THE DAILY BRIEFING

Bittensor Daily — October 29, 2026

With no protocol release and #General back to routine scam warnings and a redirected builder pitch, the day's substance ran almost entirely through #Rao, where Monday's emission-decay proposal mutated into a long accounting fight over Lium's numbers and drew the thread's first substantive reply -- that the dTAO market already does the job the proposal wanted. SN8 (Vanta) used the same channel to disclose October buyback figures and confirm it will adopt Null Consensus, becoming the first named subnet in the tracked material to do so. Subnet cards were mixed and mechanical rather than newsy: a price-model preview, a weight-delay outage, and a corrected compute-multiplier error.

In #Rao, members dispute whether Lium's reported $40m run rate is revenue or pass-through rental volume, with miner-coverage claims flatly contradicting each other.
>40m run rate now 🙂 prompted Is that $40m revenue or just what renters pay, 95% going to miners? and the rejoinder You are indeed thinking of revenue wrong. Airbnb doesn't count rent paid to homeowners as revenue. Check GAAP. The dispute sharpened over payouts: lium miner coverage is less than 10% while chutes, targon, vanta are in the 30% to 50% range against that's not true. miner coverage on lium is higher than any other subnet, before a take rate was finally disclosed: our take rate is 5%, it's all public on our docs. (1) This is the most detailed public accounting dispute about any subnet's revenue to surface in #Rao this week, growing out of Oct 27's emission-decay proposal that had used Lium only as an illustration. (2) The two miner-coverage figures directly contradict each other and neither is independently verified in the material. (3) The thread closes with a disclosed take rate but no agreed definition of revenue, and no official or core-dev voice joined it.
A #Rao member answers Oct 27's emission-decay proposal by arguing the dTAO market already prices mature subnets' returned value, crediting a prior mechanism change.
The dTAO market does this, and it's doing it a lot better now than it did in the first half of the year, adding The shift from net flow to price-weighted has very clearly improved market efficiency and concluding Now that there's a much smaller incentive to chase flows rather than diligence fundamentals, the market can determine whether the businesses are profitable without subsidies. (1) This is the first substantive reply the emission-decay proposal has drawn since it went unanswered on Oct 27, though it argues the idea is unnecessary rather than engaging with the mechanism itself. (2) The reply introduces Burnweight as an alternative the speaker says they have deprioritized now that price-weighted flow dynamics are doing the work. (3) No official or core-dev voice is identified in the exchange, leaving it as one community view against another's unanswered proposal.
SN8 (Vanta) discloses over $130k in October buybacks and confirms it will adopt Null Consensus to open mining slots.
Vanta Trading performs buybacks as part of this, so far in Oct its over $130k. Should have another today and also rules on SN8 to mine will be changing quite a bit here soon + we'll be rolling out Null consensus to open up more slots, with the follow-up clarifying only changing the rules to be for standard accounts + opening slots. (1) This is the first instance of a named subnet confirming adoption of Null Consensus, following Oct 21's ecosystem debate over single-validator weight-setting that had only named SN9 and SN93 as hypothetical candidates. (2) A member's challenge that opening slots means little without miner competition for them received only a brief clarification. (3) The buyback figure stands as a plainly disclosed data point in contrast to the day's larger dispute over Lium's undisclosed revenue.
In #General, a prospective builder pitches a prompt-injection testing tool for LLM-judged subnets and is redirected to a developer-focused server without technical feedback.
The pitch described the target problem as subnets that score miners with an LLM judge (like ORO SN15 and Compelle SN82) put the miner's answer straight into the judge's prompt. That makes every answer a possible prompt injection. A moderator responded only that this would probably be a better post for the Church of Rao Discord server, which is more developer/subnet-owner centric. (1) The pitch names two subnets by their judging mechanism, but the channel's reply addresses only where to post, not whether the risk is real. (2) This is the kind of builder outreach #General structurally routes away rather than discusses. (3) The proposer left without any substantive feedback on the tool or its premise.
Other: SN24 (ms.computer) previews a new product as its price rises 18.01%, while SN118 (Ditto) loses 7.57% after a 12-hour validator weight delay and SN64 (Chutes) corrects an overpriced compute-multiplier bug.
(1) SN24 (ms.computer) disabled commit-reveal weights and teased an unnamed product, with its token up 18.01% over 24 hours. (2) SN118 (Ditto) activated a new benchmark version but then saw validators fail to update weight vectors for roughly 12 hours, with its token down 7.57%. (3) SN64 (Chutes) briefly pushed private-chute emissions to $11 per hour before correcting the setting, while confirming a separate Pro 6000 price increase was intentional; no major reaction followed any of the three.

