Cyrus 🧠 𝜏

[Cyrus 🧠 𝜏 @magistrat70]

@Jus7Degen Gram

[Dear Self, @dear_self0]

LOCK IN.

[Previsões @trajaza]

Você vai voltar a ser feliz como era antes.

[Cyrus 🧠 𝜏 @magistrat70]

@alvin_investor Max price

[Cyrus 🧠 𝜏 @magistrat70]

@2xnmore Are we protected against the erosion of utility by monetary emission?

[2xnmore @2xnmore]

You have been told $TAO is “Bitcoin for AI.” That one line is the reason most holders still cannot explain what they actually own.

Bittensor co founder Jacob Steeves, known as Const, sat with Jason Calacanis on This Week in Startups for 80 minutes and walked through the machine from first principles.

Calacanis said on the show he put somewhere between half a million and $750K of his own money into $TAO this year. He also admitted he might be down on it.

Here is what Bittensor actually is, in eight parts you can reuse every time someone asks.

The Bitcoin idea it borrowed
Bitcoin never handed out tokens. You had to make them by spending compute and electricity.

Bittensor keeps that rule but changes the work. Instead of paying for a useless hash, it pays for useful AI output.

What a subnet actually is
A subnet is an open competition for one specific job, like serving AI models, renting out GPUs, storing data, or training models.

Anyone can join from anywhere. There is no application, only a small entry cost.

Who does the work and who checks it
Miners do the job. Validators run the subnet’s scoring code to check that the work is real and fast.

Each subnet has 256 slots. The best performers earn the most and the weakest get cycled out, the way a football league relegates its bottom teams.

Why cheaters do not break it
Validators reach consensus when more than half of them run the honest scoring code. On an inference subnet, miners attach a cheap cryptographic proof that the model answering you is the model they claim to run.

Const said designing rules that survive attack is the entire job of building a subnet. In his view, the website and the marketing come a distant second.

How dTAO ranks the subnets
Every subnet has its own alpha token, also capped at 21 million, paired with $TAO in a liquidity pool. To buy any alpha token, you route through $TAO.

New $TAO emissions flow to subnets based on how the market prices their tokens. Traders, investors, and builders collectively decide which subnets deserve funding.

The supply most people get wrong
$TAO has a 21 million cap, like Bitcoin. After the first halving, the network mints half a $TAO every 12 seconds.

3,600 $TAO per day.

What it costs to launch a subnet
Registration runs as a Dutch auction. The price doubles every time someone registers, then decays until the next buyer steps in.

Const said it sat around 68 $TAO when they recorded. That $TAO seeds the new subnet’s liquidity pool rather than going to the foundation.

There are 128 subnet slots today, and he wants to reach 256.

The flaw he did not hide
Open doors let in scammers too. Const said so plainly on the show.

In April, Covenant AI, the team behind three major subnets including Templar, left the network and reportedly sold around 37,000 $TAO. Their founder accused Const of holding too much control, while Const’s account is that the team sold and walked away.

The protocol’s answer is Conviction. Subnet owners lock their tokens, so any big exit becomes a public event that holders can see coming.

Where this is heading matters more than any of the above.

Const now runs Affine on subnet 120, which pays miners to produce better AI reasoning. He said Affine built a model that beat Qwen’s best in the 35 billion parameter class, until Qwen shipped a stronger one right before their launch.

That is the real target. Not cheaper inference, but competing with centralized labs on intelligence itself.

Bitcoin pays for proof of work. Bittensor pays for proof of useful work, and the market decides what useful means.

You can own $TAO without understanding any of this. You just will not be able to hold it when it hurts.

[const @const_reborn]

We will turn the web into a brain. We will decentralize. We will be permissionless. We will be free.

[Don @schwanald]

So much panic and fear mongering. You should only worry about what you can control. Live your life. Be happy. Log out often.

[Cyrus 🧠 𝜏 @magistrat70]

@alvin_investor Zec?

[Eli Ben-Sasson | Starknet.io @EliBenSasson]

it gets more inevitable that StarkWare needs to save the world

[Cyrus 🧠 𝜏 @magistrat70]

@0xAbit Tao?

[Will McEvoy @will__mcevoy]

@zkdragon on Zcash at 50K TPS:

"The Zcash community hasn't given up on the dream. In 2-3 years, let's get Zcash to 50,000 payments per second, fully post-quantum, with sub-second latency."

Visa-scale Zcash is coming.

[Cyrus 🧠 𝜏 @magistrat70]

@Ghooosssttttttt @realalvin911 Why not aave?
And Where zec

[Defi-Rocky 🐦 @Defirocky]

Buy 1 zec now !!! 🪙

#zcash #privacycoins

[Polkadot @Polkadot]

Digital cash should provide us the same liberties as physical cash does.

[Thomas (Tom) Lee (not drummer) FundstratDirect.com @fundstrat]

ETH is a store of value

Glad @VitalikButerin formalized this

[Cyrus 🧠 𝜏 @magistrat70]

@EliBenSasson @durov Ton or strk 🤔🤔🤔🤔

[Andy ττ @bittingthembits]

🚨 No one is ready for what $TAO is about to bring to the table.

Washington is coming for AI. Perhaps not with a ban, with something slower, Rules.

What's on the table for 2027:

• Data center power costs
• Mandatory independent testing of frontier models
• Liability when AI causes harm
• A proposed pause on new AI data centers

Think about what that means:

Giant data centers, and a handful of companies building AI behind closed doors and grading their own work.

Now look at Bittensor:

• No single mega data center to stop. Compute is spread across thousands of miners worldwide
• Independent testing isn't a new law, it's the whole design. Every model gets scored in public, every block
• Open weights. Public leaderboards.

Results anyone can check ✅️

They want AI that's accountable, distributed and verifiable.

It's called $TAO.

The closed labs will spend 2027 in hearings.

Bittensor will spend it shipping.

DYOR.

Sources:

🔗https://t.co/pLsC1XqZdo

[wutaner @wutaner]

I believe ETH will eventually replace BTC as crypto’s primary store of value. And Bitcoin’s future may look a lot more like DOGE than people want to admit.

Look at Dogecoin. Its market value and cultural influence have survived, but its underlying network has become increasingly marginal. Today, it has fewer than 500 publicly reachable nodes. In 2024, a single old laptop exploited an unpatched vulnerability and crashed 69% of its publicly visible nodes.

DOGE didn’t die. Its price, community and meme survived. Its mining security now largely depends on merged mining with Litecoin. That’s the fascinating part: a cryptocurrency can remain valuable even as its own network becomes less economically relevant.

Bitcoin is obviously much stronger today. But I’m looking at the next 10 or 20 years, not today.

Stablecoins, RWAs, lending, financial markets, digital art and countless other assets are building an increasingly substantial economy on Ethereum. Even BTC needs to become wrapped BTC to participate in much of that economy.

I believe the most important store of value in crypto will eventually be the native asset of the network that holds the most economic value, not necessarily the coin that came first.

That’s why I agree with Vitalik’s focus on ETH as a store of value. Ethereum doesn’t need to maximize fee burns. It needs to become indispensable.

BTC may survive as digital gold, or eventually as a giant meme. ETH has a chance to become the reserve asset of an entire onchain economy.