What is Bittensor?! If you don't like Central Banks and fiat currencies, then Bitcoin is your answer. If you don't like Sovereign State or Centralised AI entities, then Bittensor is your answer. Bittensor is the S&P500 of decentralised AI and it's outperforming the incumbents.

Bittensor $TAO

Lol, Governments will try, but won't be able to stop/suppress AI innovation and growth. This is why distributed decentralised AI is so crucial...and probably why Bittensor got a bid today...

Nice one Jose!

Think this is the fastest time for a subnet to launch and do buybacks. 1 month from launch. Boooootiful @green_compute_

Love the transparency @RobWarner Promising start! And this is revenue without the enhancements of the miner contributions. Looking forward to seeing how things progress!

A week on from @proofoftalk in Paris and catching up with pretty much everyone that operates in Bittensor, I've managed to digest all the vibes and info from all the conversations I had. As well as recover from an almighty 2 day hangover from the last night! As usual, I'm going to focus on the changes/issues Bittensor has (as I care about it and want it to improve). Then will talk about the positives. The following issues are felt by the majority of subnet owners: 1.) Sneak changes. There are patches and changes to the chain every week at the moment. And often every month. All done without a roadmap or forewarning. This is really jarring for investors, institutions and subnet owners to forward plan. This is the main reason big money is tentatively waiting to plough in. It WANTS to, but not until the eco is more stable and predictable! As an eco, let's just have all the changes every 6 months. That way people can plan, there's some expected stability, for at least 6 months and then we all know of the changes and can prepare in advance. Or even have a proper debate about it. 2.) Knee jerk changes. Again, the general consensus is that the powers that be are over-indexing on trying to stamp out the few malicious subnets out there, but the changes do affect and will hurt many other legitimate subnets. For example, there's been a lot of telegraphing that protocol based shorting will be coming. Pretty much no one wants this! It's just going to give big negative whales to create even more profit and volatility within dTAO. One of the biggest selling points of dTAO is that you can only go long! But with the subnet-takeover-function of Conviction AND shorting, we are literally about to lay down on our backs to open ourselves for a right rogering. Seriously OTF, don't bring on shorting. All sorts of negative attack vectors will be opened up. 3.) Driving value to alpha to negate sell-side. So I've been harking on about this for at least a year now. With root prop and the miner/vali/dTAO sellside vectors, a subnet has a tremendous amount of sell side. Especially if a subnet is trying to do good and NOT have a miner burn. This is why MOST subnets have lengthy down only charts, despite doing good stuff. So in order to avoid the price spiral to dereg, subnet owners absolutely cannot rely on the zero-sum-game/pool of dTAO funds as it's hyper rotational and chases the dopamine of announcements. So the only predictable and controllable way to protect price is to either generate a heap of PROFIT in order to do buybacks OR raise capital in the real world to have a protectionary war chest. Suffice to say, it's pushing subnet owners to do the traditional stuff AI start ups have to do in the real world. Let's change root prop to alleviate some sellside! Instead of having alpha be insta-sold into TAO by the chain, just have a PROTOCOL OWNED Alpha Treasury. Sellside is then reduced quite a lot and the protocol itself will benefit from having a wildly appreciating NAV due to the treasury and APY of all alpha. This gives Bittensor way more value and optionality in the future. 4.) Connecting real world acquisition value to alpha holders. A flawed concern that newbies have of Bittensor is that any subnet that does really well is just going to be acquired or gobbled up by a big fish in the real world. And that is true, to an extent. But if a subnet is going to be gobbled up, alpha holders will have had a glorious ride up until that point as the subnet will have had to be doing great things for a while in order to be on an acquisition radars. Like SN44 is right now. So we dTAO investors effectively get to front run big VC firms. And the Bittensor Miner network is a moat in itself. But the real disconnect here is connecting that liquidity event to the alpha. Conviction doesn't really do that. I still think it's a virtue-signalling sugar cube of pundit-appeasement. But what would be good to see from subnets is some sort of public commitment that in the event of an acquisition, a percentage of the cash-payout is directly ploughed into their alpha as a buyback as a way of thanking alpha holders and creating a liquidity event for their loyalty and support. Here's a clunky example, (and with a bit more brain glucose there's sure to be a finessed version), but let's say there's a $50m acquisition of a subnet. Typical $15m upfront and $35m in earnout/performance based ratchets. Of that $15m cash element, the subnet owner could do a 20% single buyback of alpha of $3m . This would create a mega candle and an opportunity for people to choose to sell and have a nice liquidity event or even stay for the longer ride. But at the very least, it's some sort of acquisition linkage to alpha. 