Maciej Kula

See you in Montréal everyone. Not gonna want to miss this one🫡

Right, let's make an "better" case scenario for the malicious actor. The owner has 5K TAO worth in ALPHA locked (for quite some time so their conviction decays slower). The malicious actor would need to enough to have more ALPHA than the owner (exposing themselves to price action on the subnet), then they need to lock their ALPHA for a period of time (depending on the EMA) let's take 2-3 months, so that their EMA grows from 0 towards the amount of locked ALPHA (during which they could not sell). Assuming that the malicious actor would be able to do it still. The previous owner would remove their stake, likely followed by other actors/ALPHA holders through concerns of the ownership transfer. The actor would now be the subnet owner with locked ALPHA of a subnet that depreciated in value. Once their ALPHA tokens are unlocked again they may not have the same value as previously (so a loss of investment). Instead the actor could have bought a new slot for just 1K TAO, immediately becoming the owner and building their subnet there. So that being said (I am sorry if I repeated some parts), what exactly is the incentive for the malicious actor to have done so? NOTE: In some cases like on Chutes the amount of ALPHA to own more than the owner could lock would be approximately 40K TAO (which they would need to acquire and lock into the subnet)

@mcjkula @learnbittensor @Rapido_ai Thanks for your response. The incentive is that if you become a subnet owner wouldn’t they now get the 18% of emissions that goes to them?

🔒Locked Stake & Conviction are Being Proposed for Bittensor Subnet ownership may be about to change. BIT-0011 proposes that anyone can challenge for ownership of a subnet by locking their ALPHA stake and building "conviction." Right now, inactive subnet owners can hold onto subnets indefinitely. 'Locked Stake' would make ownership a contest of commitment, giving motivated participants a path to take over neglected subnets. How it works: - Lock ALPHA on a subnet for a chosen duration to build conviction - Conviction starts at the full locked amount and decays linearly to 0 at expiry - Every 30 days, the staker with the highest conviction EMA becomes the subnet owner - The EMA smoothing prevents anyone from briefly locking a large amount to "snipe" ownership Learn More: learnbittensor.org/concepts/locke… NOTE: This is a draft proposal (BIT-0011) currently being discussed. The implementation and design may change or be replaced completely.

RE: What would incentivize the whale to become the subnet owner? The whale would need to commit their stake for at least a month if not more, so that their conviction EMA can converge with the locked ALPHA amount (of course need to own enough ALPHA in the first place). By doing that they would potentially push the owner out to just leave/unstake and so the other ALPHA holders. Owning locked ALPHA that potentially deeply depreciated in value (resulting in a loss of investment)

@learnbittensor @Rapido_ai What’s to stop a whale to become the subnet owner of a promising subnet? Wouldn’t that disincentivize the actual team that is working night and day to make it work? What do you think $TAO?

@orchidguy5 @bittingthembits That would not allow someone willing to commit (show/build long-term conviction) to become subnet owner of an abandoned subnet (like SN3, SN39 and SN81) in this case

@bittingthembits It seems a lot cleaner to just lock all subnet owner alpha (only for the owner not stakers) and require 15 day notice for the quantity they want to unlock and sell

Founder rugs have been destroying communities since the beginning of this industry 🚨 We now have BIT-0011 Proposal 3 days ago, a $TAO subnet founder dumped 37,000 $TAO worth of Alpha on his own community and walked away 3 days later, the protocol already has a proposed solution on the table. BIT-0011. Locked Stake and Conviction Let me break down what this actually means because it is one of the most important proposals in Bittensor's history, and we need to pay attention to it. Right now, a subnet owner registers once, pays the registration cost, and holds that subnet indefinitely. There is nothing on-chain requiring them to prove ongoing commitment. They can sit on the subnet doing nothing. Or worse, they can build trust, attract stakers, collect emissions, and then dump everything overnight with zero warning. That is exactly what happened with Covenant. Locked Stake fixes this at the protocol level. With math. Here is how it works. You lock your Alpha on a subnet for a chosen duration. That lock creates a conviction score. Conviction starts at 100% of the locked amount and decays linearly to zero as the lock approaches expiry. The longer you lock and the more you lock, the higher your conviction. Every 30 days, the protocol checks all conviction scores on every subnet. The staker with the highest conviction, EMA smoothed over time to prevent anyone from briefly locking a massive amount to snipe ownership becomes the subnet owner. Read that again 👀 Subnet ownership becomes a continuous contest of commitment. It's not a one-time purchase. It's an ongoing proof that you have skin in the game. What this means in practice. If a subnet owner wants to exit, their conviction starts decaying the moment they stop re-locking. Investors can see it happening in advance. The market can reprice the Alpha before the owner dumps. And any motivated team with higher conviction can challenge that ownership and take over the subnet. Learn Bittensor pointed out that Chutes currently holds roughly 37,000 $TAO worth of SN64 Alpha in a smart contract for owner emissions. If the Chutes team locked all of it for a long duration, any malicious actor trying to take over would need to exceed that amount and lock it for longer. The cost of attacking becomes massive. And if the existing community does not support the challenger, they can unstake and crash the Alpha value, making the takeover economically pointless. The conviction decay mechanic: The score starts at the full locked amount, then decays linearly toward zero as time passes. So if someone locks: 1,000 ALPHA for 100 days ▫️Day 1 = 1,000 conviction ▫️Day 50 = 500 conviction ▫️Day 100 = 0 conviction The important part: They cannot freely withdraw the locked portion while conviction exists. Only the unlocked portion is available to unstake. At expiry conviction hits zero, and everything unlocks. This means owners can not silently prepare a rug. If the subnet owner wants to keep ownership strength, they need to re-lock before conviction decays too far. Const said it himself: "Exploits are what teach a system its weak spots. The quicker you find them, the faster you learn." No other network in crypto has proposed anything like this 🔥 The standard response has always been that's just crypto or hire better lawyers. Bittensor's response is to build a cryptographic mechanism Locked ALPHA turns into a time-weighted commitment score. The more size + the more time remaining, the more ownership power you have. This is still a draft proposal. BIT-0011. Being discussed openly. The design may change. The network identified its weakness in real time, the founder proposed the fix within hours, and the community is now iterating on the solution in public That is what an adaptive stress-driven growth system looks like. Now, founder alignment market's can literally measure block by block $TAO Thanks to @learnbittensor Link: learnbittensor.org/concepts/token…

