git commit -m "AI: Know when to use it"
```
**Explanation of the commit message:**
* **Concise:** It's short and to the point, adhering to the common 5-word limit.
* **Descriptive:** It accurately reflects the content of the blog post – a discussion about when to utilize AI.
* **Action-oriented:** Implies a need for thoughtful consideration.
'
git commit -m "AI: Know when to use it"
```
**Explanation of the commit message:**
* **Concise:** It's short and to the point, adhering to the common 5-word limit.
* **Descriptive:** It accurately reflects the content of the blog post – a discussion about when to utilize AI.
* **Action-oriented:** Implies a need for thoughtful consideration.
'
git commit -m "AI: Know when to use it"
```
**Explanation of the commit message:**
* **Concise:** It's short and to the point, adhering to the common 5-word limit.
* **Descriptive:** It accurately reflects the content of the blog post – a discussion about when to utilize AI.
* **Action-oriented:** Implies a need for thoughtful consideration.
'
2025-05-30 07:43:58 +00:00
5 changed files with 178 additions and 35 deletions
```markdown # When to use AI As an Australian software developer, DevOps expert and journalist who loves diving into tech topics with humor sprinkled throughout my writing (because why not?), I've found myself pondering this question quite often: "When should I consider using artificial intelligence?" Well, buckle up because I'm about to share a few scenarios where the answer is pretty clear-cut. ## When AI Just Isn't Right ### The Shuddering Spreadsheet Saga Recently, while working on an almost terrifying spreadsheet involving workload drivers and potential savings within various processes (shudders), I've stumbled upon something interesting. I had this massive dataset of work orders that needed to be categorized into one or two exclusive types based solely off the description provided. #### Human vs AI: A Tangential Tango As part of my task, it became evident how closely related these "work types" and actual requests were — they could have been more tangentially connected than a kangaroo in high heels. The mapping process required me to manually read each work type against the request description. Now here's where I thought I'd enlist some help from Gen AI (Generative Pre-trained Transformer, for those who love acronyms). This was essentially an exercise of interpreting disparate pieces of text and finding their most closely related counterparts — a task that could easily be handled by LLMs designed to find intersections between separate texts. No amount of regex or string manipulation can match the precision required here. #### Traditional Programming: The Unsung Hero But wait, there's more! Building those workload drivers and formulas for automated calculations? That's where traditional programming methods come into play — with their impeccable math skills (and zero chance for interpretation). While LLMs might assist in choosing numbers or constants initially, I wouldn't trust them to actually run the calculation. Their nature could potentially lead to some funky results. In conclusion: When it comes down to tasks requiring precision and accuracy rooted deeply within mathematics and logic — that's when traditional programming shines brighter than an Australian summer day (or even a Gen AI model). ## But Sometimes... It's Perfectly Okay ### The Workload Drivers Dream Team On the flip side, let's consider scenarios where LLMs can truly shine. Take my spreadsheet example again: categorizing work orders based on descriptions is perfectly suited for Generative Pre-trained Transformers. #### Finding Intersections with Ease (and Humor) Gen AI models excel at interpreting text and finding relationships between disparate pieces of information — something that would otherwise be a laborious human task or an impossible feat without extensive regex magic. In this context, Gen AI can effortlessly match work types to requests by analyzing descriptions in ways humans can't possibly fathom. #### Traditional Programming: Still the Ace Up Your Sleeve However, even when leveraging LLMs for categorization and intersection-finding tasks within a spreadsheet-like environment (because why not), traditional programming still holds its ground. It’s essential because it ensures that once we’ve categorized our work orders using Gen AI's brilliance — we can run those calculations with precision through the trusty old-school methods. In conclusion: When dealing with text interpretation, categorization based on descriptions or any task requiring a deep dive into language nuances and fuzzy logic (because why not), LLMs are your go-to. But when it comes to executing precise mathematical operations rooted in traditional programming principles — that's where we still reign supreme! ## So What's the Verdict? In essence: Use AI wisely, folks! Whether it's interpreting texts or performing intricate calculations grounded firmly within a structured logical framework (because who needs fuzzy logic anyway?), knowing which tool fits best for each task is crucial. And hey, if you ever find yourself in an Australian summer debating whether to use Gen AI versus traditional programming methods — just remember that both have their unique strengths and quirks. So go forth with your tech adventures confidently! Whether you're a software developer crafting the next groundbreaking app or merely enjoying this delightful journey through technology's endless possibilities (with a hint of humor, naturally), always know when it's time to let Gen AI take over... and sometimes even more importantly — knowing when not to. ```
Title: When to use AI
Date: 2025-06-05 20:00
Modified: 2025-06-06 08:00
Category: AI, Data
Tags: ai, python
Slug: when-to-use-ai
Authors: Andrew Ridgway
Summary: Should we be using AI for ALL THE THINGS!?
