Feeling sad about AI
September 11, 2026 [Programming, Tech, Videos]This video doesn't contain any code: feel free to skip it if you like, but either way I'm going to try to keep it short, so see you on the other side :-).
Also, this video isn't about the ethics of AI - you can read how I feel about that in my article Why I don't use AI.
And it's not about whether AI "works".
It's about how I'm feeling, and some reasons for why I'm feeling like that, and some hopefully hopeful conclusions.
Feeling sad
AI has been making me sad for quite a while now. Really sad, out of proportion with how much AI has materially affected my life. Sad enough actually, that the sadness is the main effect that AI has had on my life, making it harder for me to deal with everyday life and be a good spouse, parent, friend, colleague.
Obviously, AI is a potential threat to my livelihood and career, especially if I continue to refuse to use it, but I don't think that explains quite how sad I've felt.
Disrespect
I think I've felt sad because of the disrespect. For most of my life I have loved programming: I've made it my hobby, my work, my identity. I made it my way of proving I have value, because I can be good at something, and improve at something, and it is useful, and beautiful: you can make cool things, and those things can be made in ways that are like craft or art or science or Maths.
And if you can do it, people think you're clever, and respect you.
Now some people in our industry are telling us that programming is obsolete, that we are dinosaurs, and the only exciting or viable career path is to move on from programming, the thing I love, to paying them to rent programming from their machine, because it's a commodity now.
The thing I love; the thing I made my identity. The thing that earns me the respect I am so desperate for.
I know a lot of people could be watching this and thinking, "where were you when they disrespected me?" How much respect is given to the work of farmers or nurses or drivers or parents or carers or teachers or hospitality workers or cleaners or car-washers or so many other people? I know, I am late to the party, but I've arrived now.
It was a shock to me that bosses were so keen to ditch people like me and replace us with an imitation. How naive I was, and how self-centred.
So that's why I've been sad. But I want to move on and try to provide some hope.
I've got three people in mind, who I'd like to encourage: myself, people watching my videos for the love of coding, and people watching them to learn, to build their career.
Let's talk to me first.
To myself
Andy, the first thing you need to understand is that they can't take your love of programming away from you. You love it. You just love it: it excites you, it calms you, it gives you a creative outlet, it satisfies you. That's a God-given thing about you that they can't take away. They also can't take away the fact that you are quite good at it, and can improve. Enjoy it. Don't waste your time and burn your creativity caring about receiving respect from people who don't deserve any respect from you.
To the code lovers
Now, for people watching for the love of coding, the message is the same: they can't take it away from you. Just enjoy it. If you like watching videos about coding, encourage the people making them, and you might just supply enough energy that they are able to continue.
To the learners
For people watching to learn: don't listen to people who say this skill is obsolete. Let me be clear here: even if the highly optimistic predictions about AI are true, we are going to need a lot of code, and the people who understand how it works are going to be very, very useful.
First, the type of thinking you learn by coding is going to be incredibly useful, even in a future where that thinking is applied to wider-scale problems. Training your brain to create complex systems, analyse them, remodel them, debug them is going to make you useful, and employable.
Second, the knowledge of how code actually works is going to remain useful. Did we stop understanding machine code when we wrote compilers? No! People have jobs writing compilers, and their understanding of machine code and pipelines and caches and locality and loads of stuff I don't understand is way deeper than anyone had when compilers were invented!
I have skin in the game here myself: two of my kids are learning programming and heading towards potential programming careers. Despite my deep sadness about the changes in my industry, I have no worries about them learning non-AI-powered coding at university and school - I am convinced it will set them up for whatever the world looks like over the next few decades.
Love to you
So that's all I have to say. If you are feeling some of the confusion, fear or sadness that I know many of us are feeling, I hope it might help a bit.
Let me know on Mastodon (@andybalaam@mastodon.social) or via a comment on my blog, or under the video.
Love to you: you have massive, massive value, and you deserve respect, as you are today. Whether you can code or do anything else is irrelevant. (Yes, I'm talking to myself again.)
Love to you.
Sriram 2026-09-11
I'm with you, without you, for you! - Computer Programmer turned Computer Comedian.
qarl 2026-09-11
i too am a long time programmer. about 40 years at this point.
i see AI very differently than you do - and by sharing my perspective - maybe i can give you some comfort:
for me - the love was not about the coding itself. it was about what that code did in the end. i used to explain it like this: "i make pretty pictures with code, and for some reason it makes me happy."
none of that joy is dampened by AI. it is amplified.
over those 40 years, i have had endless ideas that never got executed. because programming is hard - and the best ideas are even harder.
and now i'm able to execute every last one of them. and believe me, i am. and it's amazing, unlike anything I've ever experienced before. i tell people it feels like flying.
i agree wholeheartedly that AI is seriously dangerous. it's the new nukes. and it probably won't end well.
but in the meantime i can fly.
