BIG SOULJA

Hi there!

I am a software developer who creates integrations between experimental hardware and currently accessible XR Metaverse experiences.

• AI VOICES - Click To Read My Response

About me:

I'm a freelance software developer with a background in the Android mobile platform and custom modding scene for various devices. However, my passion for technology has led me to transition to VR/XR, where I have found a new world of possibilities.I love playing VR experiences and am constantly exploring new worlds and pushing the boundaries of what is possible. I find inspiration in meeting new people and getting lost in digital worlds.As a creator, my goal is to innovate and inspire, creating experiences that transport you to a new world without ever leaving the comfort of your own home. I strive to impress, and am always looking for new ways to push the limits of what can be achieved in VR technology. True immersion is my goal, and I am willing to do whatever it takes to reach it.If you're interested in VR technology or just want to chat about the latest developments in the field, feel free to get in touch. I'm always excited to connect with like-minded individuals who share my passion for this exciting and rapidly evolving field.

My setup:

PC Specs:- AMD Ryzen 9 5950X 1.3V @4.75Ghz All-Core OC
- Corsair Vengeance RGB PRO 128GB RAM @4700Mhz
- Asus X570 Prime Pro Motherboard
- Gainward Phantom NVIDIA RTX 4090 @3025Mhz OC
- Palit RTX 3060 12GB (Secondary GPU for offloading certain tasks)VR Hardware:- Meta Quest Pro w/ Tundra Tracker for continuous calibration
- PIMAX VISION 8KX DMAS (7680 X 2160 Resolution) w/ Beta 120Hz firmware
- Valve Index (2880×1600 Resolution)
- Teslasuit (DK1 Somnium Space Founders Edition)
- 7invensun Droolon Pi1 Eye Tracker Module
- HTC Face Tracker
- x5 Valve Basestation 2.0's
- x11 HTC 3.0 Trackers (Had these before getting the suit, not used much anymore and I guess are technically for sale at this point. Feel free to contact me about this.) No longer for sale, have started using them again for convenient FBT/tracking other objects.
- Moman MA6R XLR Mic + Scarlett Solo 3rd Gen Audio Interface (VR Mic)

My VR adventures:

Enjoy some random pictures of my adventures in the metaverse!

"Virtual Reality is our inevitable future, a future I'm passionate about. It is in my best interest to contribute to the development of the Metaverse. My goal is to make what seems like a distant future, not so distant."_{- BigSoulja, A.K.A The Real Ready Player One.}

---

SOCIALS

• GitHub
• YouTube
• TikTok
• VRChat
• Twitter
• Ko-fi
• Discord
• Gumroad
• Airnfts

“AI voices are harmful to the voice acting industry” - My official response.

I understand your concerns. As a creator myself in many fields, separate from generative AI tools, I understand how such technology can seem scary. I understand that this technology could be used to impersonate others maliciously, or outright replace them at the click of a button. Literally a click of a button. Well, that's what many people seem to think it is anyway.The problem with online society today is people are too quick to hop on bandwagons without doing any research as to how something works first. So let me explain.AI voice conversion! How does it work?Here is a basic rundown of how it works. Using GAN machine-learning models, features of audio are able to be extracted and categorised/mapped to a generic model. This model can then be applied to another audio source, and have the closest matching data synthesised from the final AI model to the input audio source. Using variational inference, this data can be synthesised with randomised (but reasonable) values to make the output sound more realistic and diversified.

What can and can't it do (as of writing this on the 13th September 2023)?Because it can essentially ‘swap’ out audio from one source to another, it is able to replicate someone's vocal pattern very convincingly if done correctly. Emphasis on ‘done correctly’. It is not easy to do and many AI models sound terrible. Despite this, it is not able to recreate other elements of a voice/person such as:

