I don’t doubt it’s possible to get better consistency but the juice is really not worth the squeeze for me. You end up churning through huge expensive models, orchestrating sub agents, writing out boilerplate hand-holding instructions (“please don’t break this, stop trying to commit to main, please lint ffs…”).

I don’t use it for Java but that would make sense with rigid enterprise patterns and VeryVerboseNamesThatAreEasierForAModelThanAHumanFactoryClazz {...

I don’t think our career is boned, moreso that all juniors trying to get in are boned. Everyone who knows what going on transition to a more hands-off architect role.

But like I said, our tokens are heavily subsidized right now. When they pull the rug, code monkey jobs will start to get listed again (with lower salaries of course).

Things I’ve realized while working with AI (Claude code):

• It’s fantastic for very small macros and medium length scripts. Think dev ops stuff, pre-commit hooks, transforming data. Keep it small enough to manually review and something you can run without destroying anything important. This can massively boost your codebase QoL. [Double bonus for not wasting tokens to solve the same problem over and over]
• It’s decent-to-good at debugging but not consistent with fixes. It can find some utf encoding edge case that might have taken you 1hr+ but suggest the dumbest bandaid fix you’ve ever seen. Also very good at spinning up unit test suites for basic edge cases.
• Due to obvious training bias, it’s pretty good with common libraries and cloud platform infrastructure. It could probably help with writing a complex cron call, debugging regex or fixing an IaC config. On the flip side it won’t bother to use the latest package version or know your niche/new library.
• It does better with greenfield because exploring your codebase introduces a ton of bias. It might try to fit in an ugly hack when a refactor to simplify everything is way easier.
• It’s absolutely garbage with UI, just throws the most disorganized HTML together that isn’t reactive or reusable. OK enough for ugly internal stuff but God help anyone relying on it for that.
• This is setting up to be the biggest rug pull in history. People that buy into it heavily just to save a couple bucks on engineer payroll are going to be fucked when they start ratcheting up the token price.

All in all it can be useful when used with care but will never be a magic bullet.

Photos are the same as most other data, you can store them pretty easily long-term in a physical medium. Of course, capturing an image is much easier and more convenient with a digital device, but that doesn’t mean it has to live digitally indefinitely. It’s simple enough to have an instant digital camera with a built in printer and access to a high quality scanner.

If you held a gun to my head, I could pick out a few dozen personal photos that I own that are worth saving physically. If you allowed me a modern flash drive’s worth of storage (64-128GB, ~5000 good quality images), I could pretty easily store every picture worth a second look from my entire lifetime.

Apple’s marketing driven perception that every single person needs a cinema quality camera (and cinema sized storage) in their pocket is ludicrous. Only a tiny fraction of people actually truly need that. Let them borrow that gear from a library if we want to preserve fair access.

To be honest I don’t have any personal digital information that I give a shit about. It’s value is only derived from its ability to identify + track me, either for my convenience or for the highest bidder’s. Computational liberty is only an issue because we’ve made everything digital by default and that mindset has leaked into critical social functions (taxes, law, logistics, healthcare, etc…).

Software and data bloat is more astronomical than most people realize. Only about 10% of persisted data is ever touched again (don’t look up the ecological implications). Amazon could capture 90% of all compute hardware and the entire human race could get by just fine on 10%. We wouldn’t have access to niceties like app stores full of niche apps, 24MP phone cameras, 4k movies, 10 sluggish layers of software abstraction, 15 years of photos you never look at, etc…

But you could run a simple message server on basically any scrap of IoT e-waste. A highly available static website can be hosted with an old phone and a solar panel. Any device (fridge/watch/calculator/pregnancy test) can run Doom. All of Apollo 11’s source code is a fraction of the size of most web pages.

We’re continously expanding our hardware usage for infinitesimally small gains. We should demand that our governments legislate digital austerity for dozens of reasons, just pick what resonates best for you. Personal privacy, energy usage, ecological damage, corporate capture, information rot, brittle supply chains, national security, etc…

Crazy that companies will do this shady stuff with client side code. At least it was slightly obfuscated at first, but that’s just incompetent fraud to leave it so obvious that a self professed non-software engineer (though clearly a smart guy) can read it and deduce what’s happening. Throw a tiny bit of random noise to the stepdown logic and it becomes much harder to find and reproduce as proof.

Idk the model but can check later. Removing/modifying it isn’t an issue, but my household wants to use it since it’s there.

I’m not personally opposed to a camera but would need to be in full control of the feed. My main goal is keeping it simple and cheap for now, so not replacing a functional camera is very tempting. Later on I can look into real alternatives but an afternoon project will do for now.

Wouldn’t be operable, but I like your moxy lol