Relationships

Here’s an analogy that only someone in Silicon Valley could come up with.

Data is king when training models. ML architectural choices that won were those that allowed models to successfully scale with data. High quality, low volume data also trumps high volume, low quality data.

Similarly, relationships will be king when building companies. Good company decisions will be directionally aligned with improving these relationships. Think hiring, distribution, fundraising.

This has always been true, but will become far more true this decade as models get better at cognitive tasks.