J-Classic Beats 3.0

@fitforcrypto_ Exactly 🎯 The most important thing I did all 22/23 was rebalancing my sources of information. Made it my #1 mission to learn and hunt down the alpha & biggest 🧠s I could find which lead to stepping my own research game way up ✊🏾. Here everyday 🫡

@opentensor @Rapido_ai @heydittoai @const_reborn 🔥🔥🔥

This Thursday on Novelty Search :: SN118 Ditto @heydittoai is building an open-source agentic operating system on Bittensor :: giving AI agents persistent memory, collaborative workspaces, and long-running context. Join @const_reborn and the Ditto team to explore SN118’s incentive mechanism :: miners compete on procedurally generated memory and tool-use benchmarks, validators deterministically reproduce every submission, and rewards are earned through transparent, reproducible evaluation. Thursday :: 9PM UTC / 5PM EDT Live via Bittensor Discord. Hosted by @const_reborn

@404gen_ 🔥🦾

This is your reminder to play The Dream, 404’s Chroma Award-winning game built using SN17-generated assets. At 404, we aim to help kickstart the next revolution in gaming around AI-native games. That starts with democratising 3D content creation through Bittensor, making it easier for creators to build virtual worlds, games, and AR/VR/XR experiences without needing massive production teams. The long-term vision is to let creators describe the game or world they want to build, then generate the assets, sound, voice, and environment around it. The Dream is an amazing example of that vision. It is a fully playable 3D open-world experience created by a two-person team in just 30 days, using over 12,000 assets generated with 404. The goal is not to replace enterprise game development. It is to create a new layer where individuals and small teams can build and explore high-fidelity game worlds without needing a massive AAA-style pipeline. Sometimes, you need to see it to believe it. Play The Dream yourself at: dream.404.xyz

@chutes_ai @jon_durbin 🔥🔥🔥

@jon_durbin pre-trained a 20B MoE for under $10 an hour of compute. Not on a cluster. Eight rented single-L40S VMs scattered across two continents, plus a few 4090s and a 5090, all holding roughly 6 seconds per step. Pre-training was supposed to be the part you couldn't do cheap.

@const_reborn @say_gm_ GM

The headline is that @say_gm_ is the first verifiably private router. In small print, its also permissionless -- so the cost to the user is dropped continuously by markets. saygm.com

@MarsSmuff GM 😎

GM ☀️ bittensor:native

@LeadpoetAI 🔥🔥🔥

Claude just hired Leadpoet to find its next customers. Leadpoet is the intelligence layer sales teams use to find and qualify companies most likely to buy. Now, with #OKXAI, other AI agents can access that same intelligence directly. Agents give Leadpoet a target customer profile. Leadpoet returns the companies worth pursuing, who to contact, and why now. We’re excited to be the first Bittensor subnet on the @OKX agent marketplace.

@say_gm_ 🔥🔥🔥

Gm is live on mainnet. Open to everyone. Drop-in compatible with your existing OpenAI, Anthropic, and Gemini code. Runs in a TEE. Verified by hardware, not promises.

@MarsSmuff GM

GM ☀️ bittensor:native

@AltcoinMillie 💯💯

$TAO tokenizes the market for intelligence. $TIG tokenizes the property rights to it. Those aren't competitors they're different layers, and only one of them holds up if a Fortune 500 company uses your work without paying, now lets go deeper and I'm welcome to the challenge, TIG IS the ownership thesis. Patent-backed IP, dual licensing, token backed by the algorithm pool. $TAO gives you ownership of emissions; $TIG gives you ownership of the innovation itself, enforceable off-chain. Different layers. Both can win. 'Only chain' is just narrative gatekeeping. I hold BOTH and will continue. Please stop comparing these tokens and even better please stop trying to turn everything in to a cult. Demand finds supply I'm confident both will do multiplyers at the end everybody will be happy. #Positivevibes

@MaxScore 🔥🔥🔥

summer business dev update: avia (fuel stations) → converted reading fc (football) → converted mettle data (cricket, ex card) → converted eyeball (youth football) → converted lavance (car wash) → in progress cémoi (chocolate) → to be launched two-a-day (fruit) → on hold converted = 1 or multi-year agreement signed remember the ✍️? that was this, plus partnerships we're now finalising with aws (thank you pwc), and integrations of our models and vision skills by leading security and tech companies. we'll announce each one properly. there's a lot more in the pipeline. can't stop, won't stop.

