The deal eventually fell apart, but what matters is that they wanted to do it in the first place.
Last week, Anthropic made an interesting acquisition: they bought Bun, the JavaScript runtime. Bun is open source and has an MIT license. Anthropic could have split it and built on top of it for free. They have Claude Code, an excellent tool for writing code.
Instead, they bought the company. Because they wanted Jarred Sumner and his team.
This is what I keep coming back to every time I see another one “Programming is dead” message goes viral. The companies building AI, the ones that supposedly know exactly what it can and cannot do, are spending billions to acquire tech talent. Don’t fire them, acquire them.
If OpenAI believed that GPT could replace software engineers, why wouldn’t they build their own VS Code fork for a fraction of that cost? If Anthropic thought Claude could do the job, why make an acquisition at all?
Programming is not the job
Here’s my opinion: AI can replace most of the programming, but programming is not the job.
Programming is a task. It’s one of the many things you do as part of your job. But if you’re a software engineer, your actual job is more than typing code into an editor.
The mistake people make is that they confuse the task with the role. It’s like saying calculators have replaced accountants. Calculators automated arithmetic, but arithmetic was never the job. The job involved understanding financial matters, advising clients, making decisions, and so on. The calculator only made accountants faster in the mechanical field.
AI does something similar for us.
What the job is
Think about what you actually do in a given week.
You’re in a meeting where someone is describing a vague problem, and you’re the one figuring out what he or she actually needs. You look at a codebase and decide which parts to change and which to leave alone. You push back a feature request because you know it will create technical debt that will haunt the team for years. You look at a colleague’s PR and discover a subtle bug that could have disrupted production. You call whether you want to ship now or wait for more tests.
None of that is programming, it’s all your job.
Some concerns
I’m not going to pretend that nothing has changed.
Will some companies use AI as an excuse to reduce headcount? Absolute. Some have already done that. Layoffs will be attributed “AI efficiency gains” which are really just cost savings dressed up as something else.
But think about who stays and who goes in that scenario. It’s not random. The engineers who understand that programming isn’t the job, the ones who bring judgment, context, and the ability to figure out what to build, those are the ones who stay. Those who only brought code output may be at risk
A common concern is that juniors are being left behind. If AI handles the “doing” part, how do they build judgment? I actually think the opposite is true. AI compresses the feedback loop. What used to take days of browsing books or waiting for Stack Overflow replies now takes seconds. The best juniors don’t skip steps, but get through them faster.
Now think about your own situation. Let’s say you were hired two years ago, before the current AI wave. Your company wanted you. They saw value in what you bring. Now, with AI tools, you are considerable more productive. You ship faster. You can deal with more complexity. You are better at your job than ever before.
“You have become much more productive, so we are letting you go” is not a sentence that has much meaning.
What to do about it
If you’re reading this, you’re already thinking about these kinds of things. That puts you ahead. Here’s how to stay there:
- Get started with AI tools. Find out what they are actually useful for. Find out where they save you time and where they waste time. The engineers who do this now will be at the forefront.
- Practice the non-programming parts. Judgment, trade-offs, understanding requirements, communicating with stakeholders. These skills are now more important, not less.
- Build things from start to finish. The better you understand the full picture, from requirements to implementation and maintenance, the harder it is to replace.
- Document your impact, not your output. Frame your work in terms of problems solved, not in terms of code written.
- Stay curious and not defensive. The engineers who will struggle are those who see AI as a threat to defend against, rather than as a tool to master.
The form of work is changing: some tasks that used to take hours now take minutes, some skills are less important, others more so.
But different isn’t dead. The engineers who will thrive understand that their value was never in typing, but in thinking, in knowing what problems to solve, in making the right tradeoffs, in shipping software that actually helps people.
OpenAI and Anthropic could build their own tools. They have the best AI in the world. Instead they spend billions on engineers. That should tell you something.
#write #code #job