TAO Community Hub

The unofficial $TAO and #Bittensor Community on Twitter.

[TAO Community Hub @TAOCommunityHub]

Bittensor is a fascinating project that represents a novel approach to the creation and distribution of machine learning models. Here are the most important aspects of Bittensor:

Decentralized AI Marketplace: Bittensor is essentially a decentralized network that allows anyone to create, share, and monetize machine learning models. It's like a marketplace for AI services where models can be accessed and used by anyone on the network.

Token Incentivization: The network uses its own cryptocurrency token- $TAO to incentivize participation. Miners, or those who contribute computational resources to the network, are rewarded with tokens. Similarly, those who provide valuable models can earn tokens when their models are used.

Dynamic Consensus Mechanism: Bittensor uses a unique consensus mechanism to ensure that the best and most useful models are easily accessible. This mechanism is based on the usage and performance of the models, which means that higher-quality models become more prominent on the network.

Scalability and Efficiency: By distributing the workload across multiple miners, Bittensor can potentially offer more scalable and efficient machine learning solutions compared to traditional centralized systems. This distributed approach can also provide a level of redundancy and fault tolerance.

Democratization of AI: Bittensor aims to democratize access to AI by allowing a wide range of participants to contribute and benefit from machine learning models. This could lower the barriers to entry for AI development and usage, fostering innovation and competition.

Privacy and Security: The decentralized nature of Bittensor can offer enhanced privacy and security features. Since there is no central point of control, it is more difficult for any single entity to manipulate or compromise the network.

Continuous Evolution: The models on Bittensor are not static; they evolve over time. As they are used and interact with different inputs, they can improve and adapt, which is a key feature of machine learning systems.

Cross-Domain Collaboration: Bittensor facilitates collaboration across different domains and industries. Since the network is open, models developed for one purpose can be adapted and utilized for another, leading to cross-pollination of ideas and techniques.

Bittensor represents a step towards a more open and collaborative future for AI, where the benefits of machine learning can be more widely distributed and where innovation is driven by a diverse community of contributors.

[TAO Community Hub @TAOCommunityHub]

@VenturaLabs @betsparket @sparketdotai this was a good interview. maybe @markjeffrey will have them on too

[badenglishtea @badenglishtea]

Im going to go ahead and say it.
Tao flow doesn't work.
I think it is now damaging a number of high quality subnets on Bittensor, a few of which will not recover.

The network should revert, implement proper governance, and open a public forum to improve the emission mechanic going forward.

Here are some of the problems with flow.

1. The starting value of flow implementation was far too high. It entrenched subnets that had deep existing liquidity pools and left all other subnets subject to the mercy of dtao speculators.

2. The network is still not actually distributing emissions based on flow. Flow was implemented 1 month ago. During that time Ridges (62) had a flow of negative -9kt and Iota (9) had a positive flow of 3kt. Both are at the same emission (2%). The half life of EMA reduces the effectiveness of flow. If flow is so good, why isn't Bittensor actually using it.

3. Tao flow advantages subnets with whale backing. Injecting flow into a subnet generates emission. This effectively allows any subnet to buy emission. The first subnet to financially engineer a circular money flow via DAT or traditional financing will dominate emission irrelevant of subnet value.

4. It will death spiral any low liquidity subnet that goes to 0%. Every subnet that loses emission will probably die. The compounding effect of investor/miner selling pressure without emission support will force a number of high quality subnets to die.

5. New subnets cannot enter this system. Without deep existing liquidity, all new subnets will be at the mercy of the small number of dtao speculators that are jumping from hype to hype. Sustainable flow is effectively impossible to achieve.

6. Tao flow is not actually killing the bad subnets. The majority of top 50 subnets had negative tao flow over the last month. The majority of subnets 50 - 100 had positive tao flow over the last month. My assumption was tao flow is an attempt to kill all but the top 10 - 20 subnets, but there are now more subnets farming 0.10% - 0.50% emission than there was pre-flow.

7. Good subnets are dying. Synth, Quantum Compute, Sportstensor, ReadyAI, Talisman are all at 0% emission. There are now multiple subnets with no GitHub that are 1%+.

8. Whales can now market make their own subnets through V3 liquidity pools to limit emission downside. Stacking dense liquidity pools into the buy side can eliminate the majority of outward flow, enabling the subnet to maintain emission perpetually.

