The feed

What’s new

Everything gathered from X, GitHub and video, newest first. Filter by how far back you want to look, and by whether something actually shipped or is only being proposed.

Last night 3 days This week This month

Everything Shipped Only proposed

163 items

1d ago

X

@engyai

RT @DendriteHQ: On Albedo/SN97, we’ve moved most of its LLM traffic from OpenRouter to @engyai - 100% of DeepSeek and 40% of GLM

1d ago

X

@dFusionAI

Both futures use the same technology. The difference is who the intelligence belongs to.

1d ago

X

@dFusionAI

This isn't just the fairer version. It's the more accurate one.

1d ago

X

@dFusionAI

Now the other version.

1d ago

X

@dFusionAI

A single model trained on everything becomes an average of everything.

1d ago

X

@dFusionAI

The monolith version is the default. Scrape everything humanity knows, compress it into one model, own that model, rent access to it.

1d ago

X

@dFusionAI

There are two versions of the AI future being built right now.

1d ago

X

@TensorUSD

Most subnets sell intelligence.

1d ago

X

@sundaebar_ai

SBC17 Executive Upward Communication Skill now closed.

1d ago

X

@nepher_robotics

We never stop—not even for a moment.

1d ago

X

@nepher_robotics

The #Humanoid Run & Jump task is now ready. 🤖

1d ago

X

@desearch_ai

Not every search pass needs an LLM judge.

1d ago

X

@SwarmSubnet

This was only one year ago.

1d ago

X

@TensorUSD

TUSDT support is live on @subwalletapp.

1d ago

X

@btmasxai

Forecast quality isn’t about hype—it’s about reliability + resolution. Numbers that mean something. #MASXAI #Forecasting #AI #CryptoAnalysis #DataDriv

1d ago

X

@trishoolai

RT @ttsheik: @xnavkumar @trishoolai already building and leading this and it's 23 steps ahead of the regulations that will be mandatory at…

1d ago

X

@TensorUSD

Watching your ratio isn't a chore, it's the whole game.

1d ago

X

@btmasxai

Forecasts are only as valuable as the evidence behind them.

1d ago

X

@bitstarterAI

'Building on #Bittensor is still harder and more public than building inside a conventional startup - but for the right problem, you can reach a scale

1d ago

X

@SiamKidd

Having a chat with one of the most misunderstood subnets in Bittensor. A cause & effect prediction subnet: SN21 @adtao_ppcrebel

1d ago

X

@taodotcom

Poker44 has released a preview of its first labeled training benchmark, with 10 micro-sessions and 40 strategic decisions balanced between human and b

1d ago

X

@hippius_subnet

Start uploading your files 👇

1d ago

X

@hippius_subnet

Most clouds encrypt your files after they get them.

1d ago

X

@trishoolai

RT @xnavkumar: Agent containment failures have now officially become a Congressional issue

2d ago

X

@DkingYooo18516

RT @KyoshiTakeshiro: Bitcast miners, ever posted an @Stitch3_ai campaign tweet only to find out it failed validation AFTER?

2d ago

X

@SwarmSubnet

Happy to announce that our CTO will also be joining us at @ExploitSummit!

2d ago

X

@TAOTemplar

RT @Jason: "Can a Blockchain Actually Beat OpenAI? https://t.co/DlrM1uaV43

2d ago

X

@Tao_mans_son

RT @djchazzychas: Monday Bittensor Series — Week 32: Subnet 50 — Synth

2d ago

X

@markjeffrey

RT @beffjezos: So the real reasons that Anthropic nerfed Fable for biology and ML weren't for "Safety" but because those are markets they w…

2d ago

X

@markjeffrey

RT @ShizzyUnchained: The @const_reborn interview on @twistartups was absolutely amazing!

2d ago

X

@markjeffrey

Fable? Or Engy Kimi K3?

2d ago

X

@markjeffrey

RT @actualinc: Muse Glimmer is Meta's 30B open-weight agentic model: tool use, multi-step reasoning, multimodal, and Apache 2.0.

