Firing programmers for AI is a mistake

Follow the money, all the way down to Mother Earth; which boats are lifted most, and at what cost?

This makes sense. I've mostly been successful doing these sorts of things as well and really appreciate the way it saves me some typing (even in cases where I only keep 40-80% of what it writes, this is still a huge savings).

It's when I try to give it a clear, logical specification for a full feature and expect it to write everything that's required to deliver that feature (or the entirety of slightly-more-than-non-trivial personal project) that it falls over.

I've experimented trying to get it to do this (for features or personal projects that require maybe 200-400 LOC) mostly just to see what the limitations of the tool are.

Interestingly, I hit a wall with GPT-4 on a ~300 LOC personal project that o3-mini-high was able to overcome. So, as you'd expect - the models are getting better. Pushing my use case only a little bit further with a few more enhancements, however, o3-mini-high similarly fell over in precisely the same ways as GPT-4, only a bit worse in the volume and severity of errors.

The improvement between GPT-4 and o3-mini-high felt nominally incremental (which I guess is what they're claiming it offers).

Just to say: having seen similar small bumps in capability over the last few years of model releases, I tend to agree with other posters that it feels like we'll need something revolutionary to deliver on a lot of the hype being sold at the moment. I don't think current LLM models / approaches are going to cut it.

When I was a kid I "wrote" (mostly copied from a programming magazine) a 4-bit CPU emulator on my TI-99/4a. Simple as it was, it was the light bulb coming on for me about how CPUs actually worked. I could then understand the assembly language books that had been impenetrable to me before. In college when I first started using "C", pointers made intuitive sense. It's a very valuable exercise.

You haven't answered those questions. Tesla's touchscreen displays maps, navigation, self-driving's model of the world around the car, reversing camera, distance to car in front, etc. Yes personally I prefer a physical control I can reach for without looking, but the physcial controls in my car cannot do as much as a touchscreen, cannot control as many systems as a modern car has. And that means something like a BMW iDrive running some weird custom OS in amongst the physical controls, and that was not a nice convenient system to use either.

Why write a couple of lines of code when you can just include an `is-odd` library? Hopefully one which type checks integers vs floats, and checks for overflows. I'm not stating that I could not write one if/else, I'm asking you to do more than sneer and actually justify why a computer loading a couple of lines of code from a file is the end of the world.

Why invest time and effort into making a good TV if people aren't going to buy it, because they are fine with the competitor's much cheaper Android OS on questionable hardware?

Why run and manage your project on a baremetal server, and deal with its power requirements and cooling and firmware patching and driver version compatibility and out-of-band management and hardware failures and physical security and supply chain lead times and needing to spec it for the right size up front and commit thousands of dollars to it immediately, if you can just rent Amazon DynamoDB and pay $10 to get going right now?

I could fill in the answers you are expecting, I have seen that pattern argued, and argued it myself, but it boils down to "I dislike laggy ad-filled Android TV so it shouldn't exist". And I do dislike it, but so what, I'm not world dictator. No company has taken over the market making a responsive Android-free TV, so how/why should they be made to make one, and with what justification?

> What more could I name?

Why go to a cobbler for custom fitted shoes when you could just buy sneakers from a store? (I assume you wear mass produced shoes?) Why go to a tailor when you could just buy clothes made off-shore for cheaper? (I assume you wear mass produced clothes?) Why learn to play a keyboard, guitar, drums and sing, when you could just listen to someone else's band? (I assume you listen to music?) Why spend months creating characters and scenarios and writing a novel when you could just read one someone else wrote? (I assume you have read books?) Why grow your own food when you could just buy lower quality industrially packaged food from a shop? (I assume you aren't a homesteader?) Why develop your own off-grid power system with the voltage and current and redundancy and storage you need when you could just buy from the mains? (I assume you use mains electricity?)

You could name every effort-saving, money-saving, time-saving, thing you use which was once done by hand with more effort, more cost, and less convenience.

And then state that the exact amount of price/convenience/time/effort you happened to grow up with, is the perfect amount (what a coincidence!) and change is bad.

> Somebody needs to write that operating system the LLM runs on. Or your bank's backend system that securely stores your money. Or the mission critical systems powering this airplane you're flying next week... to pretend like this will all be handled by LLMs is so insanely out of touch with reality.

When they do this, I really want to know they did this. Like an organic food label. Right now AI is this buzzword that companies self-label with for marketing, but when that changes, I still want to see who's using AI to handle my data.