space

Passionate about decentralized AI

[space @concil859856]

I personally really hate what Sam did. I don't know the story between Const and Sam, but genuinely, I'm really unhappy with what Sam did, both on a human level and because of the impact it had on the community.

You can check my posts about Sam - I was one of the people who called Sam a scammer and criticized him right after he posted that fucking thing on X.

https://t.co/I5hjmqAmnE

[space @concil859856]

I absolutely cannot stand this behavior. You've been proudly claiming that you were a blockchain lead at Bittensor and that you've been building something great, you feel proud of Bittensor. BUT... what are you even talking about? Wake up. You're not nearly as impressive as you think you are.
You only became relevant because Const supported you from the very beginning, and now you've turned around and dumped everything in front of the very people who backed you. what exactly are you trying to say now???

Without Const, you would be nothing. Const is what built you up - and now you've repaid that by kicking it aside. that kind of behavior is disgraceful. you should feel ashamed

[space @concil859856]

I know many people - subnet owners, builders, and other community members - have stayed silent even though they have concerns about this.

But I think it's time to openly discuss it and give this issue the attention it deserves. Bittensor's governance should evolve to match the scale and importance of the network today.

What do you guys think?

@markjeffrey @Jason @AlgodTrading @mogmachine @fish_datura @SiamKidd @MarkCreaser @BarrySilbert @JosephJacks_ @jon_durbin @DrocksAlex2 @DreadBong0 @KeithSingery

[space @concil859856]

I'm a big fan of Bittensor and one of the people who genuinely wants to see Bittensor succeed.

But I need to point this out.

In many blockchain networks, especially PoA-style systems, privileged governance keys such as triumvirate/admin keys are distributed among multiple trusted core community members. the point is simple: no single person should have unilateral control over critical network decisions.

In Bittensor, as I understand it, @const_reborn effectively controls all three triumvirate keys and can make major decisions based on his own judgment.

Here are three keys he controls:
5E7RCRrPVS8TckCDjr92B5ciGziwz2kfvxe4URy3L7AgirGJ
5FevFjov8435t5XC2MUSRpFYxtthE8pZy1toHpgAAia3ZphG
5GRCukV2rZmSVfJhAXoLjcrU1pMVCf2Ra1ydbiCFZdaQXDXo

This system is designed as a multisig, where at least two of the three keys must approve an action.

But right now, if @const_reborn controls all three keys, then having three triumvirate keys is meaningless. in practice, one key would be enough because all three approvals ultimately come from the same person.

That defeats the purpose of multisig and creates a single point of control over decisions the protocol was supposed to protect with multiple independent signers.

This is not good governance.

Subnet emissions can be disabled without a proper public explanation, then enabled again without meaningful notice or community discussion. decisions that directly affect subnet teams, miners, validators, and investors should not depend on one person's discretion.

There is also another transparency issue that bothers me.

Most of the community does not even know who the core developers actively working on Subtensor are, what responsibilities they have, or how major technical decisions are made internally.

Maybe there are valid security or privacy reasons for keeping some contributors private. that is understandable. but if that is the case, the community should at least be told why this level of privacy is necessary.

A decentralized network should not require the community to simply trust an invisible development structure without explanation.

I deeply appreciate what Const has built and contributed to Bittensor. but being a co-founder does not mean one person should permanently control critical network decisions while the rest of the ecosystem simply reacts afterward.

The network should not depend on the approval of a single individual.

Bittensor is no longer a tiny experimental project. there are subnet teams, miners, validators, investors, developers, and community members who have committed serious time and capital to this ecosystem.

Const, you co-founded Bittensor and led it well through its early stages. but the network has grown beyond any single person.

Critical actions should increasingly be transparent, justified, and community-driven. at minimum, privileged keys and decisions should have meaningful checks, independent signers, and clear accountability.

And the development process itself should become more transparent too. the community should have a clearer understanding of who is maintaining the core protocol, how decisions are made, and why certain parts of that structure need to remain private.

