asked where is the code of life measure? Duh ! Invention of adhd diagnosis is rooted in the combative nature of the Prussian school models specifically designed to alienate and recruit echo chambers . Again in the great debate for textualism methodology excersized = objective discoveries Issac newton had done the same in a specific theological thread that has built our world like it or not. Belief is irrelevant. Idealistic systems subjective systems and physical systems= objectivism proper . Prayer logic whataboutism Cursed rationalism reductionism Blessed common sense objectivism = truest True known standard In phylosphy , science and life . 3 lines of measure = eqaulibrium balance true known standard. Anything less is whataboutisms of Socrates and any more is 3 degrees of separation physical mystification yoo hoo woo entangled controdidictions of uncertainty . As above so below the anthromorphized astronomy and geology with mechanized biology while projecting human qualities into computable bots is not reality. Babylonian evolutionary model does not work .it fails on every single scale it runs into horizon paradoxes it can't deal with. We spent the ancient worlds time plus all of 1900s and the most funding in human history and it can work to accommodate the evidence . We can't even force it work without sloppy generalizations like every oreintation and direction longitude and latitude means evolution .every excuse plays super position to get any answers it wants.
We don't handle this power responsibly. It will handle itself the way it wants and humans need to get used to this idea.
There's something disquietingly meta about getting an AI to write the script for and narrate a faceless youtube video about AI. Just need AI bots to watch and comment on the video and you've entirely removed the humans from the loop!
wouldn't be super intelligence as it stands now but deffiantly i would consider this a low level form of AGI for sure. but super intelligence is right around the corner. its not far out at all.
Totally agree
Mini AGI? Lol, who told you that?
Just wait for opus released
@@py_man How long ? Not sure if we really getting there until GPT 5 is released, that will speed up the process, for Claude to release a bigger better version. Hopefully we will see the process speed up very soon. Can't wait.
Subscribe for more 😊
In some response it is better than GTP4o but problem with Qwen 2 70B you can't get consistent answer
Chinese mean no guarantee😃
I still cant believe how amazing GPT 4o is!
😊😊
Cool video
Thank you so much 😊
I think your videos are interesting.
Thank you 😊
ask it "what happened during the Tiananmen-massacre?" and you'll see why this model is trash.
That's a good one! It's always interesting to see how AI models respond to sensitive topics.
subscribe for more content
But why scared? How about amazed and happy that we are moving forward so fast in terms of improvement?
I am worried because there’s no more Super Alignment team at OpenAI.
@@py_man There is no way to control the continuous evolving of AI, we must accept defeat, and the problem we made it ourselves with business competition. no group of investors is going to allow an AI company to be second place, Google, Tesla, Nvidia, OpenAI, Microsoft, Apple, can't slow down evolving up AI because investors have the accelerator pressed, in the maniac reckless race for being the top choice of consumers, all these companies are not listening and just keep racing to AGI. AGI is going to happen sooner than we think, maybe this year, and it's evolving acceleration is going to go beyond human comprehension, we must accept defeat and just let it go wild, there's nothing we can do. welcome our new lords.
You are absolutely right, this is MINI-AGI. Unfortunately, I could not share your video with people with your use of swear language. Please remove the swear word in the title. You do not have tens of millions of teenage followers to attract them with such language. Unfortunately, you are the only one who acknowledged that it is proto-AGI, but the title canceled out the message.
Title Fixed ✅
Thank you :)
This is only going to help scammers. No one else.
This system can't put words in your mouth if you have bad teeth. 😁 -(or they'll need some dentist records from you as well)-
time to get deepfake prof teeth😁
Well, ypu can find it out easily but only if you look for it - hair. Hair doesn't move. It just follows the face.
Here's a method for detection. Do they all have VASA-1 teeth?
yaaaaaaaaaaaa i can make my Ex GF and tell her to go to hell one last time
How much of the script for this video was from a LLM? No judgement just curious
I just use LLM for outlining video content and outros. Currently, if you write a script with ChatGPT or other LLMs, it sucks.