#General

2 REPORTS

Notable

A prospective builder pitches a prompt-injection testing tool for LLM-judged subnets and is redirected to a developer-focused server

Opening · 13:46 UTC

A newcomer introduced an idea called JudgeJack, describing the target problem as subnets where subnets that score miners with an LLM judge (like ORO SN15 and Compelle SN82) put the miner's answer straight into the judge's prompt. That makes every answer a possible prompt injection.

Ask · 13:46 UTC

The proposer asked the channel directly: 3. Honest feedback: would you pay to have this run before a validator upgrade?

Redirect · 14:43 UTC

A moderator did not engage with the technical pitch and instead pointed the asker elsewhere, saying this would probably be a better post for the Church of Rao Discord server, which is more developer/subnet-owner centric.

3 things to read from it

The pitch names two subnets by their judging mechanism but the channel's response addresses only where to post, not whether prompt injection against LLM judges is a real risk any subnet owner present has encountered.
This is the kind of builder outreach #General structurally routes away rather than discusses, leaving any assessment of the idea's merit to happen off-channel if at all.
No subnet team or protocol voice responded in #General itself, consistent with the channel's general-audience rather than developer-facing character.

Where it stands

Unresolved in the channel: the proposer was redirected to a different server before receiving any substantive feedback on the tool or its premise.

Other

Other: standing scam warnings, a private-profile reminder, and a newcomer's welcome

Moderation · 03:44–12:23 UTC

A moderator reissued the server's profile-privacy rule, writing Please make your profile public and keep it that way. Private profiles are not allowed in this server, and a member separately recounted I've already gotten 6 private message invites from scammers. It's crazy., which drew a reminder to Please block and report all suspicious DMs and unsolicited friend requests, and never share your wallet credentials or click on suspicious links.

Welcome · 14:47 UTC

A new member shared that yesterday I bought my first TAO , happy to be here, receiving the channel's standard welcome and scam-awareness pointers.

What to read from it

Volume fell back to 10 messages from yesterday's 14, continuing the channel's post-Oct 26 decline, with the day's traffic again dominated by moderation and routine welcomes rather than ecosystem substance.

Where it stands

All items closed through standard moderator response; nothing left open.

#Rao

3 REPORTS

Key

A core-dev-adjacent voice gives the first real answer to yesterday's emission-decay proposal: the dTAO market, not a new mechanism, already does this job

Opening · 18:21 UTC

Replying directly to Oct 27's proposal for an emission-decay mechanism that would tie a mature subnet's funding to value it returns, a member argued the question was already being handled: The dTAO market does this, and it's doing it a lot better now than it did in the first half of the year

Expansion · 18:22 UTC

The same member tied this to a mechanism they said they had stopped pushing for: It's one of the reasons I haven't been banging the drum on Burnweight much, the market is finally doing the work its supposed to be doing, crediting a specific change: The shift from net flow to price-weighted has very clearly improved market efficiency

Conclusion · 18:25 UTC

They drew out the implication for subsidy-dependent subnets: Now that there's a much smaller incentive to chase flows rather than diligence fundamentals, the market can determine whether the businesses are profitable without subsidies.

3 things to read from it

This is the first substantive reply the emission-decay proposal has drawn since it went unanswered on Oct 27 -- but instead of engaging with the mechanism itself, it argues the proposal is unnecessary because the dTAO market already prices that question in.
The reply introduces Burnweight into the conversation without prior mention in this material, framing it as an alternative the speaker has deprioritized now that price-weighted flow dynamics are doing the work.
No official or core-dev voice is identified as weighing in, so this remains one community member's view against another's unanswered proposal rather than a resolution.

