The AI “revolution” is the thing that finally killed my imposter syndrome as a software engineer. Not because I can write better code than AI (that’s a very low bar), but from listening to all these breathless idiots talk about how they’re “10x-ing my productivity!” or how “AI has replaced search for me!” or how “In 6 months no one will have to manually write code anymore!”
Similar for me. What i find ironic is that AI already ran into a brick wall. It’s inherent statelessness by design means that AI is unlikely to be suited for anything more than isolated well defined tasks in the near future. Still usable as a tool, but without someone who is actually experienced, it will result in disaster.
and even in smaller tasks it can fucks up, especially if the person prompting it is incapable of writing the code themselves as they don’t know how to properly design it and don’t spot the issues. Like everything with AI, it looks impressive at first glance until you look at it for more than 10 seconds and spot the metaphorical 6th finger.
What we see currently with AI getting “better” at coding is more or less duct tape to make it work. Basically, they create the agents to bolt on the state, more layers between user and model. Iterative processes to make the answers better, etc, and to create “memory”, which in essence is just an ever growing prompt managed by the agent. But in the end, this won’t fix the inherent problem, so it will only do so much and is already hitting another ceiling. It introduces state decay. With the agent method its not really possible to “take away” memory, so if you gave it multiple versions of the same code (as you would if you work with AI), the AI never really forgets about old code. It can supress it through agent instructions (more duct tape), but the more there is the more it bleeds through, which can make the AI reintroduce old code or base assumptions on outdated things.
There is no fix without changing the inherent way how models work, which would introduce complexity beyond what is currently feasible in computing (and the current AI is already gobbling up all computing reaoureces as is)
If you’re someone who cares at all about the quality and consistency of your craft, there’s absolutely nothing wrong with manually writing code.
If you’re a misanthropic “techno-feudalist” who thinks of code as nothing more than an asset to sell, then pumping out as much code as quickly as possible without any human intervention is a very attractive proposition.
Tech, sadly, is absolutely infested with these people at all levels.
You will still need to learn programming manually.
The process of struggling to understand and synthesize working code is a critical part of learning. Skipping it feels easier, but you’re hurting your ability to understand coding.
Sure, you can make an LLM generate code and if you’re inexperienced it can outperform you on the basic tasks that you’re given as exercises. This is a trap that a lot of students fall into. It’s very easy to let LLMs do the ‘hard work’ part of learning while you just read the textbook or watch a video. Unfortunately, the hard part is the part that builds your skillset.
It’s just like how you can’t just watch a video about physical fitness and then use a robot to lift the weights for you. Sure, you get to the end of your sets faster and you’re not physically tired and sore but you won’t actually benefit in the ways that matter.
There’s nothing wrong with writing code manually. Over the past few months LLMs have gotten a lot better at writing code than they were before, but they can still make weird mistakes.
Preach.
The AI “revolution” is the thing that finally killed my imposter syndrome as a software engineer. Not because I can write better code than AI (that’s a very low bar), but from listening to all these breathless idiots talk about how they’re “10x-ing my productivity!” or how “AI has replaced search for me!” or how “In 6 months no one will have to manually write code anymore!”
Similar for me. What i find ironic is that AI already ran into a brick wall. It’s inherent statelessness by design means that AI is unlikely to be suited for anything more than isolated well defined tasks in the near future. Still usable as a tool, but without someone who is actually experienced, it will result in disaster.
and even in smaller tasks it can fucks up, especially if the person prompting it is incapable of writing the code themselves as they don’t know how to properly design it and don’t spot the issues. Like everything with AI, it looks impressive at first glance until you look at it for more than 10 seconds and spot the metaphorical 6th finger.
What we see currently with AI getting “better” at coding is more or less duct tape to make it work. Basically, they create the agents to bolt on the state, more layers between user and model. Iterative processes to make the answers better, etc, and to create “memory”, which in essence is just an ever growing prompt managed by the agent. But in the end, this won’t fix the inherent problem, so it will only do so much and is already hitting another ceiling. It introduces state decay. With the agent method its not really possible to “take away” memory, so if you gave it multiple versions of the same code (as you would if you work with AI), the AI never really forgets about old code. It can supress it through agent instructions (more duct tape), but the more there is the more it bleeds through, which can make the AI reintroduce old code or base assumptions on outdated things.
There is no fix without changing the inherent way how models work, which would introduce complexity beyond what is currently feasible in computing (and the current AI is already gobbling up all computing reaoureces as is)
As someone that wanted to write but never had the time to learn, what’s so bad about writing code manually?
Seems like if you can learn to do it well, you will be fairly well set with that skill.
If you’re someone who cares at all about the quality and consistency of your craft, there’s absolutely nothing wrong with manually writing code.
If you’re a misanthropic “techno-feudalist” who thinks of code as nothing more than an asset to sell, then pumping out as much code as quickly as possible without any human intervention is a very attractive proposition.
Tech, sadly, is absolutely infested with these people at all levels.
You will still need to learn programming manually.
The process of struggling to understand and synthesize working code is a critical part of learning. Skipping it feels easier, but you’re hurting your ability to understand coding.
Sure, you can make an LLM generate code and if you’re inexperienced it can outperform you on the basic tasks that you’re given as exercises. This is a trap that a lot of students fall into. It’s very easy to let LLMs do the ‘hard work’ part of learning while you just read the textbook or watch a video. Unfortunately, the hard part is the part that builds your skillset.
It’s just like how you can’t just watch a video about physical fitness and then use a robot to lift the weights for you. Sure, you get to the end of your sets faster and you’re not physically tired and sore but you won’t actually benefit in the ways that matter.
There’s nothing wrong with writing code manually. Over the past few months LLMs have gotten a lot better at writing code than they were before, but they can still make weird mistakes.