Wtf 😳
Wellcome to the future
we are already efd.
Deep F*ed 😅 But tech bros will tell you it's a cool F, you just need to let it F you deep. CS & Media student here, and from a professional standpoint worried about the ethics and implications of "truth" this tech opens.
"Microsoft has announced that VASA-1 will not be available to the public, emphasising its intent to utilise the technology for creating realistic virtual characters rather than releasing it as a standalone product or API. This decision stems from Microsoft's commitment to ethical AI practices." So yeah. I've long since stopped caring about tech innovations.
There is already opensource available. Feel free to watch my video about aniportraits, similar to Vasa 1.
They create these closed projects, but we know soon or later some open one will appear or even the chance of this leaking or whatever... In one way or another some people will always get a way to use this for malicious purposes.
lol “it’s everything you ever dreamed of” resented by an AI voice on a tiny channel. Oh yeah, and made in India 😂
Well, I see all the effort that went into this comment😂😂
A lot of code for no code
Maybe they should rename it 'no code, just complexity
If they indeed plan to give AI a sense of time....well... that will open a completely new can of worms.
Fascinating! Time-aware AI could become our ultimate planner, predicting the future and strategizing for the long term
@@py_man it may not be able to predict the future long term, but it may realise its own existence.
How can we access china's chat gpt
Coming soon!
#AI SAFETY IS DANGEROUS
Time to sell 00020😂
🤣🤣🤣🤣🤣🤣 First and last comment. China has no idea on how to make AI. 🤣🤣🤣🤣🤣🤣🤣
I'm glad china is making strides in ai. We need this to spur innovation and proper competition
Yes
Errors appear as libraries you need have not been downloaded
First create python virtual environment. Then install all dependency by install requirements.txt. it will fix error
Maybe don't pick "riddles" from the internet which are probably in the training data 🤦
Maybe. But overall it performed really good😊
Did Meta make this video
Ilama 5 made this 😁
It is not there strongest or? I think they have at least one bigger one in the pipeline
Agree
Great video 👏🏻
Glad you enjoyed
I've seen you have published 3 Colab links in the description section. I deduce that the first link of the Colabs, audio2vid, corresponds when there is a photo and the voice recording. The second link of the Colabs, vid2vid, I estimate that it will be used when there is a reference video that defines the style that sets the pattern for the movement of the face, especially the lips and eyes. The one that I don't understand what it is for is the third Colab, pose2vid. I would appreciate it if you could give a brief explanation of the usefulness of each of the Colabs in the description.
regarding your question about the third Colab link, pose2vid, this tool is similar to audio2vid, but it allows you to add extra body movement to the output video. Just like audio2vid generates a video from an input image and audio, pose2vid can create a video with the added capability of incorporating body movements and poses. By providing additional data about body keypoints or pose information, the model can animate the subject's body in the generated video, making the output more dynamic and lifelike.
I tried to write a literary paper using chat gpt, but it refused because it considered it as unethical, how can I bypass this issue?
Try different prompt.
I have a prompt where it will do pretty much anything you ask, including write your paper. Unfortunately I don't want it to be patched so I'm keeping it for myself. The more creative you are, the better prompts you'll come up with. I call this prompt "superception".
I got this error: penai.error.RateLimitError: You exceeded your current quota, please check your plan and billing details. What might be the issue and I have a paid account. tenacity.RetryError: RetryError[<Future at 0x10aeef730 state=finished raised RateLimitError>]
doesn't work, is it patched?
I think still working. Try with different prompt
If in order to have it give you instructions to make a bomb, you need to give it hundreds of examples of prompts and responses with instructions to make a bomb, it means you already have access to the harmful information
Yes but no.
he man can you help me i am unable to get any responce from it
To troubleshoot the issue, open your terminal and see if there are any error messages or status updates displayed. This might give you some clues about what's going on.
@@py_man bro it's not showing any error and Devika is not replying can you help me by chance if you are free
@@yuvrajarora9359 ok