thearrow

Felt this in my bones. It’s painful how accurately this describes my current work situation. The pressure comes from the top - leaders that have no idea what they want but they want _something_ to happen NOW. This has only gotten worse with the recent AI thoughtleadering because now the expectation has been seeded that every random brainwave should take at most one hour to implement (given enough tokens).

What can an IC do in an environment like this? If you slow down and attempt to find any clarity, you’re labeled as slow and ineffective. If you cave to the pressure and start slinging slop with the rest of them, you’re just perpetuating the spiral. Genuinely asking - how do others thread this needle?

I also feel this, and it also troubles me. Unfortunately, we may be in the minority on this site.

What the other replies seem to overlook is that it fundamentally changes the nature of the work - it’s not the next step in the evolution after an IDE, it is closer to an automated tractor, and yes that does make the farmer’s work trivial. Pressing a button and having the field get plowed is a very different experience than manually plowing a field yourself. I can no longer in good conscience say “I plowed that field”, because all I did was press a button. The tractor plowed that field.

Most people on this site seem to feel comfortable claiming “I plowed that field” after they pressed the button that started the automated tractor. Okay, you had the idea to press the button - but you didn’t actually do the work, you delegated it. Same end result, but a different lived experience. Now would I rather actually plow a field by hand or simply press a button? You can probably guess - but the two actions do differ greatly in the experience they bring me, and the feelings of satisfaction.

The test suite explicitly verifies the architecture and the data structures used? Depends on the suite, I suppose.

A 50-year-old oak is not just lumber. The fact that you think these are equivalent is a hilarious, classic HN take.

Happy Valetudo user for years. It has great integration with Home Assistant, too. Highly recommend.

Same. I was very skeptical when Macrofactor introduced this feature, but have since been incredibly impressed. The ability to give it text alongside a photo and then adjust the results (broken out by ingredients) are critical. I’ve also been taking pictures of food sitting on scales and it will take the measurement into account.

Seems like the Macrofactor team took their time developing this feature, as it felt like they were one of the last to roll it out, but the extra polish definitely shows and was worth the wait.

Absolutely hilarious! I bet you forego using calculators and do all of your arithmetic by hand, too. Such a clever analogy!

I love this essay, and it very closely mirrors my own feelings about the topic. Thanks for writing/sharing it.

I’m perpetually bamboozled by my fellow software engineering colleagues who insist on proudly shouting from the rooftops “Look at me, ma! I’m vibe coding!” as if it’s some badge of honor to see who can churn out the greatest quantities of shitcode the fastest and completely surrender any last scraps of their cognitive abilities to the best LLM provider of the current moment.

Echoing many of the other replies here - my experience has been very different (much better).

The web UI is fine and snappy and using Infuse on Apple TV is simply delightful.

The server uses file watchers to update the media library very quickly and is light on resources.

I don't need a recommendation engine because... it's my media library, presumably I added things to it that I want to watch.

And most importantly, it's open source and not likely to get enshittified in the near future like Plex.

Nice analysis! I still have an original Sega Saturn I’ve owned since 1996 that I fire up occasionally for a nostalgia bomb. The thing still runs perfectly, same as the day I unboxed it! They may have ended up with quite a complex hardware architecture, but you’ve gotta love the reliability of the older consoles. The same cannot be said of the more modern consoles I’ve had over the years - burning themselves up or failing in other ways.

Do you use a react component library? If so, which one?

I’m also curious about the answer to this! I noticed similar behavior when opening a Typescript project. Enjoy the low latency, but I’d also appreciate accurate/helpful autocomplete suggestions.