𝐖𝐫𝐢𝐭𝐢𝐧𝐠 𝐜𝐨𝐝𝐞 𝐚𝐧𝐝 𝐜𝐚𝐫𝐫𝐲𝐢𝐧𝐠 𝐚 𝐬𝐲𝐬𝐭𝐞𝐦 𝐚𝐫𝐞 𝐧𝐨𝐭 𝐭𝐡𝐞 𝐬𝐚𝐦𝐞 𝐭𝐡𝐢𝐧𝐠 🤯

𝐖𝐫𝐢𝐭𝐢𝐧𝐠 𝐜𝐨𝐝𝐞 𝐚𝐧𝐝 𝐜𝐚𝐫𝐫𝐲𝐢𝐧𝐠 𝐚 𝐬𝐲𝐬𝐭𝐞𝐦 𝐚𝐫𝐞 𝐧𝐨𝐭 𝐭𝐡𝐞 𝐬𝐚𝐦𝐞 𝐭𝐡𝐢𝐧𝐠 🤯

That distinction stayed with me while reading recent announcements from IBM, Intel and Thomson Reuters.

The same technology, agentic AI for software development, is being told through very different stories: one of greater engineering capacity, another of a smaller, AI-native workforce.

𝐓𝐡𝐞 𝐭𝐞𝐜𝐡𝐧𝐨𝐥𝐨𝐠𝐲 𝐦𝐚𝐭𝐭𝐞𝐫𝐬. 𝐒𝐨 𝐝𝐨𝐞𝐬 𝐰𝐡𝐨 𝐢𝐬 𝐬𝐡𝐚𝐩𝐢𝐧𝐠 𝐭𝐡𝐞 𝐯𝐢𝐬𝐢𝐨𝐧 𝐚𝐫𝐨𝐮𝐧𝐝 𝐢𝐭.

There are at least three reasons I would want experienced engineers, including people who built systems before AI, at the centre of that work.

➡️they know what good looks like. Not code that merely compiles, but code that can be maintained, secured and understood six months later. They are less likely to mistake AI’s confidence for evidence.

➡️ they ask the questions that shape the reasoning. When an agent proposes a change to a service, an experienced engineer does not begin with “does it work?” They ask: does it respect the interface contract? What happens on rollback? Which exception path is still used by a legacy consumer? Where is the evidence that this is safe to release?

Those questions change the output.

➡️they recognise the smell of a system becoming unhealthy. An authentication change, a retry policy and an old batch window can look unrelated until they are not. Human experience remains unusually good at connecting those dots before the incident makes the connection obvious.

AI can accelerate implementation. It can also make poor assumptions travel faster.

The value of experienced engineering is not resistance to AI. It is the judgment that turns capability into a system that can survive reality.

𝑊ℎ𝑜 𝑖𝑠 𝑠ℎ𝑎𝑝𝑖𝑛𝑔 𝑡ℎ𝑒 𝑠𝑡𝑜𝑟𝑦 𝑦𝑜𝑢𝑟 𝑜𝑟𝑔𝑎𝑛𝑖𝑠𝑎𝑡𝑖𝑜𝑛 𝑖𝑠 𝑡𝑒𝑙𝑙𝑖𝑛𝑔 𝑖𝑡𝑠𝑒𝑙𝑓 𝑎𝑏𝑜𝑢𝑡 𝐴𝐼?

#SoftwareEngineering #EnterpriseAI #TechnologyLeadership