Bittensor subnet news

Instant Inference (gpu/acc)@thinkinstant

Models that never say no. That’s our mission. Models that enable true innovation. That’s our use case. Listen in to Instant Founder @sebyrubino explain the vision with @JesusMartinez.

Quoting @jesusmartinez

"The New House is GPU's" "That's where the world is going" My full chat w/ @sebyrubino, founder of @thinkinstant a subnet on bittensor:native 0:00 Intro 0:38 Meet Seby Rubino, founder of Instant (SN46) 2:34 Why Instant skips TEEs and runs one model 4:08 Collateral and slashing: cheat and you lose your money 6:23 Uncensored inference: building a based GPT 9:43 The full stack: inference, training and first customers 13:14 What happened to RESI and Zipcode 15:15 Zipcode becomes custodial GPU loans for miners 18:14 Short the house, long the GPU 19:28 The goal: millions in ARR and a deflationary subnet

Full post on TaosisOpen on X ↗

Instant Inference (gpu/acc)@thinkinstant

FIRST BUYBACK & BURN ⚡️ 1,199.03 SN46 alpha bought back and burned onchain. Real demand for uncensored intelligence, powered by Subnet 46. Crypto payments and more listings are on the way. Onchain verification below:

Full post on TaosisOpen on X ↗

Instant Inference (gpu/acc)@thinkinstant

Track usage and burns on our dedicated dashboard: instantnetwork.ai/usage Burn #1 onchain proof: taomarketcap.com/extrinsics/9449718-0012

Full post on TaosisOpen on X ↗

Instant Inference (gpu/acc)@thinkinstant

When the clouds go down, Instant stays up. This is why we’re building a permissionless network instead of a corporation. 1B tokens served this past month and new customers onboarding steadily. As the “traditional” market falls, we rise. $TAO

Full post on TaosisOpen on X ↗

Instant Inference (gpu/acc)@thinkinstant

Why the uncensored / abliterated frontier must accelerate… We are too vulnerable and must protect ourselves. Instant is one way to fight back.

Quoting @watcherguru

BREAKING: World's largest crypto exchange Binance hacked with over $5.6 billion stolen: • $1.45 billion $BTC • $875 million $ETH • $234 million $TRX • $50 million $XAUT • $1.57 billion $USDT • $823 million $USDC • $75 million $WBETH • $88.5 million $RLUSD + $447.5 million additional cryptocurrencies This is the largest hack in crypto history.

Full post on TaosisOpen on X ↗

Instant Inference (gpu/acc)@thinkinstant

NEW RELEASE: The Instant API portal is live! ⚡️ Want to start with 25M tokens for free? Comment "Instant" below and DM your email - limited to the first 100 accounts. Access uncensored GLM 5.3 Flash at 200+ TPS, with just a few clicks:

Full post on TaosisOpen on X ↗

Instant Inference (gpu/acc)@thinkinstant

Step 1: Comment “Instant” under the first post. Step 2: Sign up to instantnetwork.ai/portal Step 3: DM @thinkinstant on X with the email you used to sign up. Happy Building ⚡️

Full post on TaosisOpen on X ↗

Instant Inference (gpu/acc)@thinkinstant

INSTANT PROVIDER UPDATE The new engine is live. TensorFold is a familair engine to many that is built for speed, cost and reliability. What changed:

Full post on TaosisOpen on X ↗

Instant Inference (gpu/acc)@thinkinstant

Proof is exact now. One bad answer gets caught, not guessed. Hosts have to clear a real speed floor, and your data never lands on disk. Images work now as well. How to run it:

Full post on TaosisOpen on X ↗

Instant Inference (gpu/acc)@thinkinstant

Miners: RTX PRO 6000 ×4, with CUDA peer access. Check your host: sn46-miner qualify Then install: github.com/Subnet46/sn46-miner-releases Onwards, faster. ⚡️

Full post on TaosisOpen on X ↗

Instant Inference (gpu/acc)@thinkinstant

Due to inbound interest from the post below, the remaining 300~ prompts from our free-trial were swiftly used. The short list of network upgrades required to scale paid customers will be finalized in the next 24-48 hours. We have extended 500 free prompts in the meantime!

