What's your AI Methodology?

Discussion
1

Not the Developer – New Paradigm shift in Projects
Recently I posted about my experience as a newbie using Cursor. In one response from SoMaCoSf, he claimed that he was never a developer while describing his career path. I claim to be a pseudo-developer because I was forced into squeezing out of technology what I needed to prove whatever concept I was introducing. Once that concept was accepted it was handed to developers and users. Each has their own focus on making the concept useful from their perspectives.
In 1993, I created a prototype application using Lotus Notes. This was a major technical shift from mainframe centric data control to a distributed model with users becoming proficient at creating supporting databases. These folks were not developers. They embraced the opportunity to expand their skills sets while learning about technology. They were, in my opinion, the first knowledge workers. Meanwhile, I took my Lotus Notes server out of the technical closet they put me in and handed it over to the production control staff/computer operations team.
The point of this reflection is that there are two distinct approaches to using AI tools. Cursor for developers and Cursor for knowledge workers.
Developers are concerned with syntax, code structure, rules, production controls and security. This is an important role in protecting a company’s digital assets. It is traditional and has been with us since the advent of computing.
To me, knowledge working is the more intriguing and compelling case for utilizing AI tools. During my career, one of the underlying goals of technology innovation was to provide “users” with the tools to make them knowledge workers. The issue has always been the burden of understanding all of the technical nuances associated with any tool set or application on the business user set against the lack of business understanding from the technical developer.
Even though these issues were addressed with the delivery of project methodologies, there was still a disconnect between user and developers. Organizational towers existed back then, and in many instances still exist today.
Out of the short time that I have used Cursor, I realized that there is a new project methodology evolving where knowledge workers using AI tools can rapidly develop quality business solutions. There still is an important role for developers in maintaining technical infrastructures, validating component competence, ensuring production accuracy and providing vigilant security. Both of these groups contribute to the on-going evolution of AI tools. What’s needed is a pathway for them to follow together to assure optimum results.

The learning activity that needs focused attention is still the same as it was 50 years ago.

  1. How do knowledge workers and technical developers communicate effectively?
  2. How do we remove the stigma of us verses them?
  3. How does an AI project methodology support all participants in becoming useful contributors?
  4. How can AI change traditional management philosophies? (Most difficult aspect)
    This document is an example of the impact of an AI tool like Cursor has on me, someone with extensive business and technical background. The possibilities are endless so we need to start with some “wagon trails”.
2

Great post, really like the detailed explaination you made and that people without coding experience use AI tools in such a more involved manner than simpler AI chat apps.

Here is my take on the points you asked about:

  1. Good communication is overall required and the best way is to use correct terminology. That does not mean making knowledge workers developers, but to understand the general meaning of what is being said so the communication can happen effectively.

  2. Stigma of us vs them is man made and artificial. It comes from fears, misconceptions, misunderstanding and perhaps from not communicating enough, no matter which side. It actually disappears when people communicate properly.

  3. There is no separate AI project methodology needed. Any decent project management and task breakdown works for coding and non-coding tasks. AI integration can use any project methodology if you just ask it to do that.

  4. Here i would also say that there is so far no change in management approach needed when using AI. Mostly AI enhances peoples abilities by automating steps that were done in a much more manual way. It does not replace critical thinking and this is the case when people rely on AI too much at this stage as it does not think at all. It just responds with a most likely correct answer, that depening on input information can be totally wrong or quite right as well.

From what I see here in the forum, Cursor is used by people without development experience, naturally developers but also people in non-coding roles like marketing or similar.

It can literally be used for any kind of writing where you want to organize information, do research, planning and the actual writing as well.

You are right in saying that a methodology is needed!
Would also like to see what experiences others have.