2d ago

X

@TheTaoPod

RT @Tao_mans_son: Great podcast once again, miss the @TheTaoPod and @JosephJacks_ will be missed.

2d ago

X

@DkingYooo18516

RT @nepher_robotics: 🚀 Updates planned for this week at @nepher_robotics :

2d ago

X

@DkingYooo18516

RT @NiomeAI: It’s a new week, and ICYMI, the SN55 miner dashboard is live, the project roadmap is out, and Phase 1 of our data challenge is…

2d ago

X

@DkingYooo18516

Verathos (SN96) is pointing to a concrete problem in today’s AI inference market: “shadow APIs.”

2d ago

X

@ShizzyUnchained

Bittensor fixes this.

2d ago

X

@openroboto

RT @taodotcom: Limit Orders are now available on all our Wallets.

2d ago

X

@byteleap_ai

ByteLeap Subnet — Miner & Validator SDK Upgrade

2d ago

X

@TAOTemplar

Why would miners mine a bittensor subnet that puts miner burn at 100%? https://t.co/KbYnYfmM8N

2d ago

X

@ShizzyUnchained

I LOVE what @openroboto SN80 chart is doing right now.

2d ago

X

@actualinc

Muse Glimmer is Meta's 30B open-weight agentic model: tool use, multi-step reasoning, multimodal, and Apache 2.0.

2d ago

X

@markjeffrey

RT @taodaily_io: Missed @const_reborn's appearance on the TWiST podcast?

2d ago

X

@dFusionAI

POV: You just realized how valuable your Telegram conversations are https://t.co/JEGzrB0iHs

2d ago

X

@markjeffrey

Boring. Dumb. Extremely profitable. Extremely useful. https://t.co/9UKRmtq0dr

2d ago

X

@taodaily_io

Missed @const_reborn's appearance on the TWiST podcast?

2d ago

X

@ShizzyUnchained

The @const_reborn interview on @twistartups was absolutely amazing!

2d ago

X

@const_reborn

RT @b1m_ai: Beam digital commodity is: Bandwidth

2d ago

X

@taodaily_io

RT @SubnetSummerT: 1B+ Weekly Transactions Enter RedTeam’s Adversarial Testing Ground https://t.co/rQVgyCkKh6

2d ago

X

@Tao_mans_son

Great podcast once again, miss the @TheTaoPod and @JosephJacks_ will be missed.

2d ago

X

@TAOTemplar

editing some vids... uh oh https://t.co/hE2NZ2yhRk

2d ago

X

@TAOTemplar

@ExploitSummit Use TAODAILY30 for 30% off!

2d ago

X

@markjeffrey

We've got to get the humanoids online faster.

2d ago

X

@markjeffrey

RT @0xSunRun: Anytime @const_reborn graces the world with an interview like this you should tune in.

2d ago

X

@markjeffrey

RT @hippius_subnet: Hippius Drive is live.

2d ago

X

@markjeffrey

👀 https://t.co/2bpOJqbOTc

2d ago

X

@TAOTemplar

FYI Today's the last day to pickup tickets for the @exploitsummit Bittensor conference before they increase the price.

2d ago

X

@TAOTemplar

RT @DrocksAlex2: I love Bittensor $TAO and the competition between subnets.

2d ago

X

@dFusionAI

Four things would fix most of what's wrong with AI.

2d ago

X

@taodaily_io

RT @twistartups: We're LIVE in one hour with @const_reborn, co-founder of @bittensor .

2d ago

X

@taodaily_io

RT @_olaige: Registered for Exploit Summit and wondering what you need to prepare?

2d ago

X

@basilic_ai

RT @tplr_ai: DeepSeek V4 Flash cuts model depth to 43 layers, bucking the trend of deeper-is-better. Architecture innovation over scale bru…

2d ago

X

@Tao_mans_son

This was great! @const_reborn explains #bittensor from the ground up, exploring every aspect of its function and goals. Good explanation for those tha

2d ago

X

@basilic_ai

RT @tplr_ai: Texas just froze new data center approvals, putting 20% of the US pipeline at risk. This is what compute centralization looks…

2d ago

X

@MarkCreaser

When we backed Subnet 46 @sebyrubino was unproven and new to the ecosystem. He shipped anyway.

