Correlation ≠ Causation https://t.co/fUI7ryxfXe
Correlation ≠ Causation https://t.co/fUI7ryxfXe
Projects that launched with InfoFi.
Most of the altcoins launched last year going to 0. https://t.co/AP28CEvO6G
Agree and NOVA checks both boxes.
- Our new scoring function builds on Boltz-2 and achieved AUC 0.81 for hit id.
- We do this with unmatched efficiency by running on #Bittensor's decentralized network.
Subnet 68 @metanova_labs is building infrastructure to decentralize drug discovery.
Not a lab. Not a single drug. A system.
The current system fails at scale:
• >$2B and ~10 years to bring one drug to market
• 1/10k molecules survive the traditional wet-lab funnel
• Most capital is burned validating compounds that fail late
MetaNova’s approach is structural, not incremental.
Through NOVA Subnet 68 discovery is run as a continuous, incentive-driven system:
▸ 65+ billion molecule search space
▸ Billions of molecules screened in production
▸ 10,000s of mammalian protein targets
▸ 2 live incentive layers:
• NOVA Compound (molecules)
• NOVA Blueprint (search algorithms)
This isn’t crowdsourcing.
It’s adversarial optimization.
Miners are explicitly incentivized to:
• explore high-variance chemical regions
• exploit model failure modes
• surface false positives and blind spots
Every exploit becomes training signal.
Every failure improves robustness.
That pressure already yields measurable output:
▸ Integr