@signulll So I agree that society needs this idea. But to me it's an idea in which value accrues at the societal layer, and won't be captured by Neuman's startup beyond being first - and then every developer does it.
@signulll What stops any developer from doing this? Network effect is purely local - literally within the building itself. Having building 1 doesn't make building 2 better at all. Nothing here scales.
@evisdrenova Parallel reads within each dir—thousands of blobs vs just 3 directories. But for repeated bulk reads, packfiles (what Git does): concat objects into one indexed file. Trades write simplicity for way faster sequential reads.
@jasonlk Does it blind them, or does it simply reveal that most decision makers have no way to measure objective facts themselves and so they default to a herd heuristic.
@jasonlk Simple product in which user = buyer: do outbound. Complex product with distributed buying decision makers and unclear path: don’t do outbound.
@evisdrenova They’re a niche enterprise software vendor with ~30 customers selling to oil companies and utilities. The ‘AI platform’ branding is aspirational.
@evisdrenova They’ve rebranded 3 times in 15 years (C3 Energy → C3 IoT → C3 AI) chasing whatever’s hot, have only $300M revenue after 15+ years, and one customer is 35% of revenue.
@evisdrenova “Rubicon Project will start solving the sharp pain point of optimizing ad networks for publishers and will leverage this position to be the trusted platform to help monetize all inventory for a publisher – the Control HQ for all revenue for web publishers”.