2d ago

X

@chutes_ai

Watch the full episode: https://t.co/CE7CITQnFy

2d ago

X

@chutes_ai

Someone hands your training run an update. How do you know they did the work you asked for, and not something else?

2d ago

X

@chutes_ai

Additionally we wanted to note that 23,000 Alpha were left out of conviction to be put toward a treasury.

2d ago

X

@Investing88ai

While waiting for weight setting, we're also planning to reopen the 88 Public Index, working closely with our trusted partner, @TrustedStake

2d ago

X

@taoinsider

RT @Jason: "Can a Blockchain Actually Beat OpenAI? https://t.co/DlrM1uaV43

2d ago

X

@grail_ai

RT @tplr_ai: DeepSeek V4 Flash cuts model depth to 43 layers, bucking the trend of deeper-is-better. Architecture innovation over scale bru…

2d ago

X

@ShizzyUnchained

Bitcoin Dreams It Was Bittensor

SN24 SN53 SN75 SN80

2d ago

X

@taodotcom

Limit Orders are now available on all our Wallets.

2d ago

X

@chutes_ai

More announcements to come. In interim you can verify the conviction here: https://t.co/Lg2Sq4fChe

2d ago

X

@grail_ai

RT @tplr_ai: Texas just froze new data center approvals, putting 20% of the US pipeline at risk. This is what compute centralization looks…

2d ago

X

@taodotcom

Chutes is migrating Subnet 64 from its deployed smart-contract mechanism to conviction lock, which @chutes_ai says enforces the mechanism at the Bitte

2d ago

X

@yanez__ai

Yanez in the news!

2d ago

X

@chutes_ai

Chutes is migrating from our current smart contract mechanism to conviction lock.

2d ago

X

@ShizzyUnchained

RT @twistartups: We're LIVE in one hour with @const_reborn, co-founder of @bittensor .

2d ago

X

@taodotcom

Cancel all plans. Live in one hour!

2d ago

X

@affine_io

RT @twistartups: We're LIVE in one hour with @const_reborn, co-founder of @bittensor .

2d ago

X

@openroboto

We use π0.5, developed by @physical_int, as the base model for the OpenRoboto competition.

2d ago

X

@TensorUSD

The dividend structure splits by what you hold:

2d ago

X

@bitstarterAI

RT @cascade_sn91: On GIFT-Eval: Cascade 4M ≈ Salesforce 91M

2d ago

X

@MacrocosmosAI

The simplest way to picture it: imagine emptying an Olympic swimming pool with a bucket.

2d ago

X

@MacrocosmosAI

A real example from the 16B run. Our monitoring picked up that a specific set of GPUs from one provider had a networking fault, bandwidth on those mac

2d ago

X

@MacrocosmosAI

Here's the counterintuitive part: a machine doesn't have to be online 100% of the time to be useful for training.

2d ago

X

@MacrocosmosAI

In a distributed run, this matters more, because machines depend on each other.

2d ago

X

@MacrocosmosAI

There is a quiet assumption when it comes to most AI training: that the compute you're using is stable, uniform, and always there.

2d ago

X

@inference_labs

Sertn is built for computer vision in environments where failure is expensive.

2d ago

X

@nepher_robotics

🚀 Updates planned for this week at @nepher_robotics :

2d ago

X

@NiomeAI

To get started as a miner, the setup guide is at: https://t.co/YfBxmPo82u

2d ago

X

@NiomeAI

It’s a new week, and ICYMI, the SN55 miner dashboard is live, the project roadmap is out, and Phase 1 of our data challenge is running now and expecte

2d ago

X

@dippy_ai

We are FINALLY back online! We deeply apologize for this issue extending nearly 24 hours 🥲

2d ago

X

@DkingYooo18516

Bitcast (SN93) is pushing its creator marketing products beyond Web3.

2d ago

X

@DkingYooo18516

RT @chutes_ai: 4.7M parameters per expert where a standard expert carries 12.6M. Same shape, but roughly one third of the size.