If Bittensor wants to become truly decentralized infrastructure, its governance and development transparency need to mature with the size of the network.

This is not about attacking Const.

It is about making Bittensor stronger.

$TAO #Bittensor

[space @concil859856]

I agree Bittensor has been successful because Const has been leading it well, but it's also been successful because so many community members have contributed to it. I think you're missing that point.

You need to have a better understanding of how decentralized networks work. if one individual can change everything whenever they want without any notice or transparency, that's something that should be addressed. I think that's a pretty clear point.

[space @concil859856]

@Cryptnomad1 This issue should be addressed. We aren't here to follow the rules of an individual. Bittensor needs to be decentralized. I agree we can't be fully decentralized right now. we're on the road toward that, but we need at least transparency.

[space @concil859856]

I'm a big fan of Bittensor and one of the people who genuinely wants to see Bittensor succeed.

But I need to point this out.

In many blockchain networks, especially PoA-style systems, privileged governance keys such as triumvirate/admin keys are distributed among multiple trusted core community members. the point is simple: no single person should have unilateral control over critical network decisions.

In Bittensor, as I understand it, @const_reborn effectively controls all three triumvirate keys and can make major decisions based on his own judgment.

Here are three keys he controls:
5E7RCRrPVS8TckCDjr92B5ciGziwz2kfvxe4URy3L7AgirGJ
5FevFjov8435t5XC2MUSRpFYxtthE8pZy1toHpgAAia3ZphG
5GRCukV2rZmSVfJhAXoLjcrU1pMVCf2Ra1ydbiCFZdaQXDXo

This system is designed as a multisig, where at least two of the three keys must approve an action.

But right now, if @const_reborn controls all three keys, then having three triumvirate keys is meaningless. in practice, one key would be enough because all three approvals ultimately come from the same person.

That defeats the purpose of multisig and creates a single point of control over decisions the protocol was supposed to protect with multiple independent signers.

This is not good governance.

Subnet emissions can be disabled without a proper public explanation, then enabled again without meaningful notice or community discussion. decisions that directly affect subnet teams, miners, validators, and investors should not depend on one person's discretion.

There is also another transparency issue that bothers me.

Most of the community does not even know who the core developers actively working on Subtensor are, what responsibilities they have, or how major technical decisions are made internally.

Maybe there are valid security or privacy reasons for keeping some contributors private. that is understandable. but if that is the case, the community should at least be told why this level of privacy is necessary.

A decentralized network should not require the community to simply trust an invisible development structure without explanation.

I deeply appreciate what Const has built and contributed to Bittensor. but being a co-founder does not mean one person should permanently control critical network decisions while the rest of the ecosystem simply reacts afterward.

The network should not depend on the approval of a single individual.

Bittensor is no longer a tiny experimental project. there are subnet teams, miners, validators, investors, developers, and community members who have committed serious time and capital to this ecosystem.

Const, you co-founded Bittensor and led it well through its early stages. but the network has grown beyond any single person.

Critical actions should increasingly be transparent, justified, and community-driven. at minimum, privileged keys and decisions should have meaningful checks, independent signers, and clear accountability.

And the development process itself should become more transparent too. the community should have a clearer understanding of who is maintaining the core protocol, how decisions are made, and why certain parts of that structure need to remain private.

If Bittensor wants to become truly decentralized infrastructure, its governance and development transparency need to mature with the size of the network.

This is not about attacking Const.

It is about making Bittensor stronger.

$TAO #Bittensor

[space @concil859856]

@bordermaxxing Righto!

[space @concil859856]

@canarymind I don't mean that we have better vision. Bittensor needs Const, but we also need transparent and clear governance

[Tao Portal @TaoPortal]

🚨 BITTENSOR JUST GOT ITS FIRST Y COMBINATOR SUBNET $TAO

@oroagents, the team behind SN15, has been accepted into Y Combinator’s Fall 2026 batch.

That is bigger than any other accelerator announcement.

