Where is the real value of asking ChatGPT?

After a number of weeks of using ChatGPT as a subscriber, its boundaries in terms of the value it can provide me specifically are beginning to become clearer. Sure, it can do a whole host of things from coding to creating Midjourney prompts, but in terms of where the value is for a knowledge worker the main benefit is in being able to organise and structure information in a variety of styles at lightening speed.

The system has still yet to make the leap to fully imaginative responses, or at least I've not seen them yet. For example, yesterday I posted an article Embracing Change as a Knowledge Worker in a post-AI World. It was suggested to me (somewhat tongue-in-cheek I think) that I should ask ChatGPT what to do. So I did, here is my prompt and the response, as you can see, logical, but not rocket science.

The initial response is quite basic, however experiments I've done on taking each of the steps and asking for suggestions on how to complete them does begin to draw out useful information.

Let's run a quick experiment. I have already decided that I need to focus on a specific niche of the construction sector. I am going to be looking into Social Housing. There is a massive effort to plan the retrofitting of the old housing stock, the government are providing funding to support the change and there is capital awaiting allocation from institutional investors and pension funds. More about this in the <Entering the Market Discovery> article.

So, I put the following prompt into the same thread:

Taking the first point "Market Research and Identification (Months 1-3) Conduct thorough market research to identify key players and trends in the construction sector supply chain." please provide more detail to help with the process.

Include a list of key players with their URLs that should be researched in more depth and provide a justification on why they have been included.

Also, outline trends that are underway and could be explored further within construction sector supply chains. Focus where possible on areas where 'walled gardens' exist and niche programme oriented delivery plans covering a number of years of planned activity are underway.

Here is the response:

Now this is getting closer to being useful. The first round of questions provided information that was logical but not particularly helpful for anyone with more than a basic understanding of how to explore new markets. This second round of responses is now providing information that would be useful to someone new into the sector. Although, again, there is nothing new in this for me as yet. I've worked on projects that have encompassed all of the trend areas mentioned and know of all the key players mentioned and in some cases have worked with them on projects. However, it has highlighted some interesting areas to narrow down on further.

Looking at the trends as opportunities to explore in turn:

1. Digitalization ((?) Digitisation) and BIM: This area is too expansive. It has vertical application across the supply chain and has very few points at which value could be added beyond that already provided by key players.
2. Offsite Construction and Modularisation: A fascinating area of innovation and one that presents enormous potential for cost saving, waste reduction and efficiency improvements (see Laing O'Rourke's Explore Offsite Factory) However the scale is too large to make for realistic value add opportunities.

The following three are however, the areas I had shortlisted for further exploration; Green Building and Sustainable Materials, and Smart Buildings and IoT (a lot of innovation by a multitude of companies here), and PPPs (here is where the walled garden concept comes into effect).

So, as a final test, let's drill down one further layer with the following prompt:

As the basis for further research into PPPs, what sources of information are publicly available on existing and prospective PPPs along with detail on consortium participants?

Finally we're getting to something that I can get my teeth into. So, in conclusion; the value of asking ChatGPT is in asking a series of nested questions that drill into the topic in question and relate responses to external sources of information. At least in this case...