Andy 2026-09-11
qarl, I am glad you are finding joy but I urge you to consider whether AI is good for the world - either the climate, the people exploited to train it, or our own brains when we use it.
It's Not AI 2026-09-11
Industry's attitude towards tech from 2010s until now was somewhat anomalous.
As an engineer you're labor. Capitalism exerts tremendous pressure on labor costs. If an innovation to reduce them exists it will be explored and perfected. Moreover the Fed's mandate to control inflation includes, wage inflation. In other words the government and industry actively collude to make sure you don't get paid too much. The upside is this collusion also prevents you from getting unemployed too much.
The other good news is that engineering is really just rapid iteration of the scientific method to solve hard business problems. The need for this isn't going away regardless of the tools. Also, it's difficult enough that most people can't excel at it.
Capital isn't going to stop paying for someone to do their deep thinking for them.
Andy 2026-09-11
Yes, It's Not AI, they jumped at the chance to take away our worker power. We should wake up and learn from all the other workers, and work together to resist.
Andrew 2026-09-11
I absolutely agree with how you're feeling. When I think about the sadness, it feels closer to mourning or grief - the loss of something that was, and that we can't get back.
Technology was an avenue of hope before. Applied technology fixed problems, made lifespans longer, enabled more people to access education and opportunity. It was a progressive force.
The script has flipped now. Technology is now regressive. It's used to surveil, to oppress. It's gone from being an tool to fight climate change to the primary driver for it. It's centralizing control and reducing opportunity, where it used to make those things more inclusive.
I still think tech can be a force for good, but only if humans can retake the reigns.
Jaye 2026-09-11
This is a timely and lovely message, and just what I needed to hear. Thank you for putting this out into the world. <3
Anonymous 2026-09-11
I so strongly relate to this video. I literally have a draft doc called "Why I don't use AI" that looks extremely similar to your post which I read for the first time today. I work at a company that has an AI mandate. I'm trying to negotiate a way to eke out a continued existence like this. I know I'm not as fast. I know I probably don't catch as many things. But I hate the alternative world that the evangelists are building and they'll have to drag me there kicking and screaming.
Andy 2026-09-11
Good luck in your struggle Anonymous and remember everyone has to compromise in this system so be kind to yourself and others, and fight with all you've got.
Muhammad Rijal Dwi Rozaki 2026-09-12
I felt better after reading this. Nobody will stop my love towards mastering programming.
Juan Cruz 2026-09-12
Thanks for this blog post <3
Sincerely, someone who's also feeling sad about AI.
Martin 2026-09-12
Hi Andy,
See this video, it will make you feel a lot better (I promise):
AI Can’t Code: 7 Myths Debunked
Kind regards
Andy 2026-09-12
Thanks Martin, I do agree with a lot of this, but I also think a lot of it is still to be determined. Even if AI miraculously became brilliant at programming, my ethical and respect objections would apply. I find the debate about whether it works somewhat unproductive right now. I guess we can wait and see...
Tiago Martins 2026-09-12
Thank you for this. I thought I was alone feeling like this, this past week. Great article.
Silas 2026-09-12
Solidarity. I've been unable to find an industry job for the last 2 years: so many good engineers have been laid off for cost-cutting in the name of "AI" that it's now harder to get noticed by the hiring managers who let's face it don't really know what they're doing. I'm still a teaching assistant, and a first-year asked me why we're bothering to learn about practices like Agile "when none of us are going to get jobs". My reply was "this is a bubble and there's still a chance the industry will come to its senses before you graduate" (he says hopefully; meanwhile powers higher than me are still debating whether to keep that module after our best lecturer for it passed away). It's certainly taking longer than I thought to sort itself out.
Read more
And it doesn't help the atmosphere when we get press items like an OpenAI executive issuing vague threats to "stop being nice" in a way that ruins a top mathematician's career, although, in the context of all the other people's careers ruined by OpenAI supporting the marketing of their service as a replacement, maybe it's just par for the course.
I'm still optimistic about the potential for the technology itself once we have it in its proper place and being done responsibly, but what I can report is, taking a more middle-of-the-road approach than you and starting to experiment with LLMs within reason, has not yet magically plopped me back into the workforce, so perhaps you're not missing out that much by going more 'hardline'.
I recently came across a "little-guy AI" company whose engineers seem to have had an interesting idea: bolt on a retrieval system that scans the chat history for anything sort-of matching, so they can use a much smaller, more energy-saving LLM and rely on the extra code to make up for its more limited context window. Sadly their execs have not yet approved open-sourcing that code, although they have open-sourced their small model so maybe one day. Meanwhile I wonder if it might make you feel a bit less bad to show them that such a system doesn't need an LLM at all: just make an application that lets the user write a diary of notes to self and auto-retrieve old ones with the most matches (it's been done by issue-tracking systems but the interface is different). I reckon LLM-assisted "brainstorming" is mostly in the brain of the user; maybe that's testable.
