Tech Debt is a lazy concept

I’ve stopped using the term “tech debt,” and I think other technology leaders should consider doing the same.

Current AI systems are remarkably capable, but they are essentially fixed points in capability space. They can reason, chain thoughts together, and reflect on their outputs, but they can't fundamentally change from experience in the way humans do. This is a critical ability to achieve Artificial General Intelligence (AGI).

The stage7 Agent project has had several enhancements added this week including:

new React UI with more visualization of the live agent network, performance metrics for the interactions with LLM,
the  LLM interface which can hold a single conversation across multiple LLMs now supports self-hosted LLM support (OpenWebUI), Mistral, and Groq in addition to Anthropic, Gemini, Huggingface, OpenAI and OpenRouter which were supported in the initial version,
more content forms (text to/from text, code, image, audio, video),

Asynchronous meetings are a great idea. They allow participants to think through their contributions before sharing, eliminate floor hogging, capture everything exchanged, and reduce unproductive time for participants. But there are additional benefits when the technology goes beyond sharing messages and tagging actions and decisions.

Here is what you should look to achieve with asynchronous meeting technology:

- Including AI participants with crucial skills/expertise that contribute to the discussion;

When I first went into business for myself, there were a few principles that I followed that were most helpful. They are:

1. Buy something that can be improved rather than start from scratch,

2. Look for a business that doesn't have remote and deep-pocketed competitors,

3. Find something big enough to survive early missteps.

Minimizing risk, avoiding strategic challenges (such as international wage differences and brand inequity) and recognizing that my own learning curve was an issue significantly reduced the demands on my attention.