Unitree R1 will be more popular than G1
Here’s a video of R1 racing faster than G1
R1 weighs 25kg vs G1’s 35kg
Less mass = faster acceleration
But price is what will make R1 more popular
R1 costs $5k, G1 costs $13.5k
These prices are for the normal versions
If you actually want to do development
G1 EDU costs $67k
That is not affordable for most
Startups, universities, researchers
Yet R1 EDU costs $15k to $30k
The more add-ons, the more expensive
But still way more affordable than a G1
1/ You can now sign a transaction before it's valid.
Not optimistically. Not with a trusted relayer crossing their fingers.
On-chain enforced, future-conditional execution — with a real simulation behind it.
That's what ERC-8211 predicates make possible. 🧵
Season 5 is now underway. ⏳
Starting on May 05, Season 5 introduced a new 7-week cycle running until June 22, 2026.
Complete tasks.
Contribute data.
Earn $VADER
Base Batches 003 Robotics Track by @virtuals_io has officially concluded at @ns.
After 8 weeks of pushing robotics onchain, here are the winners:
🏆 Champion — @opengotchi
🥈 1st Runner-Up — @shadowcleague
🥉 2nd Runner-Up — @Vader_AI_
The future of robotics is on @base.
Hundreds of industrial workers die every year doing routine tasks. 👷
🔴 Turning valves
🔴 Reading gauges
🔴 Taking samples
🔴 Climbing ladders
Not because of negligence. Because the job requires a human body to be somewhere it shouldn't have to be. 🧑🚒 The video below is our robot doing exactly those tasks, via teleoperation, with the operator safely remote. 🥽 No one should die doing a job a robot can do. And we’ll do whatever it takes to minimize deaths from industrial accidents. 🫡
Thank you to @base & @virtuals_io teams for organizing Base Batches 003!
Vader was chosen as the 2nd runner up!🥉
We had the opportunity to
i) get priceless feedback from experienced judges
ii) get to spend weeks with likeminded robotics builders
iii) work with a dedicated G1
Physical AI is not built from data alone.
Raw demonstrations are only the beginning.
To turn physical data into useful models, teams need training workflows designed for embodied AI:
🌎 World Action Models
👁️ Vision-Language-Action models
☑️ Task-specific policy networks
Transactions that wait for their funds.
You sign once. The batch sits on-chain, dormant.
When the money lands, it fires itself — approve, swap, deliver, atomic.
No polling. No second signature. No app open. Gated by on-chain state, not by you being online.
A few use cases:
→ Fiat onramp → Any token
→ CEX withdrawal → strategy same block
→ Authorization rails for AI agents that act, but never hold keys