I built significant pieces of the Copilot onboarding, purchasing, billing and settings flow. For eight months I headed up the Copilot anti-abuse effort. I then led the launch of GitHub Models, and am now working on other Copilot projects.

I’ve thought about psychology. I know LLMs can work as pseudo-therapists but I feel like that’s a field where the human connection / human element will remain important.

Due to a text predictor?

I'm a daily user of the most recent Claude and while it's amazing at presenting other people's knowledge and reducing cognitive load by filling in the gaps, it's still just a machine that predicts text and that is a limitation which won't be overcome in this generation of such tools which, including research demonstrations, are close to a decade old already.

To me the main issue is that investors are not aware of these limitations and will keep pouring money into this way beyond everyone's breaking point. But really that's a failing of the world's economic system, which relies too much on their whims.

It conflates purpose with outcome. I don't accept the premise that software's purpose is to "automate away other jobs". That may be an outcome, but software's purpose is to enable completely new possibilities. Think bicycles of the mind.

The Apollo guidance computer didn't replace astronauts. It made the missions possible in the first place because no human can continually correct the spaceship trajectory every 250ms for 10 days on end.

We're not receiving some cosmic karma. We're more like cotton pickers after the cotton gin. Someone still has to plow and plant.

The way to survive it was to 1) move to a village/small town where you could have a garden for fruits, vegetables, corn, chickens, maybe a pig or two for winter. 2) Young people lived with their parents while the parents saved up to build/buy their children their own flat or house. Children whose parents saved up enough would often start a family after getting their own place. Those who didn’t, co-lived with parents.

The secure middle class corporate employment in the US is getting severely downgraded by AI. While there is talk of universal basic income the reality is that many many companies depend on the surplus that middle class families enjoy spending and without it, vast swaths of industries will get starved as well. The solution is to show people quickly how to hunt and gather and farm as makers instead of just employees. Figuring out what is needed, taking on a small corner of needs somewhere in meat space or online, and planting there. AI has been fantastic at helping even solo founders with that. They need to encourage a cornucopia of ideas and experimenters as early as k12. They need to set up more favorable conditions for handling the admin side as well.

If the US government does not encourage cornucopia of AI-powered small business entrepreneurship and lets monopolies squash that early, they will end up with far FAR worse conditions. Any monopoly who keeps pitching “universal basic income” while actively avoids paying taxes, will end up forced into more taxes.

Big tech needs to make room for people to build and grow businesses (looking at you, Apple, for copying every successful app with a native, and you Meta for eating every social competitor) or they will end up paying for everyone’s universal basic income and then some.

If this country wants to survive the AI era, it needs to remove the pink glasses of “secure corporate job” and teach people how to plant, hunt, and gather as independent players in the market really fast.

One thing I will add: while AI is getting really good at _doing_ the software building bits, I haven't yet seen it well integrated into the decision-making and political structure of organizations. Right now, it seems best in the hands of a high-agency individual empowered and able to make changes or 'ship' something, with them acting as the bridge.

This of course, is not a technical challenge, but I would expect the change in structure of organizations to make this more efficient to be slower than the pace of improvement we've seen over the last few years.

The wheel of industry rolls forward and crushes everyone underneath it

Most IC6/7+ would not code anyways - in fact a friend of mine said "we had our own agents we just called them IC4/5" - which was ironic but funny.

I am curious if we would ever get a new programming language like rust or go, without this creativity.

In a way, we have different products that does more or less same things (postgres vs mysql for example). The reason is there's difference of thought in the process. I doubt this will go away.

Maybe companies haven't seen it yet, but most office jobs can and will be eliminated in the next decade or two.

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 :)

The job is changing, and I don't like it in many ways, but there we go. It's not the first time new tech has nuked my dev job and I had to change.

I have personal projects that I hand-code, and personal projects I hand to Claude. Depends on how boring the project is. If it's stuff I've already solved a bunch of times, I hand it off. If I have room for good learning, I code it myself.

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.

My job most likely won't exist in ten years, do you think I'm worried? You might die in car accident, from heart attack or cancer in those ten years, some lost job is absolutely irrelevant compared to your health, there are and always will be jobs available, just adapt.

