I like how you measured AI's value beyond just productivity. The most interesting examples are where AI helped uncover the real problem instead of simply answering the original question. The cat food smell and fragrance examples are perfect illustrations of that. Sometimes the biggest benefit isn't saving hours; it's making better decisions, avoiding unnecessary purchases, and discovering solutions you wouldn't have considered yourself. That's where AI becomes genuinely useful in everyday life.
Former Java engineer turned Ruby engineer who is trying to understand Ruby and Rails, MacOS and a lot of other things. Worked at Flywheel, FNBO, ACI Worldwide.
They were both truly helpful there! And I really do love my new fragrances, which is wild to think that it was AI that really helped me figure out I would like them.
I’m **Cody Simons**, the writer behind **DEV-C**, sharing ideas and interesting articles on different topics.
I enjoy exploring technology, lifestyle, travel, tips, inspiration. So Good Luck.
Civic technologist building AI tools that help systems explain themselves. Clew Suite builder, NACM 2026 speaker. I write about hackathons, product judgment, agent readiness, and learning in public.
"The useful ones end somewhere else". It seems, to me, there's always something else to connect! 🤣 Start a chat here, finish the purchase or calendar update or booking there. Still not a seamless experience exactly but once we get there. Watch out!
Reviewing actual chats instead of counting prompts is a useful way to evaluate the workflow. How did you distinguish time saved from cases where verification or correcting the answer consumed the apparent gain?
Former Java engineer turned Ruby engineer who is trying to understand Ruby and Rails, MacOS and a lot of other things. Worked at Flywheel, FNBO, ACI Worldwide.
I didn't really, It really was just me eyeballing it, thinking about about much time it takes me normally. Super not quantitative, but just an estimate of time. And you are right, in some cases it really didn't save time...and might have been a time suck in itself.
A rough estimate is still a useful starting hypothesis. The next lightweight step is to sample a handful of tasks and record elapsed time, rework, and whether the output was usable without rescue. That separates "felt faster" from work merely moved into review.
We're a place where coders share, stay up-to-date and grow their careers.
I like how you measured AI's value beyond just productivity. The most interesting examples are where AI helped uncover the real problem instead of simply answering the original question. The cat food smell and fragrance examples are perfect illustrations of that. Sometimes the biggest benefit isn't saving hours; it's making better decisions, avoiding unnecessary purchases, and discovering solutions you wouldn't have considered yourself. That's where AI becomes genuinely useful in everyday life.
They were both truly helpful there! And I really do love my new fragrances, which is wild to think that it was AI that really helped me figure out I would like them.
The cat food cabinet bit is the best argument in here. A chat that points past the question to the source is doing real work.
yeah that's right
That one stuck with me too. A pointer to the source beats a tidy answer that stops at the question.
"The useful ones end somewhere else". It seems, to me, there's always something else to connect! 🤣 Start a chat here, finish the purchase or calendar update or booking there. Still not a seamless experience exactly but once we get there. Watch out!
Reviewing actual chats instead of counting prompts is a useful way to evaluate the workflow. How did you distinguish time saved from cases where verification or correcting the answer consumed the apparent gain?
I didn't really, It really was just me eyeballing it, thinking about about much time it takes me normally. Super not quantitative, but just an estimate of time. And you are right, in some cases it really didn't save time...and might have been a time suck in itself.
A rough estimate is still a useful starting hypothesis. The next lightweight step is to sample a handful of tasks and record elapsed time, rework, and whether the output was usable without rescue. That separates "felt faster" from work merely moved into review.