Will Vibe Coding Kill LowCode? 🤖
The short answer is no, the long answer is still no, but it’s complicated. Buckle up, partner—let's decode this!
The AI Overhaul? 🤔
Lately, I’ve stumbled upon a slew of posts claiming that the Power Platform (my personal favorite LowCode platform) is on the brink of extinction due to the emergence of AI and vibing coding. The rationale is pretty straightforward:
Why draw spaghetti diagrams to create flows when a simple sentence can do the trick? 🍝
Why drag and drop components to build an app when I can just describe one and get a full React app? 🎨
What's the point of learning a LowCode language when AI can interpret natural language? ✍️
But hang on! Remember, these words are coming from industry experts who’ve been in the game longer than I have. However, I suspect that this future reality is further away than some might think. Here’s why:
💰 Cost Considerations
This is the biggie: AI is costly. Not just pricey for consumers, but companies that have gone all-in (looking at you, Microsoft! 👀) are heavily subsidizing AI development, barely making a profit—if that. In comparison, LowCode products like the Power Platform are built to be efficient.
Currently, AI solutions are costly, and they’re set to get even more expensive. Here’s the kicker: the market isn’t following Moore's Law in this race. All major players are opting for cutting-edge models, which are not the cheapest options in the long run.
For example, let’s consider Power Automate. A flow with a hundred actions is essentially free under most Microsoft 365 accounts:
- 100 actions = Free (6,000 actions daily at zero cost)
- Premium full price (most organizations don’t pay full price)
- 40,000 actions every day
- Total of 1,120,000 calls
- Cost: $10
- That’s just 0.08 cents per run! 💵
Now compare that to the Agent Flow:
- 25,000 messages monthly
- 13 messages per run
- 1,923 runs
- Cost: $200
- That’s 10.4 cents per run! 🤯
Yep, that’s 130 times more expensive. You really think organizations will abandon Power Automate for this?! Nope!
Outcomes: The Deterministic Dilemma 🌪️
AI excels in managing complex unstructured processes. Think about it—it can juggle an ever-changing environment of human inputs with finesse. However, there’s a catch: because the inputs are unstructured, the outputs can also be unpredictable.
What does that mean for critical sectors like banking and governance? Will they tolerate inconsistencies when it comes to money, or compliance issues requiring clear-cut processes? No way! 🏦
Organizations prefer to eliminate uncertainty—options like drop-down menus to ensure known outcomes beat relying on AI to