Without Systems Thinking, AI won’t deliver the gains you expect.
It’s tempting to throw GenAI at the problem - but applied in isolation, it'll likely deliver little. Worse, it could take you in the wrong direction faster and exacerbate existing problems.
https://blog.robbowley.net/2025/04/07/without-systems-thinking-ai-wont-deliver-the-gains-you-expect/
1/5 🧵
It’s tempting to throw GenAI at the problem - but applied in isolation, it'll likely deliver little. Worse, it could take you in the wrong direction faster and exacerbate existing problems.
https://blog.robbowley.net/2025/04/07/without-systems-thinking-ai-wont-deliver-the-gains-you-expect/
1/5 🧵
Comments
Fixing them is, at best, like taking painkillers instead of treating the underlying illness.
2/5
3/5
A big risk as it makes it even more alluring to get busier with busywork. Automating failure demand.
4/5
Rather than playing whack-a-mole patching problems, you’ll uncover opportunities to make changes that deliver real, lasting impact
5/5
Inside the system every interaction makes sense and has the engineering wow factor. It’s only when we step back and see the whole system and all the interactions that we see the waste.
But GenAI’s strongest use cases so far often pull in the other direction.
The use cases in the table seem vague and scattershot.
Targeted AI deployments need to integrate well with existing software systems incl. ERP suites even if the org has no dedicated eng.
(Links in alt)