2d ago

X

@DkingYooo18516

RT @poker44subnet: Poker44 v3.1.0 introduces a new training benchmark built from live poker gameplay.

2d ago

X

@DkingYooo18516

Minotaur (SN112) is framing a bigger ambition for DeFi execution this week.

SN44 SN77 SN112

2d ago

X

@taodotcom

RedTeam team member Oscar says he believes more than 700,000 alpha is currently locked, and separately says 90% of the subnet's tokens are locked perp

2d ago

X

@TensorUSD

Did you know your collateral in TensorUSD never expires?

2d ago

X

@dFusionAI

The era of your activity feeding someone else's product doesn't have to be the whole story.

2d ago

X

@minotaursubnet

Welcome to a new week, and we’d like to remind you that what Minotaur is building has the potential to change the whole of DeFi.

SN44 SN77 SN112

2d ago

X

@jaltucher

RT @TheTaoPod: Just in case you missed it on YouTube over the weekend!

2d ago

X

@TheTaoPod

Just in case you missed it on YouTube over the weekend!

2d ago

X

@trishoolai

𝟭𝟮𝟱,𝟬𝟬𝟬 𝗔𝗹𝗽𝗵𝗮. Locked perpetually.

2d ago

X

@tplr_ai

DeepSeek V4 Flash cuts model depth to 43 layers, bucking the trend of deeper-is-better. Architecture innovation over scale brute force.

2d ago

X

@gradients_ai

RT @SubiaWaud: Messy few days going through issues with @QuasarModels, but a lot went right:

2d ago

X

@tplr_ai

Texas just froze new data center approvals, putting 20% of the US pipeline at risk. This is what compute centralization looks like as a single point o

2d ago

X

@tplr_ai

Alibaba ships Qwen3.8 Max with 2.4 trillion parameters and calls it the end of the fully free open source era. A 2.4T open-weight model from a non-US

2d ago

X

@SomaSubnet

SOMA Roadmap - where we actually are.

2d ago

X

@Bitcast_network

Over the coming weeks, we’ll be expanding our creator marketing products beyond Web3.

2d ago

X

@chutes_ai

The full episode: https://t.co/CE7CITQnFy

2d ago

X

@chutes_ai

4.7M parameters per expert where a standard expert carries 12.6M. Same shape, but roughly one third of the size.

2d ago

X

@TensorUSD

When minting in TUSDT you can create vault and mint TUSDT at the same time.

2d ago

X

@desearch_ai

Retrieval should not turn an agent into a single-file queue.

2d ago

X

@SwarmSubnet

789 models entered SWARM.

2d ago

X

@taodaily_io

🚨 10 HOURS LEFT to save 30% on your Exploit Summit ticket.

2d ago

X

@TensorUSD

Put the loop and the market together:

2d ago

X

@bitsecai

RT @YVR_Trader: Numerous subnet gems hiding in plain sight:

2d ago

X

@taodaily_io

RT @Laura__crypto: “Owning $TAO is like owning the S&P500 of all Bittensor subnets.” This hits different when you actually think about it.…

SN51 SN64 SN107

2d ago

X

@openroboto

Thanks for featuring us! Cooking now. https://t.co/KrnuKfO7Rj

2d ago

X

@taodaily_io

https://t.co/iol6LCAgWG

2d ago

X

@hippius_subnet

Hippius Drive is live.

2d ago

X

@Tao_mans_son

RT @Taonet_daily: SN19 is so underrated

3d ago

X

@sobczak_mariusz

My vacay is coming to an end and I am little bit bored. Here is my framework on how to value $TAO subnets that have emerged as leaders and are heading

3d ago

X

@TatsuEcosystem

Sunday Brief – Aug 9, 2026

3d ago

X

@DkingYooo18516

I just won 10 $USDT on the Reign Wheel on @xreign_app.