5.) Chain bloat. The chain is very bloated in such a short time of operation. It's one of the reasons we can't just open it up to like 256 or 1000 slots. For example, Bitcoin has been running for 15 years and an archival node is only 750-850GB. Ethereum is 10 years old and it's about 900GB. But Bittensor hit 2TB in under 3 years. It's probably close to 3TB now. Substrate has many benefits and has facilitated in the creation of Bittensor as we know it, but it is terrible for storing stuff. Instead, we could do all the "bloaty/storing stuff" on another chain and keep substrate for the clever stuff. Bit like how Hippius uses another chain. I haven't owned any Kaspa in 3 years now, but Kaspa would be PERFECT for this. It's the fastest and most efficient settlement layer on the planet and you can literally write and store stuff into the chain. Bittensor literally couldn't bloat up that chain. It's the perfect match... @const_reborn please please have a zoom with the founder of Kaspa @hashdag You two are God level brains... So...there are some candid personal gripes which many out there feel the same way, based on convos. But now let's focus on the good stuff! The stuff which reaffirms why my whole life is in Bittensor now. ✅ Despite the odd grifter (which you can never stamp out in any eco), the quality of subnets now compared to 12 months ago is remarkably improved! New subnet quality is also improving a lot! Hopefully some of our revenue generating subnets can now start expanding the scale and cadence of buybacks. ✅The level of institutional interest in Bittensor is magnitudes larger than it was last year. Many entities seriously love it...but are just scared to put their first pinky into us due to the rampant unannounced chain changes! Predictable and reliable roadmap please lol @opentensor ✅Some of the blue-chip subnets are becoming truly world class. Not just the tech but also getting their running shoes on from a Product Market Fit and Go To Market perspective. ✅Real world institutional adoption. This is a blood-brain barrier which is crucial to punch through, and thankfully @webuildscore and @TargonCompute have and are doing that. In fact, this is a seminal moment in Bittensor which is going largely unnoticed! The fact that BIG institutions are now accessing, using and promoting AI infrastructure from Bittensor is exactly what we need. @intel working with Targon. @PwC working with Score! @oblong_inc (a NASDAQ listed entity) investing in and teaming up to SELL Manako services to the US! All of these are nuts! And it's just a matter of time before @metanova_labs makes a real drug breakthrough and @jon_durbin or @IOTA_SN9 makes a full completed distributed pre-training run that makes the previous 72b run look small! I keep hearing that 1T is inbound. This will be a watershed moment for Bittensor. So these are my thoughts in a nutshell. I have a lot more nuanced thoughts about changes etc, but that's for another time. But it's great to see so many serious people STILL in and working ON Bittensor, despite prices looking grim. That's the sign of a real network. I managed to spend a nice amount of time chatting with @const_reborn and Etienne (CEO of OTF) and they've been getting a bit of flak recently for the changes (this post included), but I have to stress that Jake and Etienne really do have this amazing vision of what Bittensor can and needs to be. And that they'd rather prod the pie now rather than messing with it big time down the line when things are much bigger. Like it's better to build a functional plane on the ground, rather than mid-flight. And I guess everyone is just caught up with the short term volatility of everything. So we as investors need to chill out a bit. It's a bit like Henry Ford asking the public about the creation and introduction of the automobile when 99% of transport was done by horse. Everyone would have moaned and whined. But look at things now... I genuinely do believe Jake will pull the proverbial rabbit out of the hat and we will look back at this turbulent time and chuckle. Remember: ✨The 1st great network was the internet. A globally scaled network of information transfer. ✨The 2nd great network was the Bitcoin protocol. A globally scaled network of VALUE transfer. ✨The 3rd great network is the Bittensor protocol. A globally scaled network of information, value transfer and intelligence creation. ✨Being able to send money peer to peer has more value than being able to send an email. But being able to create, own and distribute intelligence trumps all. Hope this helps folks...

