It’s frustrating to the point of being funny how simultaneously impressed some people are with AI and the reality of LLMs.
There are people who are convinced the AGI is here right now.
Others are convinced that either OpenAI or Anthropic are “going to build AGI”.
I’ve been a software developer for decades. And I can tell you that it’s frustrating to hear this commentary.
To hear the lies. To hear the marketing.
Because I see the trajectory. I use these models every day. And they cannot do what I can do. I know they will never be able to.
The general gist of what the AI community is trying to achieve is directionally correct.
We’d prefer the computer simply wrote the code for us.
But LLMs cannot replace developers because transformers are simply broken.
They’re too stupid and always will be.
Working with them on any serious task quickly reveals their limitations.
I’ll give you an example. I’m trying to get the Dez codebase to properly use SBP with MVC. The model doesn’t get it, even with examples of code I’ve written.
Why?
It doesn’t get it because nobody but our team at okTurtles uses SBP.
It’s too new. There’s not enough training data on it. It’s a concept these LLMs have never encountered before.
There are many such concepts. And as we move forward, the frustrations will continue.
Ordinary people don’t care about this. They just want to see a pricing page that looks good.
They don’t care about how the code is structured. So they’re thrilled and they love to tell me about how thrilled they are.
Meanwhile, I’m fighting with this dumb clanker trying to explain SBP MVC principles to it, and it’s just not getting it, and I know that it never will. And even if it does, then still, it won’t. If you understand what I mean.
😪