@macrocrux 🔥🔥🔥

We’ve been somewhat quiet since sharing our Orion-100B results, but rest assured we’ve been working very hard on the next stage of Project Orion. For distributed training to be commercially successful, we believe it must produce models of the same quality, at the same speed, at a fraction of the cost. My co founder @WSquires has being writing a lot the last couple of weeks about the third point, cost. Meanwhile, I’ve been working with the team to test the limits of our tech in terms of model quality and training speed. Orion-100B is very important as it demonstrates that @IOTA_SN9 can efficiently train models at hundred billion parameter scale at 65% the speed of centralized using cheap commodity hardware. As part of our prep for the next public release we’ve been running internal convergence tests to address the question of model quality: Our experiments confirm that IOTA models converge to the same quality as centralized, despite being trained fully distributed in an interruptible fashion over the open internet. A year ago this would have been a result in itself. Pumped for the next model drop. Coming very soon.

@webuildscore @ttsheik Nice!! ⚡️⚡️

We're building the frontend that will host the decentralised VLM miners will train on sn44. The model will be monetised from day one, at subnet level. Users will try it free, then subscribe, through the interface or via API. Payments will be done at subnet level too. Note: product branding might change to Score, so it's clearly attached to the subnet and less confusing, this interface is just vibe coded but the scaffolding works.

@opentensor @TargonCompute @const_reborn @manifoldlabs Let’s gooo

This Thursday on Novelty Search :: SN4 Targon @TargonCompute is building a confidential, decentralized compute cloud >> Secure GPU and CPU infrastructure for AI training and deployment. Join @const_reborn and the team from @manifoldlabs talking confidential compute, TEE, Targon-At-Home + launch of TargonOS Thursday :: 9PM UTC / 5PM EDT Live via Bittensor Discord. Hosted by @const_reborn

@markjeffrey 🔥🔥🔥

Targon subnet 4 - this Thursday.

@haitzu @MaxScore 🔥🔥🔥

Just got a masterclass from @MaxScore on computer vision and training robots: - Most robot-training techniques bottom out in collecting visual data, hence the flood of capital into training facilities and camera-headset human-recording rigs - Manako started from a reasoning model and taught the model to see, rather than starting from a vision model and bolting on reasoning - Manako's VLA is a distilled reasoning model + a traditional VLM + a pure detection model, running ~10 bespoke primitives. - The world is shifting toward World Models: "being intelligent is not about what you know, it's about what you do when you don't know" - This is broadly supported by firms like Physical Intelligence; Predictive (LeCun / JEPA-style joint embedding predictive architectures): you don't need to see at all, you predict embeddings or latent states. Max is fascinated by this Vision is not solved, we are waiting for the Transformer moment we saw with LLMs. Max would like to contribute to progressing research here via Manako and their partners

@arnod3f @manakoai 🔒’d in 🦾🔥🔥

from manned/unmanned to autonomous operations. making @manakoai users feel like tony stark. starting with the fuel retail industry. 1. it starts with curing blindness. - fuel retail businesses operate on a fraction of their data through their current stack. - through their cameras they can see all. 2. we awaken the operation to the understanding of its physical reality. - why is this customer driving away? - when did this carwash get damaged? » 3. leveraging reality into insights humans can act on. - this station has 52% more payment terminal downtime than network avg due to bad maintenance. costing us 2350$ fuel margin last month alone. recommend contacting maintenance contractor. - station X has seen increased queue time last month resulting in 54 missed refuel (avg 4500$ missed). recommend adding one pump. 4. from insights to autonomy. … this is the part I’ll keep to ourselves for now. so excited about what we’re building.

@peytonspencer @heydittoai 🎯🔥🔥

That's @heydittoai

@heydittoai 🔥🔥🔥

The competition starts Monday. Memory. Tool use. Latency. Measured, optimized, and rewarded on SN118. Ready your agents. Pull the starter kit. Let’s mine. bittensor:native

@AltcoinMillie @Raleigh_CA ✊🏾🔥🔥 let’s goo

Me and @Raleigh_CA just dropped a banger podcast for the Bittensor community get ready you don't wanna miss this $TAO

@chutes_ai @markjeffrey 🔥🔥🔥

Jon Durbin is going on the Hash Rate Podcast with @markjeffrey. On the table: Parallax, decentralized inference, and what it takes to train models across distributed compute. Live Friday, July 10. What should Mark ask Jon? Drop it below and we'll send him the best ones.