AI Can Write Senior-Level Code.
Originally published on LinkedIn
𝗔𝗜 𝗖𝗮𝗻 𝗪𝗿𝗶𝘁𝗲 𝗦𝗲𝗻𝗶𝗼𝗿-𝗟𝗲𝘃𝗲𝗹 𝗖𝗼𝗱𝗲. 𝗧𝗵𝗮𝘁 𝗗𝗼𝗲𝘀𝗻'𝘁 𝗠𝗮𝗸𝗲 𝗬𝗼𝘂 𝗮 𝗦𝗲𝗻𝗶𝗼𝗿 𝗘𝗻𝗴𝗶𝗻𝗲𝗲𝗿.
Engineers have a challenge and a huge opportunity with AI-generated code.
How they respond will define both their career and the success of the engineering teams they are apart of.
I'll be frank: I have zero tolerance for engineers using AI to generate code they don't understand, haven't properly validated, and then push to production.
💥 That's not engineering.
This isn't just a criticism, it's highlighting one of the biggest challenges in our industry right now: how we use AI to accelerate delivery without sacrificing quality and development of engineers.
We still need people to learn, gain experience, and grow into the senior engineers who will lead the next generation of systems and teams.
Let's take a common scenario.
An engineer is assigned a ticket they couldn't have completed manually. It's beyond their current experience level.
The 𝗮𝗯𝘀𝗼𝗹𝘂𝘁𝗲 𝘄𝗿𝗼𝗻𝗴 approach is obvious:
❌ Ask AI to generate the solution.
❌ Skim the code.
❌ Convince yourself it probably works.
❌ Ship it.
❌ Pray.
Unfortunately, that approach is becoming increasingly common.
The real question is:
🤔 𝗛𝗼𝘄 𝗱𝗼 𝘄𝗲 𝘂𝗽𝘀𝗸𝗶𝗹𝗹 𝗶𝗻 𝗮𝗻 𝗔𝗜-𝗮𝘀𝘀𝗶𝘀𝘁𝗲𝗱 𝘄𝗼𝗿𝗹𝗱?
Before AI, if you were given work beyond your experience level, you had a few options:
✅ Hand the ticket to a more experienced engineer.
✅ Pair with someone who has done it before and learn from them.
✅ Research the problem yourself, invest time learning, and deliver more slowly.
None of those options disappeared when AI arrived, you even have more. Use AI as a teacher, ask it to explain its decisions. Challenge its design choices. Compare multiple approaches. Learn the concepts behind the implementation.
𝗨𝘀𝗲 𝗶𝘁 𝘁𝗼 𝗮𝗰𝗰𝗲𝗹𝗲𝗿𝗮𝘁𝗲 𝘂𝗻𝗱𝗲𝗿𝘀𝘁𝗮𝗻𝗱𝗶𝗻𝗴, 𝗻𝗼𝘁 𝘁𝗼 𝗯𝘆𝗽𝗮𝘀𝘀 𝗶𝘁.
If you consistently rely on AI to do work you don't understand, you're not building experience, you're 𝗿𝗲𝗻𝘁𝗶𝗻𝗴 𝗶𝘁.
And rented experience doesn't create the next generation of senior engineers.
The engineers who thrive over the next decade won't be the ones who can generate the most code. They'll be the ones who can understand it, evaluate it, improve it, and take responsibility for what gets shipped.
For engineers under pressure to deliver, its tempting to hit the magic AI button and move on.
⚠️ Push back.
For engineering managers, support your teams in doing that.
The goal isn't to maximise code output it's to maximize quality output, maintaining high engineering standards.
🎯 AI should accelerate learning.
𝗜𝘁 𝘀𝗵𝗼𝘂𝗹𝗱 𝗻𝗲𝘃𝗲𝗿 𝗿𝗲𝗽𝗹𝗮𝗰𝗲 𝗶𝘁.
How are you helping engineers develop expertise while using AI?
#SoftwareEngineering #EngineeringLeadership #TechLeadership #AICoding #SoftwareArchitecture #EngineeringManagement #DeveloperExperience #CareerGrowth #ContinuousLearning #FutureOfWork #SeniorEngineer #TechCareers