Dan Hockenmaier

<start of post>

Everyone is asking the wrong question about AI in the workplace.

The question isn’t 'will AI replace jobs?'

The question is: who will build the systems that actually get work done end-to-end?

Right now, most companies are buying 'tools'.

But the winners are going to buy (or build) 'workers'.

And there’s a big difference.

Tools need you to already know what to do.

Workers figure out what to do, using your goals and constraints, and then they go do it.

CONTEXT: knowing what matters inside your company (people, priorities, rules, data, risk).

AGENCY: being able to take action across systems without someone babysitting every click.

Legacy software vendors have a huge head start on context.

They already sit on your CRM, your ticketing system, your docs, your calendars, your permissions.

They know who owns what.

They know what changed.

They know which customers are 'red accounts' and which deals are about to slip.

But most of them still ship software that feels like a cockpit.

Powerful, but you still have to fly the plane.

AI startups have the opposite advantage.

They can ship agency from day 1.

They can book meetings, draft emails, update fields, open tickets, reconcile invoices, route approvals.

They can do the work.

What 'good' looks like for your company

Which exceptions matter and which don’t

Who is allowed to override a policy (and when)

The weird edge cases everyone learns after 6 months on the job

The unspoken priorities (this quarter: retention > new logos)

An agent that has enough context to make good decisions.

And enough agency to execute them without constant supervision.

Can it proactively notice a problem and fix it?

Or does it just wait for you to tell it what to do next?

Here’s a concrete example.

Imagine you run a marketplace.

A big seller is about to churn.

Their order volume dropped 35% week over week

Their top SKU is out of stock

Their support tickets are spiking

Their payouts were delayed last month

The account owner is on PTO

open an internal incident

message the right ops person

offer the seller a credit (within policy)

update the CRM notes

schedule a follow up for next week

and send the exec team a quick summary

Most products can do maybe 2 of those things.

With a human filling the gaps.

This is why 'AI features' won’t be enough.

A chatbot bolted onto a dashboard is not a worker.

And a worker that can click buttons but doesn’t understand your business is just chaos at scale.

So who wins?

Do the incumbents move fast enough to ship real agency?

Or do the startups earn context fast enough to become trusted?

My bet is we get a new category of winners: companies that start in a narrow workflow, go insanely deep on context, and then expand outward once agency is safe.

Not 'horizontal AI'.

Not 'another SaaS dashboard'.

Workflow by workflow. System by system. Until the agent can actually run the job.

Curious what you’re seeing inside your company.

Where do you think the first true end-to-end AI workers show up: sales, support, finance, recruiting, or something else?

<end of post>