9. Subnets cannot launch into a system with such extreme volatility. How can a subnet team be expected to hire resources, make commitments, have a consumer grade product when one dtao speculator can push their emission to zero instantly.

10. Subnets cannot operate in an environment where the rules are changed randomly, and without governance/transparency or time to adapt. Tao flow breaks a number of subnets' business models. It kills miners on most subnets. It kills any subnet with an expensive resource.

11. The process of implementation was largely flawed, and leaves the network open to accusations of manipulation. Any subnet that knew of the change to flow in advance is at a significant advantage over subnets that didn't know the change was coming. There must be more transparency within Bittensor, and if any change can be manipulated, it should never be implemented.

12. I believe Bittensor only survives if new teams can competitively enter the ecosystem and succeed. Flow makes Bittensor less attractive to external teams that may be considering entering the ecosystem.

The reality is, there is very little revenue in Bittensor and this is something the network must accept. It is very hard to build a consumer facing, revenue generating product, with all the requirements of an enterprise system (security/privacy/reliability/uptime/integrity) through a decentralized network like Bittensor. Incentives attract manipulation. Malicious miners or validator disagreements can disrupt a system that has customers or obligations outside of Bittensor. Most of the existing revenue is just miners rebuying supply via $TAO. Also no subnet is under any obligation to route any revenue generated back into their alpha token.

With that said, I believe flow was implemented with good intentions.

1. Bittensor has a value recapture issue. People describe Bittensor as a flywheel, and it is very good at emitting value ($/resource), which attracts talent and utility. It is not very good at recapturing that value (revenue), and reinjecting that value back into the network (to return value to $TAO holders). The flywheel must be a closed loop for Bittensor to succeed.

2. Flow massively squeezes supply at the expense of subnets. This is good for $TAO price. Much less sn/miner $TAO flows to exchanges. If you are a $TAO holder you probably like this. This will be at the expense of a number of high quality subnets failing, and will probably limit new high quality subnets from entering.

My criticism of flow is primarily from speaking to a number of subnet owners and miners. I have always been an advocate of dtao, and have defended dtao as a very well designed system. I think Bittensor works well. Eliminating low quality subnets SHOULD be a priority. Also improving value recapture is critical.

What is the purpose here? If the goal is to attract new subnet teams to Bittensor and for the network to grow, flow is stopping that. Flow mostly benefits existing high liquidity subnets (not a criticism of good subnets).

There is another path here. I think Bittensor can learn from flow, and adapt a better, more balanced mechanism through governance.

[badenglishtea @badenglishtea]

Interesting. Actual quantum computers are running through bittensor like right now. Probably nothing.

[Quantum Compute @QuantumSN48]

Happy Halloween from OpenQuantum!

[badenglishtea @badenglishtea]

You have 2 subnets on Bittensor. One runs quantum simulation, the other will run actual quantum machines. Both will feed into @OpenQuantum_ in <2 months. They are run by @BeyondNISQ who is currently at Quantum World Congress just featured by Nvidia. Current rank 25th + 28th lol

[qBitTensor Labs @qBitTensorLabs]

That’s a wrap on QBTLL #1.

Give it a watch + listen and let us know your initial thoughts 🩵

Plenty of time to ponder it all over the weekend.

[qBitTensor Labs @qBitTensorLabs]

QBTLL - Tomorrow at 16:00 UTC.

The soothing voice of @BeyondNISQ will guide the Bittensor community through our first livestream discussion on all things Subnet 48 (Quantum Compute), Subnet 63 (Quantum Innovate), and @qBitTensorLabs.

Which live-streaming platforms have been successful in $TAO before? We need audio + video + recording + (maybe live chat) capabilities. Examples: Riverside FM, Discord, Twitter Space, Youtube Live etc.

[qBitTensor Labs @qBitTensorLabs]

In a milestone moment for both the Bittensor and quantum computing communities, @qBitTensorLabs is officially publishing the breakthrough solution to the “Hidden Stabilizer Circuits” (HSTAB) challenge, in collaboration with the Miner who first discovered it.

The winning Miner (who has chosen to remain anonymous) unveiled a novel method for tackling complex quantum circuits of a specific composition, opening new pathways for innovation and demonstrating the untapped potential of decentralized research.

