sdevonoes

Next year you may be unemployed (because your boss will cut by half the engineering department), or perhaps your salary will be cut by half (because your boss will route your other half of your salary to openai/anthropic), or perhaps you will get half insane (because your boss will ask you 10X the output you used to have but within the same timeframe)

Whatever the path is, regular engineers (90% of the people around here) will get screwed up one way or another. But hey, playing with LLM agents is cool!

It makes sense because those who can are multibillion dollar companies/people. You and me are just peasants. That’s why we should avoid using proprietary agents, just so the rich ones don’t get any richer

Problem is where to stop. Open a 3rd session? Are you sure the 4th iteration is mostly correct? Let’s try a fifth now…

I wish I could care more about what the computer wants (because it’s fun). I couldn’t care less what the product demands.

> Business people can just directly convey what they want to the AI. Today, yes, this often leads to a mess. But for how long?

As long as “business” people don’t wanna do the job and prefer to pay for it. It’s like, you can fix your own sink pipes if you follow youtube video tutorials long enough… it’s just usually not worth the effort, and so you pay someone to do it.

Business people, I imagine, don’t want to deal with software at all.

But isn’t “grooming” the hardest part? To know the details means to know the codebases behind. So you do discovery, read code, ask people, check db tables, and then once you figure it out, you write a doc with the proposed solution. It gets reviewed and then you code it.

I definitely need AI help for the discovery part… so it always starts with a simple “I need to do X”

I would replace the eng. managers

> “the AI did it" to be any acceptable excuse.

It will be depending on how much you paid. My bet is that in the near future normal people (like most of us) will be attended by AI in pretty much everywhere: the bank, the hospital, shops, hotels, etc.

You will need to pay extra premium to be attended by a human being. Actually that’s what’s been happening to the ultra rich since a very long time ago (e.g. a millionaire doesn’t book an appointment by using their phones, they ask their assistant to book for them; they don’t go online shopping, they get people do the shopping for them, etc)

Never understood the “skills” part tbh. Thse models are being trained in trillions of texts. Adding something small and particular about my codebase doesn’t really move the needle

AI as technology is cool and all, as cool as GCC or Linux or HTTP. The thing I don’t like about AI is that 99% of the time we talk about vendor-locked AI (claude, codex) and 99% of the time we use those. So you need to pay to use the models. You depend on the big corporations. It’s not anymore, I clone this repo and let’s explore freely.

Nah, LLM models are already the new compilers. A commodity only engineers know how to use (in the context of software engineering in production environments)

I think things would take longer to get done (and that’s fine for me tbh).

The main usage of AI for me is to do discovery: if I am working on service A that has a dozen other services as dependencies, then I clone all the repos, point the agent at them, and start to ask questions. I double check ofc. But without the agent I would be back to manual grep code and git logs, asking around, etc… that takes longer and you may miss things

Tip: include pictures of your machine. Not sure if I’m alone in this, but I love to see other people’s setups

Same interview process. Recruiter call, Live coding, System design, Behaviour interview.

The only difference is that in the Live coding session they now allow AI to be used. I didn’t use it and went old way (because I had already rehearsed it the old way, and found no feedback on the internet about how the company assess AI usage, so I didn’t want to be the Guinea pig)

Interview took place this summer. Major tech company in EU (not faang tho). Also Live coding is not leetcode.

I love to write code and solve problems in personal projects. I love to read interesting technical books. I love to learn certain new technologies.

I hate doing the same at work. I couldn’t care less about my CEO’s product, nor about our linting tips or dependency management rules. My PO can pretend he’s the owner of the company, my engineer colleagues that we are shipping the next google. I couldn’t care less. I do it because I get paid and because it’s my career.

So, it’s not like “you like coding or not”. It has nuances

If someone comes with a pretty decent idea and execution, it doesn’t really matter if code was written by human or llm, does it? Like if the next torvalds ships something great in the open, who cares how was it written.

I would love to buy the machine but I hate Tahoe.

So many things he got right (except you know what)

I won’t lie. I want to be Captain/Director, not the rower.

It’s very satisfying to get rich by doing pretty much nothing. And there’s some sadness when some rower does their best but doesn’t even come close to directors/captain positions.

Some excellent rowers can make it sure, but we all are not excellent. Oh well, I stop

Users don’t like shit, though.

I think it depends on the job one has. Working as senior software engineer for international companies? Any western european country is nice and all. Top p90 salary. English only needed. Working as a bank cashier? Well, then you will notice the difference between let say Germany and Portugal. Low salary. Need proficiency in the local language, etc.

The litmus test is this: do you enjoy consuming AI-generated stuff?

I don’t. Whether it’s written text, or video, audio, restaurant menus, clothing pictures, documentation, airport control, ads…

I do think there’s value in LLMs but as a sort of better search engines and q/a machines.

Im not so sure if this is a pro or a con. Take two extremes at product building: brute force vs deep research first. I guess llms allow you to experiment quickly with many approaches and see what sticks. Without llms, you need to spend more time researching and thinking what’s the most economical approach. Im biased towards the non-llm approach

Everyone can write; that doesn’t mean everyone can produce an interesting novel. If everyone can produce software via llms, that doesn’t necessarily mean that software is useful or interesting.

Compilers also write code (they write machine code based on high level langs)… yet nobody asked the same question about them (or did they?)

AI needs to be instructed and then some kind of verification of the output is needed. Programmers do that

So, if engineers out there are not even interested in the most immediate layer of code in front of their eyes (e.g., the code that is now written by LLMs), I guess more and more the kind of stuff like the one in the article is becoming like software archeology, isn’t it?

I love to learn these stuff, and pre-llm era it gave me a plus career wise. Nowadays I guess knowing this stuff pays less off (besides increasing your curiosity)

As someone who has never visited China or the US; yes. Why? Media. School shootings, ICE, crazy president… What do I see from China? Impressive multi-million cities… and true to be told, not much.

China had a worse reputation, but not anymore since Trump leads the US.

Im not saying that the above is objective, tho

Would that single person handle different topics such as: UI/UX, security, backups, distributed systems, data migrations, infrastructure, …?

Sure thing there are engineers out there that know some about all of the above (I personally do all of that on personal projects) but you still need specialists, otherwise it’s you alone with all the unknown unknowns that the llm may claim to solve, but you cannot verify

But searching on google or SO wasn’t the bottleneck either. It was searching on your own codebase (and dependant ones) for the exact place (a class, a file, a function) where to put/delete/update business logic without breaking production.

LLMs help, but they haven’t been trained on our own repos. I don’t need the LLM to help me with algos that are available online… I need them to help me with custom business logic

Got it. But then why Nike/Adidas/Quicksilver/etc cannot bring back clothes from the 90s?

Buying second hand is not that nice (smells that never go away) and “vintage” places sell min. catalogue buy way overpriced.