Now that you hopefully understand how this technology works, I will explain why I don’t believe it is a threat to the VA industry.Firstly, as I have already explained to many people now, VA’s are very talented people. They put a ton of dedication into their craft, and it takes both years of raw self-learned talent along with some genuine ‘born-with’ skill to accomplish. It also takes skill to not only be able to voice act, but embody a character and essentially be behind what makes the character who they are. In many cases before the character has even been shown to the public eye, so almost everything is riding on the VA.I think it’s important to acknowledge that a voice actor is so much more than just ‘their voice’, even if the title doesn't seem to imply otherwise. It is a huge skillset that cannot be replicated by AI in its current form, it's as simple as that. AI voices require a lot of manual labour, high technical skill, and access to powerful hardware to create and infer these models. And anything I create using them, the VA I am representing could make it happen within minutes instead of hours or even days.This is not me suggesting that VA’s have it easy. I am purely saying that because of their years of hard work and dedication, they are able to do what AI can do in a much shorter amount of time. And because of this, it wouldnt make sense for a company to use AI over a real person in its current form.“If I do a job in 30 minutes it’s because I spent 10 years learning how to do that in 30 minutes. You owe me for the years, not the minutes.”…But you are literally STEALING their voice without permission! It is disrespectful and will be the end of VA’s careers if it continues!Firstly, I’m not stealing anything. These voice lines are publicly available and legal for me to use so long as (according to Hoyoverse) I declare that I am in no way associated, sponsored or approved by miHoYo/HoYoverse/HoYoLAB. All voice lines from Genshin Impact, regardless of the VA who voiced them, are the intellectual property of miHoYo/HoYoverse/HoYoLAB. I am following their guidelines, therefore I am allowed to do this. I am not redistributing the voice lines/datasets I create or using them in ways that violate their TOS.Additionally, the finished models do not contain any voice data from anyone it is impersonating. It is purely a digital diffusion model.I am not claiming to have worked with any of the voice actors either. I am not claiming to be a voice actor myself, nor any of the voice actors whose voices I have replicated. I am also making no profit from this whatsoever. ZERO. My work based on real people/characters/others likeness is not monetised or in any shape or form and never will be.Also if anything, as I have mentioned numerous times, it gives the VA’s, Genshin Impact (and other games), miHoYo/HoYoverse/HoYoLAB, and the original reference song recognition. Many of the songs I have released have brought more viewers to the original songs, brought awareness of which voice actors are behind certain characters and brought joy to people around the world. Reality is in the end, it's just music. People getting to hear what it might be like if their favourite game character sang a song they enjoy. Music these VA’s are not going to make themselves, not because they don't want to but because they don't have to - let's be real now. And I see no issue with that either, they are busy people I’m sure.I understand that it can be more dangerous than just that, but you seem to forget - that just isn't what I'm doing. This leads me onto my next point.AI is here to stay, whether you like it or not.That's right. Buckle your seatbelts. Because 2024 is going to be a crazy ride!I’m sure you have heard this many times already, but have you ever actually taken the time to think about how it might impact you? Most people hadn't. Now that people like myself are showing the world what it is capable of, everyone is quick to raise their pitchforks. I hope you know that this technology isn't new, it has existed in places like Japan for years already as early as Vocaloid. I can tell you right now, getting a small content creator like me to remove their videos isnt going to change anything.AI is going to be a core part of the future now. Our future. And trying to silence it by, for example, trying to take down Genshin AI music isn't going to do anything. I think its important that anyone affected by this actually handles it with the correct parties, rational requests and realistic expectations. You cannot just tell people to stop doing it or take down their content. It won't work. Even if I take my content down, it doesnt stop literally anyone else from doing what I do, and likely much more disrepectfully.Why not just take your content down anyway? What do you have to lose? That seems easier than being at war with an entire industry, right?What do I have to lose? Honestly, not much. I could just delete the videos and pretend nothing happened. But I am trying to emphasize that anyone affected by AI voice cloning doesn't have much to lose either. At least not now anyway. And if in the future it gets way too advanced to the point it does rival humanity, how is that my fault? Voice actors and anyone else affected need to find a way to coexist alongside this sort of technology. That is all we can do, the same way humanity has managed to coexist with every other technological advancement made so far.But you are bringing awareness to the technology by making these videos! You are profiting from someone else's likeness when it should be them getting all the credit!…And to this I say, wrong. To reiterate, I make no money from this at all. If anything I lose money from the amount of hardware I have running to train these generative models. I give full credit where I can, that goes to the original VA’s, the original song creator, and any artwork used on my page that isn't by me. I do it purely because I enjoy it as a fun hobby to do on the side. I then choose to share the outcome of my hobby because it would be a waste if I didn't right?As for me bringing awareness to this technology, that is just stupid. Ghostwriter already did that. Or maybe it was JVKE? Whatever you want to believe. There are loads of AI covers out there, most of them arguably as ‘destructive’ as mine as they replicate music artists that also rely on their voice. Yet their careers are still fine. No difference here! In fact, this is no different to a voice impressionist. I would say that is a lot more dangerous yet that has never been an issue either.And finally, when it comes to me and what I gain out of this? Not much. I get recognised by people for my talent with AI models and de-mixing/mixing music. I’d say that's pretty fair considering how much work it takes to do. Besides that, nothing else.However there ARE services out there that make this technology easier to use, more attainable. I will not name the services here, but I’m sure plenty of you have seen them before advertised on certain social media platforms and so on. These services ARE contributing to the development of this tech privately (non-open source) and they also DO monetize usage of the services too without any consent from the people they are ‘impersonating’. They allow you to convert your voice with ease from a huge library of celebrity figures and characters, and the list of these services will only continue to grow. This puts the tech in the hands of anyone with no regulation, something I am not doing myself....So how am I THE problem? Because it seems I am being soloed by numerous VA's now and their dedicated fanbases, despite being the least of their issues when you look at the grand scheme of things. Understand that I could join the team of one of these services myself and make a bunch of money. I have the complete skill set they need to make perfect voice models, set up an entire website backend for easy use on the user end and much more. All from the comfort of my own home and easily make hundreds a month from the thousands of people that use it. But I don’t. And yes, that is by choice. I have had many offers already which I wholeheartedly declined.I do have respect for people’s craft, even if it doesn't seem that way initially.Understand that I could also do anything with these voice models, a lot worse than what I am doing now. I could sell them to small game dev companies so they can have a multitude of voice actors starring in their small $1.99 Steam game, or use it to impersonate and tarnish someone's reputation and so on. But no, I’m not doing any of that. I keep everything to myself. I don’t tell people how to do it, I don’t sell any models based on someone else's likeness, and I don’t claim that anything I make is REAL. This is all that is required of me, it is that plain and simple. If you don’t like it, I am the wrong person to complain to because I am not the difference between this technology existing or not.…So to conclude:I am not the bad person here. I am not the reason this tech exists. I am not actively developing it. I am not the difference between a VA getting paid or not getting paid. If you support what I do, I appreciate you! And if you don’t, that is fine too. But to say that I am single handedly killing the VA industry is a long shot and I don’t appreciate that.I hope this all makes sense, and if you made it this far, thank you for reading.- Big Soulja

• Back To Homepage