Andy 2026-09-12
Hi Silas, I'm really sorry it's been impossible to find an industry job recently. I am really hoping the chaos will reduce soon. Good luck!
Andy 2026-09-13
To the people tempted to make mean comments: when someone makes themselves vulnerable in front of you, it's not cool to attack them. Remember we are all on the same side.
anon 2026-09-14
This read was quite uplifting. I've been struggling with impostor syndrome lately, even though I am aware of my skills. Your encouragement makes me want to continue work on my dropped side projects. Thank you!
Hasan Ahmed 2026-09-14
Andy, thank you for this. I have been feeling the same way. I'm glad to have it spoken aloud and glad that I'm not alone in this feeling
Andy 2026-09-14
Hey Hasan Ahmed and anon, I'm really glad it was helpful. You are definitely not alone!
Abdullah 2026-09-15
I agree with how the ai models are harming it's users in terms of skill degrading, less efficiency in their job, cutting their job and more, i believe this is because companies who made the LLMs are bad. had the companies who made the current known models (alibaba, anthropic, openai, Alphabet) not scraping the whole internet to get the hunders of thousands of terabytes of data to train it's model taking people's work ideas and provide a possibly harmful model, also the companies could have instead of developing the model and train it on massive data and resources(water-electricity-HBM) fix climate change or create a better chips that runs a very small budget , or maybe creating a cooling system with magnets for example to fix the need to use gas in homes where it can be replaced or help in the development of the new solar panel or a new battery on salt (all these example are found but needs investing and researchers ) had they done any of these the people will have a better life a safer home a cheaper power a cleaner storage all of it , but instead these companies spent billions of dollars(if not trillions) to develop a bot that can write code, draw an image or a video, chat in personas, while none of these help humanity in anyway. on top of that it's more of harmful according to the studies. but i disagree when it comes to say all llm are bad (by llm i mean neural networks in general) for example speech to text is a very useful piece of software that could be used in many ways (of course it has a possibility of misinterepting the speech)
Andy 2026-09-15
Hi Abdullah, yes I agree with you completely. This software is a tool, and the vast amounts of computing resources, and the useful properties of these tools in particular, could have been used to help people. But this is not what is happening.
Dave 2026-09-16
Thanks. This captures a lot of how I feel. I have been going through a very challenging period of life the past year; first child, having to interview in a very bad market, first tech lead role, new job at a giant international company with a "challenging" culture and trying to adapt to this new LLM obsessed world.
All of these have been difficult to navigate at the same time and I have had multiple times that I genuinely thought I was going to have a breakdown in very public spaces.
But without a doubt or a lick of exaggeration, mourning the death of something I love and cherish so much has been the hardest. And seeing everyone around me at my company seem okay with it all has been quite isolating. It has made me feel like I am crazy or something.
I've found today particularly hard to deal with these feelings. Close to tears honestly. While staring at a copilot session trying to argue with an array of glorified statistical models to do something correctly.
Then coming home looking horribly upset to the point that my partner looks panicked asking me what's wrong. Then trying to explain my feelings towards all this and it not hitting home. She can't understand it, but she tries and kindly sympathises.
So thanks. I really really really needed to read and watch this today. It's made me feel at least less isolated.
Andy 2026-09-17
Dave, that is all incredibly hard. Love to you. Our craft is not dead: many people are living in a fantasy world, and at some point things will become clearer. In the meantime, let's try to get through and support those around us.
Vincent 2026-09-17
Andy, thanks so much for this. I am sad too. I am 65 years old and being a programmer is who I am. I run my own small company and I use AI tools but it has taken much of the joy from it. Already, we are pivoting, trying to guess where all this will land. I hope your kids can navigate whatever it is to come and I agree with you that knowing how code works, how systems are built will serve them in the future.
Phong 2026-09-28
I'm in a program to become an expert of AI. The more I get into it, the more I feel it's not for me. AI forces many people to become "perfectionist" (productive face cheater). Some people have become playing too safe, too good to be true, not willing to take risk. I, as a risk taker, feels ashamed of myself when AI changes the way I work and interact with the world. AI brings fears, fear of making mistake, fear of missing something... Perfectionism doesn't make us happy, AI ruins it. I'm sad.
RM. 2026-09-11
Thank you. So on point, this article hits exactly on what I've been feeling in the past few months. I have also been telling myself that I program for the love of it, and must not rely on validation from an industry that has ever-shifting trends. If you love to code, just code like before all of this.