Pay only for measurable model uplift. Earn when your data improves models. Proof-of-Data rails with on-chain origin, use, improvement & contribution.
Joined December 2023
- Tweets 521
- Following 25
- Followers 38,538
- Likes 3
246 Photos and videos
Pinned Tweet
22 Dec 2025
Syncora → Strova 🟠
New name. Same mission: prove what data actually improves AI, and pay the people who contributed it.
Derived from "strong" "prova" (Italian for proof) 💪🔐
We're live → strova.xyz
Built on @base ⛓️
#Strova #ProofOfData #AIData #Web3 #DePIN #Base
32
11
149
27,289
The missing link:
Proof of Origin
Proof of Use
Proof of Improvement
Proof of Contribution
129
models don’t lie.
they:
train
evaluate
improve
stall
regress
performance is feedback.
measure the cause.
#ProofOfData #AIData #Web3 #DePIN #Base
1
83
Massive bet on hardware.
The next bet should be
measurable data contribution.
Without it, compounding slows.
Feb 27
Tech giants are burning $2 trillion on physical infrastructure before the AI revenue even exists.
It is a massive physical bet.
Will it pay off?
148
2023: “Just scrape it.”
2024: “Just fine-tune it.”
2025: “Just add more compute.”
2026: Prove which data worked.
#ProofOfData #AIData #DePIN #Base
1
1
86
Meta revisiting stablecoins is a strong signal.
When platforms at that scale lean into on-chain settlement,
it’s no longer experimental, it’s infrastructure.
The next question:
what else should settle on-chain?
#Meta #Stablecoin
Feb 24
⚡️ LATEST: Meta is reportedly planning a stablecoin comeback in the second half of 2026 with third-party payment integration and a new wallet.
1
161
deleting this tweet in 5 mins if it doesn’t hit 1k likes
but real talk:
proof-of-data > “trust me it worked”
93
Most AI platforms incentivize more data.
We enforce better data.
– Uploading isn’t enough
– Usage is recorded
– Improvement is measured
– Payment requires proof
As models get smarter,
accountability tightens.
Scarcity won’t be compute.
It’ll be verified contribution.
96
if your friends aren’t talking about:
Ai agents running 24/7
prediction markets printing
looksmaxxing protocols
creatine gains
proof-of-data rails
time to find new friends
#Proofofdata #AI
156
When your data trains the model
but nobody credits you.
158
AI models are worth billions.
The people who trained them?
Often invisible.
No attribution.
No impact measurement.
No performance-based payout.
That’s not innovation.
That’s extraction.
#ProofOfData #AIData #Web3 #DePIN #Base
1
73
g
gm
gme
gmea
gmeas
gmeasu
gmeasur
gmeasure
gmeasure
gmeasure impact
Strova.
120
AI training has a measurement problem.
Not the one people think.
Let’s separate what we KNOW
from what we DON’T KNOW
about AI training data today.
🧵👇
1
93
6/7
Who deserves what share?
If 12 contributors improved a model by 10%,
how do you fairly distribute $11,200?
Flat splits?
Heuristics?
Gut feeling?
That’s not infrastructure.
That’s convenience.
Why This Gap Matters
AI is moving into regulated domains:
Healthcare.
Finance.
Legal.
Robotics.
In these domains, “trust us” isn’t enough.
1
53
AI teams after a successful training run with zero attribution.
111