Bittensor Overview - Deep Dive Part 1
ฝัง
- เผยแพร่เมื่อ 4 ส.ค. 2024
- Part 1 of the Bittensor Deep Dive series. In this video we cover Bittensor and what Bittensor is as well as covering how Bittensor works and some of the incredible subnets that have been built out so far.
Timestaps
Introduction - 0:00
The Problem with AI - 0:33
What is Bittensor- 3:15
How does Bittensor Work - 6:43
Bittensor Subnets - 10:42
Is Bittensor useless? - 21:10
As always, nothing in these videos constitutes to financial advice.
Make sure you smash the like button and subscribe if you're into weekly crypto videos covering the entire Defi space! Drop a comment down below regarding your favourite topics.
How to build your crypto portfolio in 2024 - wajahatmughal.substack.com/p/...
Airdrop Referral links - use these so you can get airdrops:
Mode L2 Chain - ref.mode.network/zYpNT9
Velo Data - velodata.app/ref/owfayw4o
Grass referral link - app.getgrass.io/register/?ref...
Hyperliquid referral link - app.hyperliquid.xyz/join/0XMUGHAL
Ethena - app.ethena.fi/join/hpl9f
DeBridge - app.debridge.finance/r/7747
Follow me on Twitter - it's the place I am most active and will get back to you quickly if you have any questions or messages!
🐦 CryptoTwitter - / 0xmughal
🗞️ My Free Newsletter - wajahatmughal.substack.com/
🌿 Follow me on Lens! - www.lensfrens.xyz/wajahat.lens
📸 Instagram - / 0xmughal
🕺🏽 Tiktok - @wajmug
excellent school analogy for Bittensor! thanks
Thank you!
Im doing research on Bittensor and your video explains Bittensor very clearly. I cannot understand what Bittensor actually does when reading its whitepaper, so much mathematic formula
thank you!
What do you use to present the graphics / text? I like that program whatever it is and would like to use it as well for videos
@@ThomasRush-m8n it’s called Tome
Are you familiar with detensor which is supposed to be built on Tao...? Any thoughts? And separately also the Zero Labs project (DEAI). Would appreciate your thoughts.
Not so familiar with it, there’s so much happening on Bittensor! Zero labs, I’ll add to my deep dive list
@@WajahatMughal Great looking forward to that! If you at some point could share any thoughts on Detensor, that'd be sweet. Thank you
Yes broski
🙏
Wow, great video. I really have been waiting for something like this 😁 Question: let's assume you are interested in protein folding because you are a pharmaceutical company. You then go to that class room and then you pay the students and the teacher and the school for the results? And you pay in TAO as it is the gas token for that computation?
They're just using alphafold bro, nothing new
You create the subnet - you pay for the costs there
After that, you earn TAO and TAO gets given to the teachers (validators) & students (miners). This is via the network itself so you as a company don’t need to fork this out.
This is why so many are trying to build subnets because it takes away so much of the costs associated with running AI models
@@WajahatMughal Many thanks. Will definitely read about it.
@@shillyvanilli4114 my pleasure
Will this really get to £5k ? And when?
is staking TAO easy ? as like solana or Atom? thanks
Yup super easy. I’ve done a video on it before it’s on my channel
@@WajahatMughal thank you.i will search for find your video
@@muataz_mohammad great!
How many actual users are there? Inference costs? You didn't address Eric's points just called it FUD.
Will try to get this data on the exact number of users. I’m not sure exactly because not every subnet has a consumer app for retail to use.
What I can do is go through every single consumer subnet and find out.
Having said that, do you think any of the subnets have value?
@@WajahatMughal Thanks! So far I've just seen basic stuff like fine tuning Llama 3 models and other things that are easy to do. Who's going to train a brand new model that can compete? It's not as trivial as throwing some gpus and data at it, you actually need a team of experts, you are going against economies of scale etc. Probably the best would be to ask ai researchers about the whole thing.
@@Sawa137 agree, not every subnet is providing actual value, lots of improvements still needed.