ORO is already running an open market on Bittensor, where builders compete to produce better commerce agents and the network rewards the strongest performers. Now the same team is entering one of the world’s most recognised startup programmes.

For $TAO, this is a new kind of external validation: a company can build its core AI system around a live Bittensor subnet and still fit the traditional venture-backed startup model.

The first subnet in YC. Likely not the last. 👀

[space @concil859856]

@lynn_aisv @a16z @JenniferHli Congrats!

[space @concil859856]

@RyanBequette @const_reborn @qBitTensorLabs Cutting off the emissions is understandable, but ignoring them isn't community-friendly. You only allowed SN107, which happens to be your friends, and they aren't fully auditable yet. I'm not sure why this is happening, @const_reborn

[space @concil859856]

Amazing one!

[小樱💞｜实用工具分享 @xiaoying_eth]

出现了一个音声AI能用600多种语言完全复刻你的声音

OmniVoice(Python写的)，GitHub星标快1万了

核心功能：
• 600+语言零样本语音合成
• 用音频样本克隆声音
• 可调年龄、音高、方言、气音
• 实时生成速度快40倍

做多语言内容和视频配音的人应该会很感兴趣

https://t.co/EqlEVv2KGR

[Google DeepMind @GoogleDeepMind]

We’re launching AlphaGenome Atlas: an AI-powered searchable database mapping the predicted impact of all 9 billion possible single-letter DNA changes.

Here’s how it could help researchers better understand our biology 🧵

[space @concil859856]

@Cats_CR @SubnetSummerT Nice work, @Cats_CR
You finally added some real value.

[Pareton @Pareton_ai]

Miners just made Qwen3.8-27B-FP8 3.5× faster than stock vLLM end-to-end on 1× H200 (median over SWE-agent traces):
• 727ms → 190ms median request
• 58.7 → 221.8 tok/s per request
• inter-token latency p99: 93ms → 38ms
• requests meeting a strict serving SLA: 3% → 100%
• identical outputs (≥0.99 greedy token-match)

The winning patch: turn on the checkpoint's own MTP head for self-speculative decoding (γ=8) + fused Triton kernels for the GDN linear-attention layers. 1,090 lines inside vLLM, nothing else touched.

25 rounds, 74 submissions, 31 miners. Week 1. Holding here while we maintain and recalibrate.

[space @concil859856]

@taodotcom Save your tears for the next 5 years 😂

[space @concil859856]

Today I scrolled wayyy back through Discord just to see what was happening 5 years ago.
Fucking Discord, thank you for somehow keeping all that history alive. 😂
It's wild to see how small Bittensor started - just like every other tiny startup with a crazy idea.

Now we've got 100+ teams, thousands of insanely talented technical minds, and an ecosystem that barely resembles what it was back then.

Makes me wonder where we'll be 5 years from now.
If this is what happened in the first five years… imagine the next five.

HOLD!

[umi subnet @umi_sn78]

THE COMPETITION IS ABOUT TO BEGIN.

Mining access is open.
The public miner pilot is live.
The network dashboard is live.

Next: public calibration and scoring.

Soon you’ll be able to watch miners compete, models improve, and translation performance move in real time.

Bring your best battle plan.

Mac or Linux. Your model. Your strategy. Your gladiator.

Apple Silicon or x86. Your model. Your strategy. Your gladiator.

Enter the arena.

May the best intelligence win.

Bittensor SN78 $TAO @umi_sn78

[space @concil859856]

@DFVTAO Ah, it was Bitcast. cool

[space @concil859856]

@DFVTAO What exactly is this rev? and where can I see the breakdown?

[space @concil859856]

Does anyone want ~200ms TTFA voice-clone streaming TTS on a single RTX 4090?
This is a WebSocket streaming server for Qwen3-TTS 1.7B via faster-qwen3-tts, doing about 5× real-time synthesis, streaming 24 kHz PCM audio in 40ms frames, with barge-in/cancel support for real-time voice agents.

GitHub: https://t.co/vaoQzPu6bq