For anyone that wants to get an idea of what the discussion will be about on this week with Const, this one is worth to read and form opinions/prepare questions about 🫡

@oscarmigula1 Sorry not liquidity to emission ratio, but the injection to emission ratio (which could result in the same thing, but depends on the subnet's trading which is why it's different)

@oscarmigula1 Yeah and with reaching the ALPHA halving earlier and actually outpace the TAO halving the liquidity to emission ratio on these subnets will be 1:1 while it continues to increase for newer subnets registering later in time

@RadoTsc Normalized to 0-100 x.com/mcjkula/status…

@mcjkula it's based on what the number? how does it determine 20-40-60?

One thing I’ve been working on lately was a DTAO Fear & Greed Index. Since the launch of DTAO, the index shows a clear swing from fear in March to greed by mid-April, with a recent move back toward neutral. Chart below. Do we want to add this to TAO.app @JosephJacks_?

Block #4,920,351 to Block #7,537,718 1 year. No going back.

@CryptoZPunisher @theTeacher_PP @Barbarian7676 What you said is correct. If as a holder one holds 10% of the total staked ALPHA in the subnet then at deregistration one would get 10% of the TAO in the pool.

If you stake on a subnet that is at risk of being deregistered, you will receive TAO instead of your Alpha tokens. The amount you get back is proportional to the TAO available in the pool at the time of deregistration. However, if the subnet is transferred to a new owner, you keep your Alpha tokens, there is no automatic conversion in that case. I’ll still tag @mcjkula just to make sure I’m not saying anything stupid and to confirm the details. That said, I don’t really understand the point of your last question?

Bittensor >> $TAO >> $dTAO Generation Z will always surprise me. @Barbarian7676 is back after the launch of the auctions on SN123 Mantis. We’re about to discover the next chapter of his work and the true launch of SN125, 8Ball. Stay tuned.

7 months. 55,000+ conversations. 128 subnets. One Savant. Here's what we learned from building the first MCP for @Bittensor.

This data seems to be wrong from my perspective. It doesn't align with the values on the chain which don't show this drop. On December 7 (Block 7047726): 1,668,662 TAO On December 9 (Block 7062126): 1,666,063 TAO Today (Block 7126966): 1,740,669 TAO Where as the chart (today) still deviates by ~200K The amounts of liquidity provided by LPs was also to my knowledge minimal and definitely not ~300K TAO

An eagle-eyed @taostats user noticed an interesting chart. A 300k tao drop in tao_in (the yellow line). 👀😱 The reason: The chain dissolved all liquidity positions on 12/8. So 300k tao in LPs were pulled out of the pools and returned to the hodlers. taostats.io/analytics/tao-…

@markjeffrey @SiliconJose @admiralagrippa @ExploitSummit Did I hear "Bittensor is too complicated"? Consider me summoned

Hash Rate - Ep 151: Bittensor EXPLOIT Summit 🧙‍♂️Guests: @siliconjose @admiralagrippa of @ExploitSummit Tickets: exploitsummit.com March 30-31 | San Francisco

@Aerials_its @DreadBong0 The staking rewards are unchanged

The first ever #Bittensor halvening is complete Block rewards have been cut to 0.5 $TAO per block History. $TAO

The TAO halving happened at #7,103,976

Happy Bittensor Halving everyone! 🥳

@DrocksAlex2 @xavi3rlu @taohalving Only the ALPHA injected into the pool, not the ALPHA emitted to participants

@xavi3rlu @taohalving Are you sure? Latest commit by Const seems to do exactly that, no?

49.99% of supply is now circulating. $TAO

@Luciano0Laki @taoapp_ @lium_io @chutes_ai @bittensor The TAO emissions aren't the raw inflow EMA, it has to be normalized to the sum of the EMAs.

@taoapp_ @lium_io @chutes_ai @bittensor Emission = Tao Flow Shouldn't all subnets for pic above receive chain buying in that case?

The chain is buying $52.5k (156 TAO) of @lium_io and $42.7k (127 TAO) of @chutes_ai per day at the moment tao.app/subnets $TAO @bittensor

@SiamKidd @lium_io @taoapp_ @chutes_ai @bittensor Because the TAO emissions > ALPHA price

@lium_io @taoapp_ @chutes_ai @bittensor Wait a minute, I must have missed a key part of the pie here. I thought alpha-in was only tweaked when price dropped below emissions? What's making the chain reduce alpha-in for Lium?

@fish_datura The ALPHA gets recycled*

With the new taoflow update, if a subnet's normalized net flow is higher than their price, the chain will start buying those subnets so that the price converges to flow. We added a new column on taomarketcap that shows the daily buy pressure of the chain. Ex: 130 extra tao per day is used to buy and burn sn51.