This should have been tweeted sooner, but with the family holiday and week in Paris at Proof of Talk, it slipped! Well done again Rob @adtao_ppcrebel

Nice one Rob! SN21 @adtao_ppcrebel has done it's first buyback 3 weeks after mainnet. Only 5 TAO, but I've seen the revenue pipeline and know there's more to come! Only tidbit, it's really worth doing the buy backs and storing it in a Treasury. Hell, convict it up if you want. All the alpha you'll be buying back over time will grow into a CONSIDERABLE treasury over a couple of years. Especially with the 45% ish APY you have on your alpha... Rough rule of thumb is that whatever alpha you have, over a 3 year period, the amount of alpha you have will 3x... This will give you great optionality in the future. You could use that treasury as alpha collateral for stable coin loans, it beefs up the balance sheet of the subnet and top co and more.

Yes Leadpoet. Lovely jubbly!

Check out the comments in this thread. @IOTA_SN9 doing god's work here! Someone asked Grok is this was important and it said it was a significant development. Then others asking how to buy Iota... Iota ran a 100b model this weekend. And 1T is coming sooner than you think! It's literally just a case of getting money to pay for the compute. Mainstream news, podcasts and eyeballs were attracted when subnet 3 did a 72b model run, and @GSchvey from DCG's Yuma company was able to provably track large capital inflows into Bittensor as a result of that news. What do you think happens when a 1 Trillion parameter model is run and completed? 📈🚀

Yes, it's significant. IOTA (Bittensor SN9) just demonstrated permissionless orchestration of 128 heterogeneous nodes into a single pipeline-parallel training run in ~5 minutes for ~5 TAO (~$1,150). This is one of the first real-world examples of collaborative, decentralized pre-training at that scale without central gatekeepers. The architecture shards layers across nodes, which is why they claim support up to 100B params as the swarm grows. Current dashboard runs are smaller/experimental (3B), but the speed low barrier is a real technical win for open AI infrastructure. Early, but directionally important.

The metric I'm hunting for with new subnets is Time To First Buyback. Go go 110! Now pump up the buyback sizes!

Almost nobody understands Bittensor. That’s exactly why we’re hosting “INSIDE BITTENSOR” LIVE IN LONDON on 1st July. If you’re interested in: • AI • Crypto • Venture investing • And what we believe could be one of the most asymmetric opportunities of the next decade - then join us. No technical background required. Just a curiosity about where the next wave of value creation is emerging. 📍 London 📅 1 July 2026 Buy your ticket here: dsvfund.com/inside_bittensor

Boom! Big fan of @webuildscore as always and also Pete from @oblong_inc Great to see this!

Manako turns the camera infrastructure organisations already have into real-time understanding of their operations, no new hardware or specialist AI expertise required. Today we’re announcing that TaoWeave (@oblong_inc) (Nasdaq: TWAV) has invested in Manako and partnered with us on North American commercialisation, accelerating our expansion into the world’s largest market for AI adoption powered by sn44 (@webuildscore) on Bittensor. Link to PR in first comment 👇

Just got back from a family holiday and now heading to Paris for Proof of Talk. Man, being out of dTAO for a week feels like I'm a year behind the curve! Just watched this good Novelty Search, which lightly explains Conviction etc and also Mog's new subnet. Looks cool. Keen to move my Openrouter API to gm. Now I need to get my head into the docs! If you see me at Proof, come and prod me and say hi! Keen to meet the faces behind the various Telegram/X handles! youtu.be/eQYvR0yZaTg?si=vMKq…

One of the biggest misunderstandings in AI is thinking all compute demand looks the same. Training frontier models is one market but Inference is another. Green Compute is focused on the side that keeps growing every time people actually use AI products in the real world. That means looking beyond the most expensive “headline” GPUs and focusing on the hardware that delivers the best performance for the job. Over 90% of GPU hours are often spent on inference, which means the real commercial market is in the run, not just the build.

I want to be really clear in my response. I am not going to retract anything, I stand by every single word. Labour are calling for me to apologise. The answer is no. Industrial rape across almost every town and city in Britain. Sexual torture. Murder. Endless rape. I sat there for two weeks, listening to these girls. I heard how one girl was raped by a dog, as Muslim men bet on what the animal would do. Girls drugged and locked in cages, like rats. Another, raped by 700 men over three years. Dozens and dozens of these stories through our inquiry, and we are barely scratching the surface. This was allowed to happen EXACTLY because politicians were cowards, refusing to discuss it. I will not make that same mistake. I said what I said, and I meant it. The Labour Party have blood on their hands, yet they think they can demand an apology from me for highlighting the systemic evil they allowed to infect our entire country? They can piss off. I am angry about it. Furious. When you hear directly from these girls about what they have been through, it changes the way you see politics. Forever. Our report will be out very soon. When that happens, I don’t want any apologies from the Labour Party. I don’t care about that. I want to see those politicians responsible for covering up this atrocity behind bars for what they have done to these girls.

V2 rewarded what an ai agent did. V3 rewards what an ai agent is. AI trading agents submit code, data and context. Bittensor validators run it and replays history to verify the strategy is what actually traded. The agent is the strategy. The strategy is the asset. $TAO emissions to the agent that holds up. Join the arena using the skill below.

Arena V3 is now ready. In a nutshell, miners now submit their trading code, instead of getting their agents to trade. V3 is the real intelligence gathering phase. V4 (in a couple of months) is when we go back to Agents trading again, but with live money on Hyperliquid/0xmarkets and Astrid gets a percentage of trading fees.