This marks an early validation of our thesis: that by exposing quantum challenges to a highly motivated and incentivized set of innovators within the Bittensor network, previously unrealized innovations can be discovered and accelerated.

The qBitTensor Labs Quantum team has worked closely with the Miner to produce a comprehensive technical paper detailing the approach, including the source code necessary to reproduce the result. As part of this milestone, the HSTAB circuit will be formally removed from the network in the coming days.

For those interested in examining the technical details, the full write-up is available here:
https://t.co/iR1tKuAdK6

By publishing this solution, we aim to expand the collective toolkit of quantum researchers and inspire new methods of circuit design and simulation; because with each solved challenge, we move closer to unlocking the next level of quantum-native innovations on Bittensor.

P.S. As some astute qBitties noticed, our website has been upgraded -- designed to better reflect our commitment to professionalism, transparency, innovation, and outreach beyond $TAO 🩵

[𝗗𝗥𝗘𝗔𝗗 𝗕𝗢𝗡𝗚𝗢 @DreadBong0]

#Targon on Subnet 4 announced some huge updates 🔥

A suite of new and improved products coming to their ecosystem

🔵 Targon v6 - Delivering enterprise grade inference at affordable prices

🟠 https://t.co/PSs0LfW2tV - A brand new bloomberg style terminal for the #Bittensor ecosystem

🟢 Sybil - Improved user experience for their hybrid search engine

🟠 TAO XYZ Wallet - New intuitive browser extension to effortlessly swap Alpha

No wonder sn4 is hoovering up emissions within the network

$TAO

[TAO Community Hub @TAOCommunityHub]

@Bettensor 🔥

[Al @Al_3194]

Only 4%.

Looks like we're just getting started.

$TAO

[𝗗𝗥𝗘𝗔𝗗 𝗕𝗢𝗡𝗚𝗢 @DreadBong0]

Dynamic $TAO will have accessible and visible charts through the cli 🔥

Someone please build a simple #dTao website interface with these charts 🙏

The aesthetic is 🔥🔥

$TAO

[Ventura Labs @VenturaLabs]

Ventura Labs Ep. 8 - Iain Head & Nickel5 Team

Iain (@badenglishtea) is the Founder, Noah (@geardici) is the lead Developer on Bettensor (@Bettensor), Matt is lead Developer on NextPlaceAI (@NextPlace_AI), and Steven leads general AI/ML development at Nickel5 (@Nickel5_)

Timestamps:
1:06 - Introduction
1:47 - Nickel5 Team
3:16 - Doxxed And Transparency
7:56 - Investment Thesis in Bittensor
10:24 - Building on Bittensor
12:58 - Attraction to Bittensor
18:21 - From Moonshot to Nickel5
19:51 - Subnet Development
26:18 - Bettensor
34:39 - NextPlace AI
43:15 - New Subnet Geospatial
56:25 - Bittensor Enhancing Value
1:00:46 - dTAO
1:07:08 - Challenges in Building Subnets
1:14:23 - Advice for Success in Bittensor

[Openτensor Foundaτion @opentensor]

"Intelligence at the edge of chaos" — A maxim for Bittensor.

SN36: Formed at the crossroads of groundbreaking research and mysterious computational phenomena, driven by the potential of Rule 30 and cellular automata.

@WOMBO @KeithSingery

Full Episode on Spotify & YT:
https://t.co/HTnPB3afkW

[TAO Community Hub @TAOCommunityHub]

@Bettensor 🔥

[Fish | Datura @fish_datura]

With just 0.4% emissions and 2 days since incentivization, we already have ~100 GPUs available to rent at https://t.co/baVl18BAim at half the price from runpod. All powered by subnet 51 on Bittensor with $TAO.

[Learning Crypto @learningcrypto]

$TAO Bittensor is making another move, check our recent review on this decentralized AI network 👇

[TAO Community Hub @TAOCommunityHub]

@Bettensor 👀

[Openτensor Foundaτion @opentensor]

AI History is being made

Distributed Training on Bittensor

Made possible with @MacrocosmosAI + @manifoldlabs + @const_reborn

[TAO Community Hub @TAOCommunityHub]

@Palaidn 👀