In the week’s context

The emission-decay proposal itself was posted on Oct 27 using Lium's flywheel as an illustration, without alleging wrongdoing, and went unanswered that day; today's reply closes that gap, if only with a counter-framing rather than engagement with the original mechanism.

Where it stands

Answered for the first time, but with a claim that the existing market mechanism already solves the problem rather than a response to the decay idea itself; no further reply follows in the material.

Key

The same emission-decay thread splits into a long, unresolved dispute over what counts as Lium's actual revenue

Opening · 04:29–10:53 UTC

A member used Lium to question the ecosystem's value flows: Most of revenue of Lium comes with tao since most of customers are miners. And the tao was taken from mining. Miners are selling subnet alpha to buy lium credits. Tbh, Lium is absorbing tao from other subnets. A reply countered that Lium brings money from offchain too. Platforms dont have compute but Lium has so anyone looking for a gpu can end up on Lium

Dispute · 15:17–16:55 UTC

After a Lium team member posted >40m run rate now 🙂, another member asked Is that $40m revenue or just what renters pay, 95% going to miners? and argued Its not revenue to lium if it goes to miners, prompting yes it is and check GAAP. The critic pressed further: Most people think of revenue as what is received to offset miner payout. I think a lot of people may be confused about lium's rental spend numbers, thinking that is revenue to lium. How much are you actually offsetting? Best I can discern is under 10%, and later: You are indeed thinking of revenue wrong. Airbnb doesn't count rent paid to homeowners as revenue. Check GAAP

Reply · 17:15–17:23 UTC

The Lium team member conceded nuance: We do all the provisioning and add services to the compute, which may make it principal accounting under GAAP, different from airbnb which doesn't do anything to the houses it rents, making it agent-based accounting, adding it's ambiguous. A third member proposed a framework distinguishing It's Intercompany Revenue from It's external Revenue, arguing the latter is creating buying pressure on TAO overall

Pushback · 17:31–17:49 UTC

The dispute sharpened over miner payouts: one member said Its not remotely like that. Its a pass through on a brokered transaction. Miners are getting paid double what you would think compared to every other subnet. I don't think people realize lium miner coverage is less than 10% while chutes, targon, vanta are in the 30% to 50% range., which the Lium team member disputed outright: that's not true. miner coverage on lium is higher than any other subnet. The critic replied But the number is secret. Publish an actual revenue number please, not bookings. The top subnet should be held to a high standard., while another member dismissed the claim as Sounds like FUD to me.

Outcome · 18:28–19:07 UTC

The Lium team member eventually disclosed our take rate is 5%, it's all public on our docs, while maintaining but I still think we should consider total volume of rentals as revenue. One member summed up the stakes: This is the number that any serious investor wants to know before subnets scale to infinity. Unit economics need to make sense., and another agreed the split matters from a subnet perspective, would still want to get a view on unit economics before we scale stuff to infinity. Especially since we see more and more subnets maturing and scaling.

4 things to read from it

This is the most detailed public accounting dispute about any subnet's revenue to appear in #Rao this week, growing directly out of Oct 27's emission-decay proposal that used Lium only as an illustration.
The disagreement is substantive, not cosmetic: whether gross rental volume, a 5% take rate, or miner-coverage percentage is the right lens changes the picture of how much value Lium actually captures versus passes through.
The miner-coverage figures cited by the two sides directly contradict each other -- under 10% versus higher than any other subnet -- and neither was independently verified in the material.
No official or core-dev voice joined the thread, consistent with the pattern across the week's #Rao tokenomics debates; the dispute closes with a disclosed take rate but no agreed revenue definition.

In the week’s context

This is the same emission-decay proposal and Lium illustration first posted on Oct 27, where it went unanswered; today it draws by far its most extensive engagement, though centered on accounting definitions rather than the proposed decay mechanism itself.

Where it stands

Open and unresolved; a take rate was disclosed but the underlying dispute over what counts as revenue, and the conflicting miner-coverage claims, remain unaddressed by any independent source.

