mrbreo

Tech Entrepreneur || Founder of @morcastnet
Building at the Intersection of Decentralized AI & Web3 Infrastructure

[mrbreo @mrbreo]

@shardiban Are you guys focusing on @oroagents ?

[THOR Group @THORMachines]

The people helping robots learn should be rewarded for their work, whether collecting, annotating or verifying training data.

Open networks and crypto incentives can help make that happen at scale.

Thanks @0xSammy for the breakdown.

Full report via @KhalaResearch

[Swarm @SwarmSubnet]

SWARM quarterly update to stakeholders.

A year ago, SWARM started with a thesis: open competition on Bittensor builds better autonomy than a closed lab, and a physical lab turns that autonomy into products people pay for.

The thesis now has customers. Two contracts signed in two months.

Time to lock in. 🧵

[mrbreo @mrbreo]

@SwarmSubnet @Ever_BlueGreen Wonderful!!!

[mrbreo @mrbreo]

@LeadpoetAI Awesome, Poet!

[Leadpoet @LeadpoetAI]

We just killed cold outreach.

Introducing Leadpoet: the AI sales agent that finds buyers who need you now, writes personalized outreach, and follows up on LinkedIn and email.

Used by Dropbox, OnePatch (YC26), Muttdata.

Comment “Poet” for 100 free lead credits to find qualified buyers.

[mrbreo @mrbreo]

@SwarmSubnet

[mrbreo @mrbreo]

@DucCryptoX @LeadpoetAI Alright

[mrbreo @mrbreo]

Most AI projects are still searching for product-market fit.

@LeadpoetAI (SN71) is going after something businesses already spend real money on: customer acquisition.

Leadpoet V2 launches tomorrow, Oct 20.

What matters is that this is no longer just a lead-finding tool. It is moving toward a more complete revenue

Here is the workflow:
Buying Signals => Qualified Prospects => Personalized Outreach => Follow-Ups => Qualified Meetings => Revenue => Alpha Buyback
That is the real product shift.

V2 expands the system from simple lead discovery into a more operational sales engine:
Detect real buying signals
=> Identify qualified decision-makers
=> Generate personalized outreach
=> Execute email and LinkedIn sequences
=> Automate follow-ups and track replies

A simple way to think about it:
V1: Find potential buyers
V2: Start conversations and move them toward conversion
That is a much stronger commercial proposition.
And the economics are important.

Leadpoet already has a SaaS model and real customer-facing use cases. V2 increases the value captured per customer because it moves closer to the part of the workflow that actually generates revenue.

Better workflow coverage = higher product value = stronger monetization potential
Now add Bittensor.

SN71 is not just another AI SaaS wrapper. It is built on decentralized competition.

Miners compete to build better sales-research agents.
Validators measure output quality.
Rewards go to measurable performance improvements.

So the flywheel looks like this:
Better agents => Better prospecting => Better outreach quality => More qualified meetings => Higher customer ROI => Stronger product demand

That is the part people should pay attention to.
The key metric is not just how many leads are generated.

It is:
- Lower cost per qualified meeting
- Higher conversion into real pipeline
- More revenue generated for customers

If Leadpoet V2 improves those numbers, then SN71 becomes much more than a subnet story.
It becomes a real example of Bittensor producing commercially useful AI.
This is the kind of real-world AI application I want to see emerging from Bittensor.

$TAO #Bittensor #SN71 #Leadpoet #AgenticAI #SalesAI #DeAI

[mrbreo @mrbreo]

This is why setting expectations upfront matters so much. A simple agreement on drop-off, pickup time, and whether parents are staying can prevent most of these awkward situations.

Hospitality is always appreciated, but respecting someone else's time and boundaries is just as important.

[mrbreo @mrbreo]

@DucCryptoX Dmed.

[mrbreo @mrbreo]

@gavinzaentz HUGE 🔥

[Gavin @gavinzaentz]

Leadpoet V2 drops tomorrow.

V1 found the buyer.
V2 starts the conversation.

The signal. The person. The first message. The follow-up. The reply.

All of it, in one place.

[THOR Group @THORMachines]

Shenzhen is the home for hardware AI

[mrbreo @mrbreo]

@jollygreenmoney Could you share a bit more about this?

[THOR Group @THORMachines]

The embodied AI data market is forecast to grow from an estimated $536M in 2025 to $4.9B by 2032. Roughly 9x in seven years.

But volume alone isn't enough. A demonstration recorded for one robot arm can be useless to a lab training a different one. Labs need data cut into individual actions, labelled with language instructions and checked for consistency, all to their own spec.

THOR starts with what the lab needs, then turns those requirements into paid annotation and verification tasks for providers on Bittensor. Outputs are checked against the lab’s specifications.

We're building so labs can scale their data operations without coordinating every provider themselves.

[mrbreo @mrbreo]

Robotics is approaching the same inflection point that LLMs experienced with large-scale pretraining.
But there is a fundamental difference.
LLMs had the internet. Robots don't have an equivalent dataset of physical interactions, manipulation trajectories, contact dynamics, and failure recovery.

That is why Bittensor's @THORMachines (SN112) caught my attention.

THOR is building a decentralized market for robotics training data, starting with temporal action annotation of egocentric video.

The architecture is interesting:
- Robotics labs define data requirements and quality thresholds.
- Miners compete to produce structured training data.
- Validators evaluate submissions against quality criteria and hidden reference samples.
- Bittensor emissions incentivize accepted contributions.
- Verified datasets become reusable infrastructure for robotics research.

The bigger opportunity is connecting this to Vision-Language-Action (VLA) models, imitation learning, simulation, and eventually real-world deployment.
But the critical metric won't be dataset volume.
It will be whether independently produced data improves downstream robot task success at a lower total cost.

If THOR can demonstrate that, Bittensor could become a meaningful coordination and incentive layer for physical AI.

Time to all ears and eyes on that.

bittensor:native #Bittensor #SN112 #THOR #Robotics #PhysicalAI #DeAI

[MSB Intel @MSBIntel]

Grayscale's Bittensor Trust had 3,247,300 shares outstanding on October 13, up 14.25% from 2,842,300 on September 15. Its latest new shares were issued for TAO worth $1,451,029.

@opentensor

[mrbreo @mrbreo]

@wolfwammer @Tom_A_Lynch He is in sleeping mode for Bittensor now. xD

[mrbreo @mrbreo]

@yellowpantherx @playcambria @cyberpunk You should hop on board!
You have got some killer skills, and you would be a great fit for this network.
https://t.co/3QzgJAx7cP