Former Uber CEO: AGI Has Not Arrived, Change Management is Key
Former Uber CEO Travis Kalanick stated that AGI has not yet arrived, claiming that those who assert it has are "foolish." He noted that while current AI agents have made progress, their intelligence level remains limited.
He emphasized that the challenge of transforming large enterprises into "autonomous enterprises" lies in change management: dealing with middle management, technocrats, and undocumented complex processes, which is fundamentally a human organizational issue rather than a technical one.
Source: Public Information
ABAB AI Insight
Kalanick's assessment hits the core bottleneck of AI implementation: technological maturity is secondary to organizational inertia. Large company processes are often "black box legacies," and AI needs to reverse-engineer this tacit knowledge, while resistance to change comes from vested interests, which is more time-consuming than model training.
The concept of "autonomous enterprises" points to a shift from human coordination to agent autonomy, but practice has shown that the speed is far slower than theory. The truly agile are tech companies with a "pro-AI culture," which accelerate deployment through developer culture, while traditional enterprises are mired in hierarchy and lack of documentation.
This divergence reinforces the two poles of the industry: AI-native companies capture productivity dividends, while legacy enterprises gradually become marginalized. The narrative claiming AGI overlooks the limitations of agents' "narrow intelligence," with actual value lying in incremental efficiency rather than disruption.
In the long run, change management will become a core service track in the AI economy; those who master "human + agent" hybrid governance will dominate enterprise transformation profits, rather than merely selling models.