August 1, according to 36 Kryptonian, Cognition Engineer, former Meta EngineerMape WonIN INTERVIEWS, IT WAS NOTED THAT SILICON VALLEY AI ENTREPRENEURSHIP HAD ENTERED A PHASE OF RAPID CONSENSUS FAILURE. THE REDUCTION IN MIDDLE-LEVEL MANAGEMENT AND THE INCREASE IN THE NUMBER OF GENERAL-LEVEL ENGINEERS ARE NOT JUST ORGANIZATIONAL FLATTENING SLOGANS, BUT PART OF THE MANAGEMENT WAS REDIVIDED AFTER AI BEGAN TO ASSUME RESPONSIBILITY FOR INFORMATION COLLATION, SYNCHRONIZATION OF PROGRESS AND CROSS-GROUP COLLABORATION。
He is particularly opposed to using Token consumption to measure whether an engineer is "AI native". Public rankings induce employees to run empty for numbers, leaving only higher bills; enterprises need to compare AI spending with how much time they spend on work time and business results. Cognition therefore tries to convert every conversation from Devin to an "Equivalent Engineer Hours" and reconnect abstract models to productivity。
According to Marpeon, model worship is also declining. It would be more feasible to move cheap and front-line models along mission lines, while adding sandboxes, detectability and output monitoring to intelligent bodies. The opportunity for vertical intelligence comes from complex processes that cannot be completed at a time: products must continue to absorb the details of the tax, legal and other industries at the customer ' s site, and these sedimentation does not automatically disappear as the generic model upgrades。
In his judgement, such companies were not necessarily able to replicate the valuation multipliers of base model companies, but were still able to rely on continued delivery for a stable return. Entrepreneurs need to avoid the path that has become a consensus and find processes where common models are difficult to directly annex and clients are willing to pay for long-term solutions。