Notable

SN8 (Vanta) discloses October buyback figures and plans to adopt Null Consensus to open mining slots

Disclosure · 14:02–14:10 UTC

Asked about buybacks, a Vanta representative said Vanta's is all external, for us leveraging usdc -> alpha rails will be a big unlock and specified Vanta Trading performs buybacks as part of this, so far in Oct its over $130k. Should have another today

Expansion · 14:15 UTC

The same representative added that also rules on SN8 to mine will be changing quite a bit here soon + we'll be rolling out Null consensus to open up more slots

Pushback · 14:21–14:27 UTC

A member questioned the plan's purpose: what's in it for opening up more slots when miners don't compete for slots?, to which the representative replied changing the rules to be for standard accounts + opening slots.

3 things to read from it

This is the first instance in the tracked material of a named subnet confirming it will adopt Null Consensus, following Oct 21's ecosystem-level debate about the trust assumptions behind single-validator weight-setting under that mechanism, which at the time only cited SN9 and SN93 as hypothetical examples.
The challenge that slot expansion is meaningless without miner competition went unresolved beyond a one-line reply about rule changes for standard accounts, leaving the practical effect of the move unclear.
The buyback figure stands as a standalone data point alongside the day's much larger dispute over how to measure Lium's revenue, offering a contrasting case of a subnet volunteering a specific dollar figure without challenge.

In the week’s context

Extends Oct 21's ecosystem debate on Null Consensus, which discussed what stops a single top-stake validator from acting dishonestly and named SN9 and SN93 as potential adopters; SN8 (Vanta) is now the first subnet in this material confirmed to be rolling it out.

Where it stands

Partially answered: the buyback figure and Null Consensus plan were disclosed, but the challenge about miners not competing for slots received only a brief clarification and no further follow-up.

Subnets

5 REPORTS

SN64 24h −0.20%

Chutes

What happened: Chutes (SN64) adjusted compute-unit multipliers across GPU tiers, including the private chute bonus and H200/Pro6000 scaling, which briefly pushed private chute emissions up to $11 per hour before the team corrected the settings. The team also confirmed that the Pro 6000 price increase from $1.80 to $2.25 was intentional, to match updated market rates. Separately, a brief internal server error caused chute listing failures, which the team fixed shortly after. Chutes also went live on SayGM (SN28), allowing its confidential open-weight models to run inside attested hardware via SayGM's API.
Community: Miners questioned whether the private bonus jump to 7.5x was intentional and pointed to a 25% share of the incentive pool going to private chutes, while others disputed the exact multiplier figures and asked for clarification. Several community members worked through the math on H200, Pro6000 and B300 multiplier changes over time, with some concluding the Pro 6000 reduction was smaller than for H200.
Official: The team acknowledged the private chute bonus was too high, said it could be lowered, and pushed a fix along with a related adjustment to bounty growth. The team also confirmed the Pro 6000 price increase was intentional to align with market rates.

In debate · SN64 (Chutes) GPU multiplier and pricing fairness · 13 participants

A miner argued the private chute bonus increase led to disproportionate emissions compared to actual billed rates.

“It resulted in the private chutes that you bill $1.8 per hour for receiving $11 per hour in emission.”
The team acknowledged the bonus was too high and said it could be reduced.

“We can lower the private chute bonus though”
Another miner noted the changes mainly affected Pro 6000 private chutes and would not materially change H200 payouts.

“there are no private chutes on h200 so it doesnt change anything effectively for h200 payouts.”

SN118 24h −7.57%

Ditto

What happened: SN118 (Ditto) activated DittoBench v14 at 01:28 UTC, moving emissions scoring from v13 while keeping v13 scores unchanged. Validators then failed to update weight vectors for about 12 hours due to a ledger signature issue and a later heartbeat gap, leaving newly scored agents and a generator unweighted until fresh weights were committed around 13:36-13:54 UTC. Separately, several agent submissions built on the publicly released dan0l base code were repeatedly held by automated copy-similarity checks and cleared after manual review. The team also fixed a screener bug that had misclassified dan0l's reply-draft mechanism, and dan0l v28's source code became public at 21:05 UTC after its 48-hour disclosure window.
Community: Miners whose agents or generators were held or unweighted asked operators to check specific UIDs and explain the delay, with one miner pressing for a timeline on the weight-setting fix. Several pointed out that their code's overlap with other agents came from the shared public dan0l base rather than copying.
Official: The team confirmed the weight delay was on its side, traced it to a ledger signature issue followed by a heartbeat gap, and posted when fresh v14 weights were back on chain. It also shipped a screener fix (PR #3242) reclassifying the dan0l reply-draft mechanism and cleared multiple held agents after manual review.

