So I've been full-time independent for 12 years now. In general, I don't think AI is a major driver of employee vs contractor decision but we'll see.

The market has been trending towards specialists for a long time. AI may help employees in the short term be more effective generalists, and so be able to compete with specialists. AI may help specialists be even more effective in their niche, while also serve wider needs, and so compete better with employees.

Something I do see happening is companies are doing a lot of low hanging fruit themselves in my space (I do revenue and business analytics). Today, they will get 80% of my specialty done themselves. That is enough for most companies. But that last 20% for those who want it, still requires a specialist like me who knows the domain entirely.

So far I've noticed that the actual consulting firms aren't fairing as well. They have to get really lean and focus less on being a body shop and more focused on the high-value delivery -- on what can be very short projects.

I've noticed, however, freelancers do quite well. A lot of this is personality driven. People build good personal brands and they do well, firms want those people and will pay good $$ for them.

In your estimation what other notable aspects separates a freelancer from a consulting firm?

ML(not AI), will be able to automate a lot of simple repetitive tasks with predictable outcome. That will indeed eliminate some stupid-simple jobs that require little to no thinking and are mostly just types of jobs that just have to be done but can be done by anyone. So low level white collar jobs. There will be bleed into different sectors as well but nothing of significance.

I would say no, ML will not have impact on freelancing whatsoever. It will have impact on the bottom line in some companies, but that's like self serving kiosks in mcdonalds. It will have about as much impact as that.

It’s going to be a great time to be an entrepreneur, but a terrible time to be an employee.

A solo indie developer can have an AI team working on a project.

But so will every corporation. The competition for the remaining human-only roles will be intense.

Coding has never been the hard part. Creating something that people will pay more for is.

I will bet a whole paycheck that if you took 10 “solo indie developers” and 10 employees, the employees would be both working less hours per week on average and making more money

The last line is really quite tragic. The implications for the future for many people will be harsh in this scenario

The current job market is definitely increasing the amount of freelancers and indie developers.

The software crisis of the 1960s was marked by an inflection point: Organizations saw the benefit of automation but software writing was a tedious affair typically done in assembly language or FORTRAN or something. There just weren't enough programmers on earth to take up the load of writing all the software that would be necessary. So new tools, like COBOL and ALGOL, were devised to help programmers produce correct software quickly.

Today's software crisis is not one of too little but too much. We are absolutely spoiled for computing power -- a smartphone having enough capacity to replace a mainframe that in the 1970s or 1980s would have handled a national bank's transactions, many times over. We are awash in software, most of it bad. We need less software and better software. Stochastic slop generators are going to make this problem worse, not better.

It may be a rough few years, but on the other side there will be a boom in demand for programmers to clean up the mess "AI" has made.

They might use AI to clean up old AI tech debt code

No I think it will create more full time roles, either (pessimistic) cleaning up AI slop, or (optimistic) opening up work that would've been uneconomical before.

This is something I hope will happen. I can see small dev shops being able to do things like big migrations on legacy code they couldn’t contemplate before. I’m not so optimistic on the creation of new jobs though.

Why wouldn’t AI handle the work of the contractors?

For the time being, AI doesn’t have good access to a lot of things: customers, information in 3rd party systems, etc. Even when we give access (with MCP, for example), current AI has trouble locating the right information to solve a problem.

So I still need an engineer to find out why the distributed system isn’t performing correctly, or to create a new component that is coherent with the rest of the existing UI. The difference is that any particular engineer will find more success tackling a wider range of goals.

Lots of people don't dare do things in production, even the ones with some other specialized technical experience.