Standard data models for business software

Being able to build, maintain, monitor, execute, or use business software is a pillar for intelligence AND artificial intelligence. Things like CRMs, ATS, Payroll system, billing systems are more than just a social construct. I think they're solving fundamental problems, and not just human problems. I think if robots ruled the world they would also need to build similar systems.

We need to inject our knowledge of business software into artificial intelligence.

Unfortunately, this knowledge is not easy to get from unstructured data sources. Because traditionally business software lives in relational databases, and these relational databases are rarely exposed on the web. And even if they were, it's very hard to use these data models because they all have different schemas.

This is why I think that the most impactful way to accelerate AI is to have companies agree on a standard data model they use accross all the software they produce.

We're already seeing initiatives like that but they're too low level: think about open telemetry for example.

The last point is also the reason why everything will go open-source: because open standards become ubiquitous.