It’s very clear AI writes great code and will even review it for bugs much better and faster than a human can. With cloud agents like Devin, you can spawn an army of these agents at a moments notice. Teams who figure out how to do this will simply out compete any teams who don’t.
Now humans operate at a higher level of abstraction. Our focus area is now ensuring the high-level architecture will accommodate future needs well, ensuring the final product meets requirements, and most importantly, ensuring the final product has been validated. It’s important to use every strategy in the book to test the output via unit tests, smoke tests, integration tests, and end to end tests. On our team, we’ve been investing a lot in setting up full test environments that include the entire stack at a level simply unachievable before AI. Now we can merge code changes at an unprecedented velocity without losing confidence in the system.
Be careful, these statements are aging faster than milk. Most skeptics of even a few months ago suddenly got very quiet. New models are getting very advanced; the pace is stupefying. LLMs absolutely can and do write much better code today than the majority of senior software engineers, or at least at comparable levels. If you're not seeing that, you either aren't using the greatest models, your harness needs some work, or maybe you specialize in some very niche domains where LLMs are legitimately helpless.
jonator · · focus · HN ↗
Now humans operate at a higher level of abstraction. Our focus area is now ensuring the high-level architecture will accommodate future needs well, ensuring the final product meets requirements, and most importantly, ensuring the final product has been validated. It’s important to use every strategy in the book to test the output via unit tests, smoke tests, integration tests, and end to end tests. On our team, we’ve been investing a lot in setting up full test environments that include the entire stack at a level simply unachievable before AI. Now we can merge code changes at an unprecedented velocity without losing confidence in the system.
pydry · · focus · HN ↗
It's very clear they dont. Maybe they will one day. That day isnt today.
jonator · · focus · HN ↗
iLemming · · focus · HN ↗
Be careful, these statements are aging faster than milk. Most skeptics of even a few months ago suddenly got very quiet. New models are getting very advanced; the pace is stupefying. LLMs absolutely can and do write much better code today than the majority of senior software engineers, or at least at comparable levels. If you're not seeing that, you either aren't using the greatest models, your harness needs some work, or maybe you specialize in some very niche domains where LLMs are legitimately helpless.