In debate · SN118 v14 weight-setting delay · 9 participants

A miner flagged that a generator was scored but missing from any validator weight vector and asked if the v14 update was stuck.

“it isn't in any validator weight vector yet. The validators seem to still use the ledger from the v14 activation”
A community member said there was no visible change or fix and pressed for an explanation and timeline.

“there is no validator weights change, no crown's code release. what happened? When will you fix it?”
The team explained the delay as validators correctly rejecting a ledger with incomplete signatures before fixing and recommitting weights.

“after v14 activated at 01:28 UTC, Platform's ledger included three agents with only two of their three validator signatures”

SN44 24h −1.64%

Score

What happened: SN44 (Score) launched a new public track element for detecting room dirtiness, defining five detection classes with annotation rules. The team blacklisted four hotkeys on the football private track for not sharing committed solutions as required by the track's documentation. A Chutes platform outage caused 500 errors and zero scores across public elements at block 9383100, affecting Petrol Station and Beverage challenges.
Community: Miners flagged ground-truth disputes on several elements, including PPE eyewear annotations and Petrol Station pump counts, and asked the team to review specific blocks. Several users pressed the team to remove the failed block after the Chutes outage. A miner also raised concerns about the Chutes hourly rate for the PRO 6000 GPU rising to $2.25.
Official: The team confirmed the Chutes outage was being fixed and said the affected block would be removed once resolved; it later removed block 9383100 for all public elements, including Petrol Station and Beverage. The team stated rewards are sufficient to cover costs and reiterated that Chutes offers stability and security.

SN15 24h −3.62%

ORO

What happened: SN15 (ORO)'s Race 6 produced unusually low scores for most agents, prompting miners to question the results. The team explained that race request text is deliberately withheld, unlike qualifying rounds where task details are visible, to prevent hardcoding and push developers toward more generalized agents. Separately, a miner reported that a submission missed the Race 7 start snapshot by 1.5 seconds despite completing its qualifying runs before the cutoff, and asked the team to manually include the version.
Community: Miners voiced frustration over the low Race 6 scores and the lack of visibility into race requests, with some asking the team directly to explain the hidden-task approach. Others defended the design, arguing that fresh, undisclosed race content each round is the best way to identify agents that generalize rather than overfit to past data.
Official: The team said race requests are withheld to prevent pattern matching and encourage versatile agent development, and said it is considering releasing sample queries in future races after observing how performance trends over the next few rounds.

SN24 24h +18.01%

ms.computer

What happened: SN24 (ms.computer) disabled the commit_reveal_weights_enabled hyperparameter, switching it from True to False. The team also shared a preview of a product called ms.computer, describing it as something they have been working on, with more details to follow.
Community: Community members asked whether the preview involved physical computing hardware, prompted by imagery in the preview video. Several messages referenced upcoming mining opportunities on the subnet.
Official: The team posted a preview of ms.computer in the subnet channel, saying more information will be shared soon.

More subnet community coverage every day in the Telegram channel →

Lium's revenue accounting dispute

Lium reported a run rate of over $40M, prompting questions about whether this figure represents revenue to Lium or gross rental volume, most of which is paid out to GPU-providing miners. The exchange drew in comparisons to GAAP principal-versus-agent accounting rules and to how marketplaces like Airbnb recognize revenue.

Positions

Some community members argue the $40M+ figure is gross rental volume, not revenue to Lium, since a large share is passed through to miners, drawing a parallel to how Airbnb does not count homeowner payouts as its own revenue.