We have saying in my country "There were, are and (always) will be the money", which basically means you should not worry too much about money, just find a way to get them when you need them.

I strongly recommend LISTENING to this song - Baz Luhrmann - Everybody's Free To Wear Sunscreen

https://www.youtube.com/watch?v=sTJ7AzBIJoI

I quote one of the most important part regarding the article (especially 2nd sentence): "Don’t worry about the future; or worry, but know that worrying is as effective as trying to solve an algebra equation by chewing bubblegum. The real troubles in your life are apt to be things that never crossed your worried mind; the kind that blindside you at 4:00 pm on some idle Tuesday."

However I am not quite so defeated, I think that developers will continue to find employment in tech even as AI augments the roles. Experienced developers are the obvious pick for a hire to run agentic AI development tools, and even the obvious pick for managing a no-code endgame scenario as they are just smart technologists with strong problem solving skills.

I think the devs who were only here for the paycheck and would not reasonably pick software if it didn't pay so much, will probably be happy to retrain into something else but be disappointed by the paycut.

I am also excited by the prospect of being able to take on bigger scale side projects solo as that's really where my passion lies.

I think general purpose technologists will really excel in this new ecosystem as the industry will be back to moving fast and breaking stuff for a while, for better or worse. A lot of them call themselves programmers right now but will evolve pretty quickly.

Pragmatism, small teams and fast pace will best deliver software based projects, and the bottleneck in big orgs will become (or already was) the bureaucracy and communication layers. Small team, greenfield projects have a huge advantage in getting an MVP to market, which is pretty exciting for someone excited mostly by solving problems with technology.

Time will tell though, this is not career advice and times are chaotic. At the end of the day, there are other careers, and you were smart enough to get into software. You will be smart enough to find a new career.

I like the architecture analogy. An architect is not really focused on doing the actual building of a design, but understanding what's possible, what tools and techniques and materials are available, and figuring out how to put the pieces together to make a thing.

Right now, you're the architect who designs a house, but you're also the cement mixer, framer, drywall installer, plumber, electrician, and so on, all at once.

In this analogy, it's hard to design anything big like a skyscraper if you're bogged down by all of the minutia like picking out what type of nails to use for the framing and then installing them.

I think going forward, AI is going to do a lot of the non-architecture part of software engineering. We will all become architects.

The difference between us will be the scope of what we're qualified to design as we go through our careers - new grads will cut their teeth on the likes of designing shacks, principals design skyscrapers.

I think this also unfortunately means there's gonna be a lot less people in software. The industry will still exist though, but it's going to look way different.

I look forward to this being settled out, the uncertainty sucks.

Software validation was always the interesting part vs software verification. Validation asks the question, did you build what was actually necessary?

In 2024 (so now, one).

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

On the other hand. Unless we have a breakthrough hardware/physical innovation, GenAI in its current forecast is not energy-efficient, cost-efficient, compared to human/deterministic methods. It has shown no capacity so far to « create » in the sense we animals do.

And all that, still being: highly subsidized (your subscription does NOT cover the costs of the service as of now, we are still in the market creation/capture phase), and without mesurable economical benefit.

Things are still far from being over.

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.

First, this kind of phrasing is repeated again and again since Cobol was launched. A machine code writer said to Grace Hopper: "everyone can write programs!"

Second, as Dijkstra said once "computers are for programmers like telescopes for astronomers", programming isn't about writing lines of code, programming is about knowing how to deal with requirements and putting them to work as a computer program.

> I'm a daily user of the most recent Claude and while it's amazing at presenting other people's knowledge and reducing cognitive load.

'Presenting other people's knowledge' is enough to get the job done when that knowledge encompasses the entire internet.

But I don't think the demand will ever be zero, or that laypersons will ever write (useful) software using AI, because most people do not understand what software is, what it does, what it can do, where to start, what to ask, what is data, what is input vs. output, etc. They are incredibly clueless, and it's not a problem of intelligence. Some of the most clever people I know have no idea about this. (Maybe they don't care enough to understand, or maybe it's a mindset that you either have or don't have, IDK.)

I just don't see how we could do without people to think things through.

this could be one of the silver linings to AI disrupting the industry. tech was better for the world when it was run by nerds that were in it for the love of the game.