There was a quiet shift in @PythNetwork last week that most people haven’t fully priced in yet
On April 22, OIS emissions stopped
~1.93M PYTH per week gone. No drama, just governance doing its job
At the same time, nothing changed on the other side:
the protocol is still buying PYTH from the open market every month using real revenue
That’s the part people are missing
For the first time, the flow flipped,
more PYTH is being bought than distributed
Before, the system constantly leaked supply
Now that leak is gone, but demand didn’t go anywhere
And it’s not static either
DAO revenue has been ramping up for months.
Pyth Pro alone has sent close to $500K since September, and new streams are still scaling
So you end up with a simple dynamic,
less supply coming in
steady (and growing) buy pressure
No big announcement. No hype cycle
Just the math quietly changing underneath
This is what protocol maturity actually looks like
most people still don't understand what an oracle actually does
Every time you trade on a DEX, borrow against your crypto, or get liquidated — something had to tell the blockchain what the price was. The chain itself is blind. It has no idea what BTC is worth, what EUR/USD is trading at, or how to value an ETF
That data has to come from somewhere
For crypto assets, oracles solved this years ago. But traditional financial data spot FX, precious metals, fixed income, ETF valuations, that world ran on a completely different system
Institutions generated the data. Vendors like Bloomberg repackaged it. Distribution was gated behind expensive terminals. You paid for access. Middlemen took their cut at every step
That model just started breaking
Six of the world's largest financial institutions Fidelity, Euronext, Tradeweb, SGX FX, OTC Markets Group, and Exchange Data International - just chose @PythNetwork as their onchain data distribution layer
First time any of these firms have published proprietary data via onchain infrastructure
What they're putting onchain: > Spot FX benchmarks > Precious metals pricing > ETF intraday valuations (iNAVs) > Fixed income & crude swaps > OTC pricing & reference datasets
The hardest-to-source corners of a $50B industry. Data that used to require a Bloomberg terminal and a six-figure contract.
Through the Pyth Data Marketplace, they now distribute directly - across 100+ chains and 700+ apps, while keeping full control over attribution and pricing.
That's a new revenue stream for the Pyth DAO - on top of Pyth Pro already running at $122K/month, growing 36% MoM
Three continents. Six institutions. One network.
This isn't TradFi slowly "entering crypto." It's TradFi choosing onchain infrastructure as the new standard for distribution
The shift is already happening. Most people will notice years too late
Hot take: most prediction markets have been pricing assets with garbage data
Polymarket just fixed that. They brought in @PythNetwork - the oracle running 60% of perps protocols,
backed by Jane Street, Cboe, Jump.
Gold. Oil. TSLA. NVDA. All live, all institutional, all verifiable
When millions of dollars hinge on one price - you don't cut corners on the source
BREAKING: @Polymarket just went live with traditional asset markets, powered by Pyth Pro 🔮
Major equity index ETFs, commodities, single-name US equities up/down + daily close contracts. All relying on Pyth data.
Here's how it works and why it matters 🧵
how to claim:
sign up here: https://t.co/ohZ2zxiY3j
use the builder for free during the event
claim your $100 api credits on the site
full event details: https://t.co/9GoqEroh5H
RT + Like for the alpha. let's build 🤝
FREE $100 FOR YOUR CLAUDE CODE API
tomorrow, march 8th, @Lovable is dropping the ultimate alpha for vibecoders:
• 24 hours of UNLIMITED access to the platform @Lovable
• $100 ANTHROPIC API CREDITS (@claudeai code) for your balance
• $250 @stripe credits if you’re building a product
this is a massive gift for anyone into the ai space
don't miss out.
watching a few videos on the obsidian + claude code workflow and it looks pretty interesting. having a direct link between your digital brain and an llm is a vibe
going to try and implement this for testing
tested @NotebookLM today and i can say for sure: if you're just starting with ai, this is the first tool you should check out.
it’s a real game-changer in the ai space. i kept putting it off, but it turns out i had a goldmine right under my nose.
detailed post later
for example, if you want to dive into "how to use claude code @claudeai for beginners." just upload your links or files, or ask the tool to find the right sources. confirm the selection, refine your prompt, and you're good. you get a clear, concise summary and you're done studying with @NotebookLM
no need to spend hours watching videos (if visuals don't matter) or reading through a stack of websites
or let's say you're choosing a country for travel or living. just pick a few videos like "pros and cons of living in ukraine" - you can even bulk upload an entire channel if it's relevant. you get a full list of opinions in seconds, instead of watching every blogger for 15 minutes each
the simplest way to use @NotebookLM
my goal - as i mentioned before, i’ve been trying to find a specific video by margulan youtube with a great idea i heard a few months ago but didn't quite catch
so i just bulk upload his last 50 videos with one click and drop a totally newbie prompt describing what i'm looking for - just a stream of consciousness.
tested @NotebookLM today and i can say for sure: if you're just starting with ai, this is the first tool you should check out.
it’s a real game-changer in the ai space. i kept putting it off, but it turns out i had a goldmine right under my nose.
detailed post later
you can upload your own files, info sources, videos, and articles this way
it lets you analyze massive amounts of data instantly, and here’s the best part -
the ai doesn’t hallucinate because it stays strictly within the sources you provide, giving you clear citations for every thought or idea
decided to get back into obsidian. going to structure all my thoughts and notes here
watched a bunch of videos on how people make it look clean and efficient, so i’ll start implementing it step by step. obsidian is basically god-mode for notes
first thing - moved my ideas folder and started filling it with ai-related stuff.
later on, i might code some scripts to sync obsidian with my other services