Art of Message
GI/GO
When you use LLMs to code, especially if you’re a sh*tty coder like myself, you might feel beset upon by the Garbage-In/Garbage-Out principle. My experience here is mostly in:
- Copilot/Starcoder
- ChatGPT with GPT3.5/4
- OpenAI API tools usually on GPT3.5
- Claude+/Claude-100k
BTW, coding is where LLMs most impress. I say that having made 15,000 prompts a month since last fall.
But I feel that my LLM-coding experience is unusual – partly because I’m so low-level; it’s hard to evaluate the code and advice LLM gives me. In contrast, I can detect value from LLMs as they predict, categorize, critique, ideate, graph, write, etc.
But also, I’m not the “ideal customer profile” – I’m a generalist full-stack designer/developer, not an actual professional developer. I just want my tool to work, be simple to use, and be secure – I care nothing about things like efficienc, scalability, or other best practices that don’t directly benefit my current users.
But the LLMs are OBSESSED with software development best practices.
Why? Because the authority-nerd-majority prevails over the content on which major LLMs develop foundational knowledge. (For example, 11 million C4 tokens are from StackOverflow.) And as professionals, of course these people obsess over efficiency and scalability – and “separation of concerns” (eye-rolling hard right now). And so LLM coding assistants inherit that bias.
But just because a bias for professional software development best practices is painful for me, and wastes my life, doesn’t mean it’s objectively bad. Engineers do need to make sure buildings don’t collapse and personal data is secure.
I’m just using the wrong tool for the job.. albeit the best available.
LLM products have 4 places, IMO, where they can take in “garbage”:
- Pre-training, as on the C4 cited above
- Training/fine-tuning
- Hardcoded prompting (from product owner)
- Non-hardcoded prompting (from product user)
But here’s the thing – one man’s garbage is another man’s treasure: the challenge will be affordably creating generative AI that works for you and enough others like you.
