Microsoft’s Seven Pillars of AI Readiness – but where’s the wisdom?

Skyline view of Paris with Eiffel Tower in background.

Jerry Bartlett of Content Pro Tech Ltd delivers another guest post for the Firehead blog. This time, he’s looking into the role wisdom plays for AI readiness.

There’s a well-trodden quote from the systems analysis world that offers: “Information is not knowledge, knowledge is not wisdom…” 

I thought of this after reading Content Wrangler Scott Abel’s Substack post about Microsoft’s seven “key” pillars of AI Readiness Assessment. Microsoft’s assessment “measures an organization’s preparedness for AI adoption across seven key pillars”. In his Substack post, Scott reports, with some consternation, that Microsoft only vaguely touches on documentation in considering AI-readiness. To paraphrase him, has Microsoft constructed seven pillars but forgotten wisdom?

I’d like to use the rest of this short piece to take some of the points Scott makes and add my own slant. It’s a fairly deep dive by Scott. You can read it here: 

https://open.substack.com/pub/thecontentwrangler/p/technical-documentation-the-missing?r=69zf&utm_campaign=post-expanded-share&utm_medium=web

I am a paid subscriber, so I’m not sure if this post is behind a subscriber wall.

Right at the outset, Scott puts over his core opinion: “AI readiness is a documentation problem wearing an AI Halloween 🎃 costume”. And I agree. (I lack the cultural background to comment on the Halloween part, but it can be a pretty scary problem.)  

Of course I agree with Scott. My company is called Content Pro Tech, and I’ve developed a program with Firehead we are calling The Content Forge , which is about transforming your content so it is AI-ready. Here, I take “content” to be a generalization of Scott’s “documentation”.

But, as baffling as not taking content into account may seem, Scott points out Microsoft’s take is a different kind of readiness. It is an organizational readiness that is only tangentially connected to content. 

If I’m following this logic, it suggests that, at enterprise level, it would be cavalier to go ahead with AI without considering Microsoft’s pillars: business strategy, governance and security, AI strategy… and the rest. I’m fairly sure, if you follow Microsoft’s assessment, you will receive some useful recommendations. But as Scott alludes to, will you arrive at knowledge, let alone wisdom? Because after all, wisdom (insight, if you like) is where the value of an organization lies, not merely in the information (data).

Here are Microsoft’s “Seven Pillars” of organizational AI readiness:

  1. Business Strategy
  2. AI Governance and Security
  3. Data Foundations
  4. AI Strategy and Experience
  5. Organization and Culture
  6. Infrastructure for AI
  7. Model Management

Scott acknowledges that these Seven Pillars deal with areas that documentation and content people have little or no direct control over. However, they/we do have a great deal of say in how knowledge is built into an organization’s content. We do this through good content structure and provenance, and by creating dependable and identifiable content building blocks for the content.  From there we can add in semantic context, in effect, we build in what it all means and whether and how it can be used – and who does it. Only then can we say with any confidence that AI can produce output that is of any value (that is, knowledge) reliably.

Our approach at The Content Forge is to transform content and ways of working with it, so that content, including technical documentation, works reliably with AI tools over time. What we call AI-friendly content. 

The first step is to get to content AI-readiness. This is where content is well-structured, in short: organized into distinct, reusable components, and identified with metadata tags such that each component can be used as required in its correct context, clarified and made consistent. This makes it easier for AI tools to process it reliably and produce consistent, accurate outputs. 

What we do to make AI-friendly content is no secret sauce, but it is an extra element. We think calling it “AI-friendly” makes a dry-sounding term relevant to AI, and well, more memorable. That dry term is content operations

But specifically, it is content operations with the intent of maintaining content AI-readiness. It is designed to build a way of working with AI-ready content that ensures our hard-won state of AI-readiness is maintained over time. Content keeps being self-consistent, well-structured, and semantically rich, so AI tools can work reliably over time: ongoingly (because that’s a good, succinct word!). A commitment to AI-friendly content is a way of acknowledging to yourself that you’re in it for the long term! 

Content operations is a  systematic way of managing how your content is created, maintained, and used. It’s a way of working that informs what we do and when. Specifically, in relation to AI-friendly, we mean using all the tools at our disposal to keep our content AI-ready: consistent, relevant, and in sync with your workflow.

By all means, use Microsoft’s Seven Pillars of organizational AI readiness, but wisdom would be to build those pillars on a foundation of AI-ready content. And on top of that to build in content operations to make your content AI-friendly would be especially wise.

Are you curious about how AI-friendly content operations can help your organization for the AI long-run? Setting your content up for this can be tricky, especially for SMBs. Where are you on this journey? Let us know in the comments or contact us here for a chat.

The Content Forge. AI-friendly content operations for humans.

Leave the first comment

CJ Walker

Related Posts

Call to action

Introducing The Content Forge 🔥

AI doesn't fix broken content. It amplifies it. The Content Forge helps you build AI-friendly content operations that actually work. Find out where to start....

CJ Walker

Unlocking New Career Paths: How Termbase Skills Empower Technical Communicators

Post 9 in our series on localization skills for modern technical communicators Your translation vendor sends you a quality report. The same product feature appears under six different labels across your documentation. The translators resolved the inconsistency independently, producing six…...

CJ Walker