Before a partner promises a business anything, Ditto checks it against 95 features. Here's the check running on a made-up 12-person landscaping company: – Fit first – Which features are actually live – Questions for the first call – Clear next step The check shipped this week, as part of our partner program.

Your root stake, reborn as buybacks - our validator reinvests its commission in subnet alpha 🔄 Stake with @bitstarterAI, make a difference for new teams @adtao_ppcrebel @provenonce_ai @DeSciClaims @cascade_sn91 @carbonphysicsai

1/ ASMR Research Preview Bittensor subnets in aggregate (the YCX index) have substantially lower volatility than the S&P 500. The likely driver is large-cap subnet tokens having greater liquidity (quoted in $TAO) relative to their size than many S&P 500 index members. Market discussions often start with the assumption that more liquidity is better. Yuma’s research, recently presented at @Exploitsummit, demonstrates how excess liquidity can suppress price discovery, obscuring quality and risk information from capital allocators. Key slides from the presentation follow in this thread. This work is shared as research for public discussion and it explores a solution we’re calling ASMR: Alpha Subnet Market Reloaded.

FIRST BUSINESS CLIENT Less than two weeks after V2 launched, Ditto has its first business client! A multi-six-figure company, and the first custom build Ditto is doing. Like a lot of growing businesses, they run on a mix of apps that don't talk to each other, and every question still ends with a call to the owner. Ditto becomes the one place it all runs through. Agents take over the day-to-day work, and every member of staff gets what they need from the Ditto mobile app, under their own permissions. Know a business that runs like this? Our partner program opens soon, and partners earn commission on every client they bring to Ditto. Our business development team is already on it. Revenue is our focus now, because it's what fuels our buyback and burn program.

Seeing so many retail traders getting constantly burned with trading young/newly launched subnets. There's a bunch of different telegram groups full of "New subnet launched, get in early" type bollocks. Everyone is trying to take knowledge shortcuts and instead jumping into any new shiny object because someone on a chat group said they have insider knowledge that a is b and x worked at y and is launching z. It's just pure gambling folks, and no one is calling it out. I doubt any meaningful amount of people have got rich through these telegram groups and blindly aping into whatever bag someone is trying to pump. Bigger money actually use these groups as inverse sentiment indicators. I know I do. Subnet 46 was the latest scammy pump n dump. Dodgy behaviour + Chinese whispers = pain. I literally got about 4 DMs from people saying they had the inside scoop on something massive about to launch on 46...I called BS straight away. Bittensor really is something special, but people will still find a way to ruin their bags by gambling on immature/premature subnets. 99% of the time, the price ends up tanking. Avoid the knowledge shortcuts. I've explain several times in videos how subnet pricing works, the various glass ceilings there are and high Expected Value bands that are best to operate in. This is why I NEVER touch a subnet that: ✖️ Is less than 45 days old. ✖️ Has less than 1500 TAO in it's LP. ✖️ Hasn't price tanked yet. ✖️ Hasn't tanked and formed a stable base. ✖️ Tanked in emissions, formed a stable base and there's a nice gradual growth in the emissions chart. ✖️ Isn't clear what the subnet actually is/does/doesn't have literature explaining their subnet. I mean these are basic basic filters folks. It's enough to avoid most portfolio ending moves. Yes you will miss out on the 1% of subnets that launch and just go up to begin with... But they always tank eventually once this retail echo chamber has run out of buying power. And it's the alpha heavy participants like miners that end up just dumping into that initial retail strength. Don't be the exit liquidity! Again, anyone talking about subnet market cap/FDV doesn't really get the dynamics. Those metrics are utterly useless in dTAO. Market cap = Price x circ supply. The narrow band of price that subnets move around in and circ supply is almost negligible. It's the price of TAO that is the main vector in subnet market cap because subnets are priced in TAO. If TAO does a 10x overnight, guess what, all subnets will 10X in market cap overnight. If TAO tanks 50% overnight, same thing, subnet market cap tanks 50%. There's a subnet that all the chat groups are frothing over at the moment but no one knows what it actually does. The chart looks like someone took some viagra, but it will tank. TAO is going to do epic things. It will 10-20 X easy. You don't need to chase 100Xs with gambling on newer subnets. There are dozens of amazing subnets with great teams, great products/services and real potential. Buy them! Rant over... hopefully it saves at least 1 person from blowing up!

With Ditto: You can use smarter models in your agent. You can use cheaper models in your agent, or turn on Auto and let Ditto pick the model for each message. You can use a different model for every workflow step. You can switch models in the middle of a conversation, and the history comes with you. You can choose which apps and chats your agent draws on. You can delete anything it remembers, or chat without saving at all. You can run it while your laptop is closed. You can reach it from Slack, Discord, your phone, and the AI tools you already use. You can choose the model, how hard it thinks, and the instructions it follows. You can see every tool it has, and turn off the ones you don't want. You can teach it something once and never explain it again. You can see what it did, and when. You can read the exact prompt your coding agent sends through the Ditto Router. You can export your memory as a signed file and check it yourself. You can read and fork the memory harness. It's MIT. Your memory is yours.

One month after mainnet launch, we asked ourselves: How does Claims (SN111) compare with the other Bittensor subnets? We built a dataset to measure it - on-chain data, Taostats, GitHub, X, websites, podcasts. Here are the five measures where we stand out. Data and code are public, so you can check 👇

Miners on Subnet 118 compete to build a better memory harness. They have maxed the benchmark within hours of us publishing it, and we moved the version the same day. That is not the benchmark failing. That is the research lab working, and it is why retrieval improves here faster than it would inside one company. The article explains how the subnet and the product fit together. Most people have only ever seen one half.

Carbon is live! Subnet 116. Physics AI is now on Bittensor Let’s go!!!