Hands-On AI Is Here

I think we just crossed an important line.

Not because a model got a little better.

Not because another company added AI to a product page.

The bigger signal is that OpenAI and Anthropic are moving closer to the actual work.

OpenAI is reportedly moving forward with The Deployment Company, a new joint venture backed by 19 investors, including TPG, Brookfield, Advent, Bain Capital, SoftBank, and Dragoneer. Bloomberg reported that the venture has raised more than $4 billion and is valued at $10 billion before the new money.

Anthropic announced a new AI services company with Blackstone, Hellman & Friedman, and Goldman Sachs. Reporting from Fortune and the Financial Times via Investing.com puts the backing at roughly $1.5 billion.

That is not just another enterprise partnership.

That is hands-on AI.

The model companies are not only saying, “Here is the tool. Good luck.”

They are saying, “We will finance, staff, and scale the work of putting this into the business.”

I think that is a huge deal.

The bottleneck moved

For the last couple years, the obvious question was model capability.

Which model is better?

Which one can code?

Which one can reason?

Which one is cheaper, faster, safer, easier to integrate?

Those questions still matter. But I think the bottleneck is moving.

The hard part is no longer only whether the model can do useful work. The hard part is getting that work into a company without turning the company into a mess.

That means understanding the workflow. It means permissions, data access, evals, reliability, handoffs, approvals, security, training, change management, and maintenance.

It means sitting close enough to the work to know where the time actually disappears.

That is why these announcements matter. OpenAI and Anthropic are both pointing at the same thing from different angles.

This is not only a product launch.

This is capital formation around AI deployment.

AI adoption is becoming an implementation problem, and the model labs are building vehicles to own more of that implementation layer.

Consulting is not dead, but it is changing

I do not think this means every consultant is done.

That is too simple.

There will still be room for people helping businesses use AI, especially smaller businesses that are never going to have a direct line to OpenAI, Anthropic, Accenture, PwC, or one of these private-equity-backed AI services firms.

The local business still needs someone who can show up, understand how the business works, clean up the process, pick the right tools, set up the workflows, and make sure the owner is not left with a shiny demo nobody uses.

But the bar is going up.

The person who “knows ChatGPT” is not enough.

The valuable person will be the one who can take a messy business process and turn it into a working AI system. Not a slide deck. Not a prompt pack. A real system that saves time, reduces friction, and keeps working after the first week.

That is a different skill set.

It is part software engineering, part operations, part product thinking, part training, part trust-building.

This is going to create a lot of jobs

I think this wave creates jobs.

Not vague “AI jobs.”

Actual hands-on deployment jobs.

Companies are going to need people who can:

• Map workflows and find the right places for agents.
• Connect AI systems to existing tools and data.
• Build evals so teams know whether the system is working.
• Set permission boundaries so AI does not become a security problem.
• Train teams to supervise agents instead of just using chatbots.
• Maintain systems as models change every few weeks.
• Turn one successful pilot into a repeatable deployment.

Some of those jobs will have familiar titles. Some will look like forward-deployed engineering. Some will look like solutions architecture. Some will look like internal tools. Some will look like operations people who got very good at AI.

The title matters less than the posture.

The valuable people will be close to the work.

They will be able to watch how a team actually operates, identify the bottleneck, build the system, verify that it works, and keep improving it.

That is going to be a serious category of work.

It is also going to displace work

The honest version is that this will also displace jobs.

I do not think it is useful to pretend otherwise.

If AI services teams are going directly into companies to automate documentation, coding, compliance reviews, reporting, customer support, internal operations, research, and back-office workflows, some work is going to disappear.

Usually the tasks disappear first.

Then roles get rebundled.

Then teams get reorganized.

Some companies will use AI to do more with the same number of people.

Some will use it to do the same work with fewer people.

Some will move faster and hire more because the new tools unlock work they could not do before.

All of those things can be true at the same time.

That is why the fear version is too small and the hype version is too shallow.

This is not just “AI takes jobs.”

It is also “AI changes what the job is.”

The people who can adapt into the deployment layer will have opportunities. The people who ignore the shift because it still feels early are taking a real risk.

The next few weeks are going to feel strange

I do not mean the whole economy changes in two weeks.

But I do think the next few weeks are going to feel strange.

Announcements like this create permission.

A board sees Anthropic, Blackstone, Goldman Sachs, and Hellman & Friedman forming an AI services company backed by roughly $1.5 billion, and the question changes from “Should we experiment with AI?” to “What is our AI deployment plan?”

A CTO sees OpenAI reportedly building a $10 billion Deployment Company around AI adoption, on top of launching Codex Labs and partnering with global systems integrators, and the question changes from “Should developers try Codex?” to “How do we roll this into real engineering work?”

That shift matters.

Once large companies believe the deployment motion is real, the market moves fast. More partnerships show up. More case studies show up. More executives start asking for plans. More teams are asked to find workflows worth automating. More workers realize the tools are not staying in the sandbox.

This is where things can start to feel crazy.

Not because everyone loses their job tomorrow.

Because the default expectation changes.

AI stops being an experiment on the side and starts becoming part of the operating model.

What I would do right now

If you work in software, operations, support, finance, marketing, sales, HR, compliance, or almost any knowledge-work function, I think the move is simple:

Get closer to the tools.

Not in a passive way.

Do not just read the announcements. Do not just try the chatbot. Do not just wait for your company to tell you what the AI process is.

Pick a real workflow and try to improve it.

Take something you do every week and ask:

What parts are repetitive?

What context does the work need?

What decisions require human judgment?

What could an agent draft, check, research, route, or execute?

How would I know if the output was good?

What would make this unsafe or annoying?

That is the muscle.

The future job market is not only going to reward people who can use AI. It is going to reward people who can deploy AI into real work responsibly.

That means taste. It means judgment. It means technical ability. It means understanding people well enough to build systems they will actually use.

The point

I do not think these announcements are just about OpenAI and Anthropic wanting more enterprise revenue.

Of course they want enterprise revenue.

But the larger point is that the AI market is moving from access to deployment.

The first wave was: here are powerful models.

The next wave is: here is how those models become work.

That is going to create a ton of opportunity. It is also going to create pressure. It will make some teams faster, some jobs different, and some jobs unnecessary.

I do not think the right response is panic.

I also do not think the right response is casual optimism.

The right response is to pay attention, get hands-on, and start building the skills that sit between AI capability and real business value.

Because that middle layer is where a lot of the action is about to be.

Strap in.