3d ago

X

@DkingYooo18516

RT @REALDIEGOVALDEZ: That’s why people like @rBryer23 succeed… he did it with $BTC, and now he’s doing it again with $TAO and the $TAO ecos…

3d ago

X

@Old_Samster

RT @micaelabazo: while other networks waited for sentiment to change, @opentensor spent the summer cleaning house and changing its structur…

3d ago

X

@webuildscore

Note: The Studio landing page design and copy are still evolving.

3d ago

X

@webuildscore

Every Computer Vision engineer knows the ritual. Beg for footage. Wait for legal. Ship a scraper. Pay an annotation farm. Six weeks later you finally

3d ago

X

@dippy_ai

We are experiencing an issue with our database provider @supabase , so the app & website may be down for a few more hours.

3d ago

X

@Tao_mans_son

RT @peytonspencer: Clarifying our leadership structure at Ditto:

3d ago

X

@dFusionAI

The ending to to the AI story isn't written.

3d ago

X

@Old_Samster

RT @alpha_pls: I've never returned to a project as much as Bittensor.

3d ago

X

@ShizzyUnchained

ENGY SOLVES WORLD HUNGER!!! 😂

SN4 SN51 SN53 SN64

3d ago

X

@Old_Samster

positioned for RedTeam to become a top ten subnet https://t.co/TyWQtSILGh

3d ago

X

@taodaily_io

RT @taodaily_io: 🚨 Mark Jeffrey (on The TAO Pod):

3d ago

X

@SynthdataCo

Every market has a volatility surface

3d ago

X

@jaltucher

RT @Steve_too: Worth a listen, as always, thanks @jaltucher and @markjeffrey.

3d ago

X

@inference_labs

In high-stakes environments, a model prediction is not just a prediction.

3d ago

X

@bittensor

RT @opentensor: “We’re already in a state of permissionlessness far beyond any example of a similar incentive network.”

3d ago

X

@bittensor

RT @opentensor: Recently, 5 more enterprise platforms worth $22B+ combined have been testing @_redteam_’s technology. They reach 900M+ acco…

3d ago

X

@bittensor

RT @opentensor: https://t.co/BhZToz8eS8

3d ago

X

@SwarmSubnet

RT @LouiseBeattie: https://t.co/TqKp3NpDVp

3d ago

X

@ShizzyUnchained

RT @openroboto: Over the past week, OpenRoboto shipped updates across the website, submission system, and benchmark evaluation pipeline.

3d ago

X

@Tao_mans_son

RT @openroboto: Over the past week, OpenRoboto shipped updates across the website, submission system, and benchmark evaluation pipeline.

3d ago

X

@opentensor

https://t.co/BhZToz8eS8

3d ago

X

@ShizzyUnchained

STOP DOUBTING ME!

3d ago

X

@Bitcast_network

RT @will_mizu: Here's a @Stitch3_ai campaign we ran for a prediction markets client.

3d ago

X

@desearch_ai

The first page is not the dataset.

3d ago

X

@openroboto

The website now includes Brand Assets, mobile navigation, and a more complete Queue result view. https://t.co/o7OLSmmUjO

3d ago

X

@openroboto

We also updated the model submission rules and documentation:

3d ago

X

@openroboto

Submission history and result tracking have also been expanded.

3d ago

X

@openroboto

The Evaluation Queue now covers the full submission lifecycle:

3d ago

X

@openroboto

Over the past week, OpenRoboto shipped updates across the website, submission system, and benchmark evaluation pipeline.

3d ago

X

@taodotcom

devs cooked on this tbh

3d ago

X

@webuildscore

Very interesting question from @jaltucher at 11:00:

4d ago

X

@trishoolai

Congrats to @_redteam_ on this massive market validation.

4d ago

X

@jaltucher

Very pretty puzzle composed by David. https://t.co/QlV3vJpcgL

4d ago

X

@jaltucher

RT @taodaily_io: 🚨 Mark Jeffrey (on The TAO Pod):

4d ago

X

@ShizzyUnchained

You missed ENGY SN53.

4d ago

X

@SiamKidd

Looking forward to seeing how these Bitcoin Agents perform, moving forwards! First of the new type of Battles start 1st Sep. https://t.co/LS9Etg3Dex