I don't know if my job will still exist in ten years

> I don’t think there are any genuinely new capabilities that AI agents would need in order to take my job. They’d just have to get better and more reliable at doing the things they can already do.

The "just" here is minimizing what has been the crux of the problem for the past ~5 years.

This technology has been capable of producing code all this time. The end result has been improving due to massive scaling efforts, and some relatively trivial engineering ("reasoning", "agents", etc.).

And yet reliability is still a massive problem. The tools still hallucinate, still lead the user in dead-end directions, and still do so confidently and randomly, without any discernible reason. Expert users are able to guide them to a certain extent, but whether the prompting incantation is done manually or via the trendy Markdown file of the week, it's all guesswork based on feelings and anecdata.

I'm personally not too worried about being replaced by these tools, even though my skillset is nothing remarkable. My opportunities might shrink, but this is a two-way street. Companies that use "AI" indiscriminately don't interest me either. The demand for quality human work and ingenuity will always exist, even within a sea of mediocrity.

I'm much more concerned with the societal impact of the mountains of shoddy software being produced, deployed into increasingly more critical infrastructure, and put into hands of incompetent and malicious people. There is very little thought and discussion on this topic, let alone any guardrails. "AI" companies are now attracting governments and advertisers, both full of malicious and incompetent people. The next decade is going to be interesting, that's for sure.

Let me be the optimist here: Programmers are both extremely unproductive and know very little about everything else. What percentage of capable programmers can actually sit down and write code for 8 hours and then do it 5 days in a row? The time not coding is filled with many different words for recovery. Some of this recovery involves trying to better understand the domain. (more on this later)

Most of the startups that get the attention are attempting to be the next big thing but a startup can just be a startup. It doesn't have to be big or glorious.

Someone who sells hot dogs (on a small scale) cant really hire a programmer but if he could (or could write it themselves) there would be plenty of software to write. You can make a nice interface with all of the sales statistics, inventory management, maps with competition and demographic data, work schedules, etc etc There is infinite complexity to even the simplest job. You could hire help and have an app talk them though every step in great detail with pictures, videos and animations. You can encode all of the little tricks that could normally take decades to learn. Say, on a busy spot you might not have time to spend 8 minutes properly cleaning the grills every 47 minutes but you could wipe down the glass every 4 minutes and clean part of the grills with alcohol every 11 minutes then clean it properly every 3 hours. The app might instruct to google location related news or other topics to talk about with the customer. If people are walking their dog they expect you to guess what breed it is and where it comes from then ask how old it is.

You might build a tech stack to help recognize their face, remember their name, what they ordered last time and how long ago. You're not suppose to but you know you want to.

You wont be coding for a glorious salary but will earn depending on the sector your chose. The software will be pure dog food of the finest quality in the world.

Grilling hot dogs is also very relaxing, can let the mind float a bit and have software ideas the way you should. Lots of bad ideas will come, I can show people pictures of themselves eating my hotdogs!

You can basically look at programming as the new literacy. You might want a fancy job writing letters for a nobleman but it is hardly the only application.

I'm not yet convinced that the positions are going away. Unfortunately, I find it more likely that it's going to be a bad few years and then surge in demand to fix things that have been done in those years.

Maybe I've just had bad luck but over last 2 decades I've only worked in places where 80% (at least - probably closer to 95%) coworkers (in development related areas) had negative productivity - making software more complex, brittle and abstract than necessary. With AI assistants the same people can be more "productive" and gatekeeping is mostly a lost cause.

I hope to find a decent alternative in those few years and never go back to software :)

Actually, Anthropic reckons you've got three.

In 2024 (so now, one).

https://lifearchitect.ai/my-last-three-years-of-work/

When dockworkers went on strikes in 2024, one of their demands was to forbid automation of their jobs. HN responded negatively. How the tables have turned now it is white collar jobs' turn on the chopping block!

https://news.ycombinator.com/item?id=41704618