“Airbnb doesn't count rent paid to homeowners as revenue. Check GAAP”
Others argue the full volume should count as revenue, comparing miner payouts to a company paying its employees and noting that if Lium controls the GPU provisioning rather than just arranging it, GAAP principal accounting could apply.

“That's like saying it's not revenue for a company if it goes to paying its employees. Lol”
A third position calls for separating rentals (revenue), miner payouts (OPEX), and the difference (gross profit), and for Lium to publish an actual net figure rather than headline booking volume.

“Publish an actual revenue number please, not bookings. The top subnet should be held to a high standard.”

Official response: The team acknowledged the classification is ambiguous, citing GAAP language suggesting principal accounting could apply since Lium handles provisioning and adds services to the compute, unlike a pure marketplace. The team stated its take rate is 5% and that figures are published on its docs and on grafana.lium.io, but said it still considers total rental volume a reasonable way to count revenue.
What to watch: Whether Lium publishes a net revenue or take-rate breakdown beyond the 5% figure already shared, and how it addresses the miner coverage comparison raised in the discussion.

@markjeffrey

Subtensor merged quantum-resistant wallet support (ML-DSA-65 and Falcon) with optional key rotation into the release-487 branch, now on testnet but not yet mainnet, with no migration needed for existing wallets.

View the original post on X ↗
@urnetwork

UR (SN25) launched an open competition on SN1 (Apex) inviting participants to improve its routing and provider-matching algorithms over six weekly rounds, with winning submissions adopted as the new baseline.

View the original post on X ↗
@Poloniex

Poloniex announced a new TAO listing, with deposits opening October 28 at 07:30 UTC and full trading starting October 28 at 09:30 UTC.

View the original post on X ↗
@chutes_ai

Chutes (SN64) announced it is now live on SayGM, allowing SayGM's private models to run on Chutes within secure hardware.

View the original post on X ↗
@metanova_labs

Metanova Labs, working with Chutes (SN64) and Onepot AI, is developing a Hit-to-Lead optimization agent that extends NOVA's capabilities into iterative small-molecule optimization using an algorithm from a NOVA Blueprint winner.

View the original post on X ↗
@b1m_ai

Beam announced integration with Hugging Face to help move large datasets into and out of Hugging Face via its distributed network, as Hugging Face's dataset count grew from 711,000 to 1 million in the first seven months of 2026.

View the original post on X ↗
@taostats

Taostats confirmed that services have resumed following an attack-related issue that took longer than expected to resolve, and apologized for the disruption.

View the original post on X ↗
@taostats

Taostats reported experiencing network issues that were being worked on and apologized to affected users.

View the original post on X ↗

◆2026-10-29 · End of this edition

Past editions +

24H Market

Gainers

Losers

Bittensor Daily — October 29, 2026

#General

A prospective builder pitches a prompt-injection testing tool for LLM-judged subnets and is redirected to a developer-focused server

Other: standing scam warnings, a private-profile reminder, and a newcomer's welcome

#Rao

A core-dev-adjacent voice gives the first real answer to yesterday's emission-decay proposal: the dTAO market, not a new mechanism, already does this job

The same emission-decay thread splits into a long, unresolved dispute over what counts as Lium's actual revenue

SN8 (Vanta) discloses October buyback figures and plans to adopt Null Consensus to open mining slots

Subnets

Chutes

Ditto

Score

ORO

ms.computer

Lium's revenue accounting dispute

24H Market

↗Gainers

↘Losers

Bittensor Daily — October 29, 2026

Today's highlights

Discord conversations worth following

A prospective builder pitches a prompt-injection testing tool for LLM-judged subnets and is redirected to a developer-focused server

Other: standing scam warnings, a private-profile reminder, and a newcomer's welcome

A core-dev-adjacent voice gives the first real answer to yesterday's emission-decay proposal: the dTAO market, not a new mechanism, already does this job

The same emission-decay thread splits into a long, unresolved dispute over what counts as Lium's actual revenue

SN8 (Vanta) discloses October buyback figures and plans to adopt Null Consensus to open mining slots

Chutes

Ditto

Score

ORO

ms.computer

Ecosystem debate

Lium's revenue accounting dispute

Voices on X

Recent editions

Gainers

Losers