Uniswap fee switch proposal is killing the decentralized DAO model.
Uniswap foundation activities move to Uniswap Labs, meaning...
...decision power moves from a non-profit organization governed by $UNI holders to a Delaware centralized corporation.
- Most Foundation employees move to Uniswap Labs
- The Foundation only keeps a tiny grants team
- After the remaining ~$100M grants are deployed, the Foundation shuts down
Thus $UNI token is no longer a DAO token but a token purely valued by buybacks/fees Uniswap will be able to generate.
It's not a criticism but admitting the facts that:
- The DAO model was indeed just pretending decentralization due to regulatory struggles
- DAOs are inefficient at governing and allocating resources
----
Uniswap isn't the first to do it either:
- Scroll fully shuts down the DAO and moved to centralized governance
- Arbitrum's "Vision for the Future" moves many decisions to the core group of Arbitrum Foundation and Offchain Labs to 'fix inefficiencies'
- Optimism Season 8 centralizes power by moving real decisions to curated stakeholder groups and councils while tokenholders only keep veto rights
- Lido’s BORG model centralizes execution into legal foundations run by appointed directors while the DAO only sets high level direction
-----
The famous a16z "Progressive Decentralization" model of finding PMF and exiting to the community for sufficient decentralization is dying.
Or it was just simply pretending in the first place.
The math behind liquidity intelligence: bounded updates, exponential convergence, MEV-proof by design. Here shows you how adaptive markets actually work. From whitepaper to mainnet. 📐🔥 @EndlessProtocol@sliswap https://t.co/Kpmsdpskmu
What if liquidity pools could learn? Sliswap gives AMMs a learning rate — adaptive curves that evolve with every trade, slashing slippage and starving MEV bots. DeFi just got intelligent. 🧠 @EndlessProtocol@sliswap https://t.co/gGZoMvd35I
🎉 Big Milestone!
Luffa has officially onboarded 1 MILLION users!
We’re just getting started — thank you to our amazing community for joining us reach this incredible milestone.
#Luffa#CreatorEconomy#Milestone#SocialFi
New blog post: Multimodality and Large Multimodal Models (LMMs)
Being able to work with data of different modalities -- e.g. text, images, videos, audio, etc. -- is essential for AI to operate in the real world.
This post covers multimodal systems in general, including Large Multimodal Models. It consists of 3 parts.
* Part 1 covers the context for multimodality.
* Part 2 discusses how to train a multimodal system, using the architectures of CLIP and Flamingo, and examples from GPT-4V.
* Part 3 discusses some active research areas for LMMs, including generating multimodal outputs.
As always, feedback is appreciated!
https://t.co/5KUi1cNMqP
If you are using jupyter notebooks for Python and Data Science, try these 7 magic commands that will save you a ton of time:🧵
1. Jupyter AI: Select any model and chat with it right from the Jupyter Notebook.
Get up at 6:30 am to participate in the discussion on the impact of artificial intelligence on society. It is indeed a bit cruel, but it is very, very interesting.
https://t.co/1URxvwvb0x