AI Is No Longer Just Learning to Think. It Is Learning to Work.

[#1]

AI is developing fast, but I think the most important change is not that models are getting “smarter”.

It is that AI is moving from answering questions to actually doing work.

We started with:

Human → Prompt → Answer.

Then came copilots:

Human → AI assistance → Human action.

Now we are moving toward:

Human → Goal → Agent → Tools → Actions → Result.

And the Atlassian ecosystem is a good example of what that means in practice.

With Jira, JSM, Confluence and Rovo, AI is no longer only about summarizing text or drafting a reply. Agents can increasingly use organizational knowledge, interact with work items, support automation and participate in real workflows.

That changes the discussion completely.

A chatbot giving a wrong answer is annoying.

An AI agent taking a wrong action in Jira, changing content in Confluence, or acting on service-management data is a governance, security and accountability problem.

So the future of enterprise AI will not be won only by the company with the “best model”.

It will depend on:
• permissions
• identity and access
• reliable knowledge
• clean data
• APIs and integrations
• auditability
• human approval
• clear ownership

This is also why I do not think AI makes experienced admins, consultants or platform owners less relevant.

If your Jira permissions are chaotic, AI will not fix that automatically.

If your Confluence knowledge base is outdated, AI can simply surface outdated information faster.

If nobody understands the process, adding an agent does not suddenly create accountability.

AI amplifies the environment you give it.

And that may be the biggest shift for IT professionals.

Less time spent on repetitive execution.

More time spent on architecture, governance, process design, integrations, risk and deciding what should be automated at all.

That is why I think the real AI skill is not “prompt engineering”.

It is:

Domain expertise + systems thinking + AI literacy.

The human role is slowly moving from:

Executor → Collaborator → Orchestrator.

For people working with Atlassian platforms, this could become very visible over the next few years.

We may no longer only administer tools used by humans.

We may increasingly administer environments where humans and AI agents work together.

And that leads to the question I find much more interesting than “Will AI replace my job?”:

What should my job become when AI can do more of the execution?

[Viswanathan Ramachandran]

Hi @Arkadiusz Wroblewski 

Will AI replace my job?”

What should my job become when AI can do more of the execution?

It's a fair question. I would acknowledge and agree that execution heavy roles are changing. Denying this creates distrust.

Before AI

80% of current day was spent on manual execution, we started writing codes/simple programming like excel macros to a massive code, that core routine changed dramatically. We evolved. 

With AI

When AI handles the execution like drafting code, generating baseline designs, writing initial reports, human value moves to two distinct places: upstream (problem framing, strategy, intent) and downstream (judgment, context, ethics, and accountability). We will evolve. 

The job doesn't disappear - it evolves from Operator to Director.

In summary

• AI answers the questions. AI is for scale, speed, and patterns.
• Humans are for intent, judgment, and connection. Human have the vision and intelligence. 

[Arkadiusz Wroblewski]

@Viswanathan Ramachandran  That's the reason I think AI will cut some Jobs and maybe have some bad impact, but overall will allow us faster development then even before 

[Viswanathan Ramachandran]

It is worrying as it is cited as reason for job lay-offs. All we have is to acknowledge and accept the truth, like a hard-pill to swallow. But develop and outgrow ourselves. Instead of fighting with the waves, swim with the waves. It's tough ride.