Language models had the internet for text
physical AI needs spatial data and lab collection just doesn't scale @axisrobotics
solves this with a compounding data engine
> browser simulation -> teleoperate robot arms with zero hardware required
> mobile capture -> turn daily human actions into real-world datasets
> post-training fixes -> step in when policies slip to feed high-value corrections back into the loop
every trajectory is verified and logged on-chain via Base for full cryptographic provenance
we aren't just watching robotics evolve we're actively training tomorrow's physical AI
what task are you training on @axisrobotics today??
$AXIS tokenomics are finally out @axisrobotics
but the number I kept looking at wasnโt the 1B supply or the 16.5% TGE unlock
the contributor rewards part was more interesting to me
10% of the total supply goes to contributors, but itโs not really structured like one big airdrop
4% is for Season 1 / pre-TGE simulation
and only half of that unlocks at TGE
so at launch, only 2% of the total supply is actually unlocking on the contributor side
the other 2% comes at T+3 and T+6, while the bigger 6% allocation is saved for Season 2 / post-TGE sim-to-real
I think that split matters
Axis doesnโt just need people doing tasks until TGE
if the data engine is supposed to keep working, people still need a reason to keep producing new data after launch too
this is basically the loop Iโve talked about before:
collect data โ train โ find new failures โ collect better data
and now the tokenomics are starting to show how they want to incentivize that loop
another detail I liked is that team and private investors get 0 unlock at TGE
their 31.67% allocation starts after a 12 month cliff, then unlocks linearly over 24 months
so the 165M AXIS entering circulation at TGE is mostly coming from ecosystem and liquidity instead
the part Iโll probably watch the closest though is the 51.08% Ecosystem allocation
โ61% goes to the networkโ sounds great as a headline, but most of that is sitting inside a pretty broad ecosystem bucket
where those tokens actually go over time, what they fund, and what kind of real contribution gets rewarded will matter more than the headline percentage itself
for now, the part I like is that most of the contributor allocation is being saved for future data production instead of just rewarding past activity
for a data network, that makes a lot more sense to me
https://t.co/GvODyJjo9A