Quoting @sebyrubino

@entrptaher Uncensored GLM 5.3 Flash does the trick. @thinkinstant you can try it for free.

Full post on TaosisOpen on X ↗

Instant Inference (gpu/acc)@thinkinstant

500 Free Prompts loaded for you to test our network! >instantnetwork.ai< The Instant paid chat & api portal will be available as soon as our expected upgrades to the network are live. The update will also include an upgraded usage dashboard.

Full post on TaosisOpen on X ↗

Instant Inference (gpu/acc)@thinkinstant

Instant MinerOS will be a major milestone for our mechanism. Private and uncensored Inference, end to end, minus the cost or slow downs of TEEs. Prices stay low while tokens per second grow. $TAO will be the venue for the highest performance spectrum of inference.

Full post on TaosisOpen on X ↗

Instant Inference (gpu/acc)@thinkinstant

Access (discounted) accelerated and uncensored inference on AntSeed: antseedstats.com/sellers/0x8dfd8d7d93702738d985…

Full post on TaosisOpen on X ↗

Instant Inference (gpu/acc)@thinkinstant

Instant Miner Service Announcement: Prepare to update your machines in the coming hours as our new engine configuration becomes available. Happy mining! ⛏️

Full post on TaosisOpen on X ↗

Instant Inference (gpu/acc)@thinkinstant

We’ve reached our first 10,000 requests on @AntSeed in the past 7 days. Marking 400% WoW growth. ⚡️🌱 The updated Instant Network revenue dashboard and paid api portal will be published to our website before the end of tomorrow.

Full post on TaosisOpen on X ↗

Instant Inference (gpu/acc)@thinkinstant

There are many ways to enable privacy that don't require TEEs. This is the path to the maximal intelligence per second. Listen in to @sebyrubino explain our stance on privacy and performance.

Quoting @VenturaLabs

Is a decentralized AI miner any more private than ChatGPT, and do users even care?

Full post on TaosisOpen on X ↗

Instant Inference (gpu/acc)@thinkinstant

The Instant API has been live for ~2 weeks… Next, we will enable local models to run on company owned machines and for those machines to be rented by the network. Instant Inference is going local, giving businesses the infrastructure they need.

Quoting @sebyrubino

Subnet 46, Reborn. How @thinkinstant re-enters the top 20 on $TAO.

Full post on TaosisOpen on X ↗

Instant Inference (gpu/acc)@thinkinstant

Instant Inference reservations will soon be available on the Liquid Marketplace! Businesses can no longer afford the disfunction of variable rates and reliability. Cost and price are not the same. Great work @Architect_Fi / @BrettHarrison.

Full post on TaosisOpen on X ↗

Instant Inference (gpu/acc)@thinkinstant

Every answer on Instant is checked by the network, and those checks no longer expose any of your data. Next, we're making those checks cheaper to run. That keeps quality high and pushes cost down.

Full post on TaosisOpen on X ↗

Instant Inference (gpu/acc)@thinkinstant

Up next is a faster engine. Speed, cost and reliability are how we win, and our upcoming engine update moves all three forward. Onwards, faster. End of thread.

Full post on TaosisOpen on X ↗

Instant Inference (gpu/acc)@thinkinstant

Privacy isn't a promise anymore. It's built in. Your prompts stay encrypted in memory and are erased minutes after you get your answer. Nothing is saved to disk. If a machine could leak your data, Instant won't run on it.

Full post on TaosisOpen on X ↗

Instant Inference (gpu/acc)@thinkinstant

Instant is now uncensored AND untraceable. Now: prompts are wiped the moment they are answered. Nothing is stored, ever. Next: a faster engine and less costly checks. More speed, less cost. More below🧵

Full post on TaosisRead the thread on X →Open on X ↗

Instant Inference (gpu/acc)@thinkinstant

Instant distribution and development is underway. Read our weekly report below for a snapshot of where we are, and where we are going. x.com/i/article/2111133934402625536

Full post on TaosisOpen on X ↗