"Preparing For The Future" - in his article, Marty Cagan describes what's ahead of product teams thanks to generative AI. I cannot recommend it more also for engineering leaders.
Here are key takeaways:
- Because more things will be generated for engineers, their overall cognitive load should go down. It'll lead to leaner product teams and/or broader product responsibilities.
- The more we get from generative AI, the less we'll outsource (If you can explain your needs to an external company, you can explain it to ChatGPT).
- Generative AI won't take engineers' jobs. While the amount of boilerplate can be reduced, engineers will still play a vital role in aligning technical solutions with the company's unique vision, mission, and strategy, making it a full-time creative job.
- Generative AI will produce artifacts, but critical thinking and overall judgment will remain within engineering teams.
Will generative AI be a giant leap in software engineering? Yes and no, in my opinion.
Yes, because it'll reduce a lot of tedious and repeatable tasks. But it's likely the same level of significance as:
- all kinds of SDLC automations (building, testing, deployment, etc.),
- PaaS or SaaS (AWS, GCP),
- Open Source (all kinds of libraries we add to our code),
- better hardware (M1 chips or cloud machines), and others.
Each of these constantly evolves, changing weeks of our work into hours.
Generative AI, we'll be just another (yet powerful) tool in this collection. This means all creative work, tinkering in code, debugging, and architecting systems is still on us. 🧑💻
Sure, AI will become more "intelligent" over time, and much of our work will be done for us. But it'll only mean we can build even more sophisticated solutions within a shorter time. That's it. This loop will never end.