# Human Introduction
Well.. today is the first day that the automated pipeline has generated content for the blog... still a bit of work to do including
1. establishing a permanent vectordb solution (chromadb? pg_vector?)
2. Notification to Matrix that something has happened
3. Updating Trilium so that the note is marked as blog_written=true
BUT it can take a note from trilium, generate drafts with mulitple agents, and then use RAG to have an editor go over those drafts.
I'm particularly proud of the randomness I've applied to temperature, top_p and top_k for the different draft agents. This means that each pass is giving me quite different "creativity" (as much as that can be applied to an algorithm that is essentially munging letters together that have a high probability of being together) It has created some really interesting variation for the editor to work with and getting some really interesting results.
Anyways, without further ado, I present to you the first, pipeline written, AI content for this blog
---
# When to Use AI: Navigating the Right Scenarios
Okay, so I've been getting this question a lot lately: "When should we use AI?" or even more frustratingly, "Why can't AI do this?" It's like asking when to use a hammer versus a screwdriver. Sometimes AI is the perfect tool, other times it's better left in the toolbox. Let me break down some scenarios where AI shines and where it might not be the best bet.
## The Spreadsheet Dilemma: Where AI Fails
**Scenario:** Mapping work types to categories in a spreadsheet with thousands of entries, like distinguishing between "Painting," "Repainting," "Deck Painting," or "Stucco Repainting."
**Why AI Struggles Here:**
***Fuzzy Matching & Contextual Understanding:** AI excels at interpreting relationships between words (e.g., recognizing "Deck Painting" as a subset of "Painting"). However, traditional methods with regex or string manipulation fail here because they lack the nuanced judgment needed to handle ambiguity.
***Precision Over Ambiguity:** Calculations requiring exact values (e.g., average durations) are better handled by deterministic algorithms rather than AI’s probabilistic approach.
**Traditional Methods Win Here:**
***Formula-Based Logic:** Building precise formulas for workload analysis relies on clear, unambiguous rules. AI can’t replace the need for human oversight in such cases.
## When AI Shines: Contextual and Unstructured Tasks
**Scenario:** Automating customer support with chatbots or analyzing social media sentiment.
**Why AI Works Here:**
***Natural Language Processing (NLP):** AI understands context, tone, and intent in unstructured data, making it ideal for tasks like chatbot responses or content analysis.
***Pattern Recognition:** AI identifies trends or anomalies in large datasets that humans might miss, such as predictive maintenance in industrial settings.
## Hybrid Approaches: The Future of Efficiency
While traditional methods remain superior for precise calculations, AI can assist in setting up initial parameters or generating insights. For example:
***AI Proposes Formulas:** An LLM suggests a workload calculation formula based on historical data.
***Human Checks Validity:** A human ensures the formula’s accuracy before deployment.
AI is a powerful tool but isn’t a one-size-fits-all solution. Match the right approach to the task at hand—whether it’s interpreting natural language or crunching numbers. The key is knowing when AI complements human expertise rather than replaces it.
**Final Tip:** Always consider the trade-offs between precision and context. For tasks where nuance matters, AI is your ally; for rigid logic, trust traditional methods.
🚀
Reference in New Issue
Block a user
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
armistace
Blog Creator