Why H.I. Should Not Be Replaced By A.I.


I have been serving the land development industry with software technology since the late 1970’s. Back then, off-the-shelf software was rare, and every computer came with instructions on how to create programs. Software development started off as a hobby that morphed into an unplanned large business. I never had a course on how to write software, only the documentation that came with the computer on how to write code.

I wrote exclusively on Hewlett Packard Basic which was simplistic but in HP’s form had very strong engineering and drafting applications. This made it easy to develop higher level solutions. Eventually Hewlett Packard contacted me to form a collaboration that lasted through the 1980’s and 90’s. We sold to thousands of engineers and surveyors in both the private and public sectors including the military.

In the 1990’s I directed my financial consultant to NOT invest in any of the ‘Wall Street darling’ software companies that were attracting stupid public money to hire overpaid programming graduates to develop vaporware. This ultimately led to the Dot-Com crash at the turn of the century. In the same time, Hewlett Packard began its pivot from having the most advanced, fastest, and most dependable products specific to the sciences, to a more consumer based company that it remains today – just another mass market PC.

In the 1970’s the software we developed reduced tasks that could take days or weeks to just minutes or hours. In the 1980’s and 90’s we reduced those minutes or hours to seconds for the same tasks. It was at that time I started noticing built land development patterns getting worse–not better–compared to the days when a draftsman drew by hand on paper a subdivision plat. Instead of creating better designs with the tech, they were faster subdivision plats.

I addressed these concerns in a 2010 New Geography article and won’t replicate here what was previously published.

In my opinion, the problem with A.I. is that it’s certainly not ‘artificial intelligence’. It’s a master of plagiarism of information out there on the web maybe, but to be intelligent, it would require coming up with an entirely new idea that has never before been invented; that it is not.

The software industry has dumbed down the entire land development industry. What we need is more H.I. (Human Intelligence) and less A.I.

Here are a few real-world examples:

Architecture. Today, all one needs to do is select ‘typical modules’ in software such as Revit to quickly come up with a house or apartment design. This is one reason why most new homes by different builders all look like the same designer was hired. This is apparent in the sprouting of mid-rise apartments so popular today where a designer can whip out another mindless highly visible structure with the least amount of H.I.

Education. Again, not to repeat past articles, like this 2017 New Geography article, we see graduates in Urban Planning lacking any real world design skills, instead educated more on social engineering. Now imagine the near future, when these same professors embrace A.I. as the ‘design’ tool destroying any hope that H.I. will help eliminate the mundane cookie cutter of today’s growth patterns.

Land Planning. For the most part, land ‘planning’ does not exist today, only mathematical subdividing to regulatory minimums. A ‘paint by numbers’ approach to create the living environments in which we dwell. We wrote about the dangers of automated subdividing in the above ‘lots-per-minute’ race which I predict will get much worse with A.I.

Engineering. We have seen over the last few decades the inability for many of the younger engineers to properly ‘engineer’ a site plan, again over dependence on having a software make the designs. In far too many cases and too often, the clients complain about the excessive earthwork costs, which is ultimately passed onto the home buyer (i.e., you). Before computers (old school H.I.) an engineer would painstakingly design final grading that could take weeks or months of turn-around time to balance the earthwork and minimize costs. Today, we often see earthwork quickly calculated using a pre-programmed CAD system along street grading without taking the time and effort to work and re-work the grading and utilities to locally balance the earthwork. On hilly sites we have seen a street on top of a slope and one on the bottom of the slope completely ignoring the proposed topography of how the homes will sit between the two streets, just contours automatically generated by software. I tell the developer they did not hire an engineer but instead a ‘CAD’ operator that can press automated buttons. A real (H.I.) engineer would work the grades of every home pad along with experimenting with street grades until the earth balanced. Having an old school engineer rework the plan might take a few weeks, or even a month, but could reduce construction costs by many millions of dollars. It would not be that unusual that earthwork could be reduced by $10,000 to $20,000 a lot in such situations. Could A.I. do better? I seriously doubt it.

Are there applications for A.I.? Sure. When I create educational videos, I narrate into Word and make modifications, then I use an A.I. voice for the video. Changes are quick and easy, much more so than if I did the voice work and then needed to make changes. But is that really A.I. or just good programming on voice manipulation?

My fear is that we have been dumbing down certain industries. All one needs to do is look at most of todays suburban patterns, not just more cookie cutter than the past, but most are incredibly wasteful because we have come to depend too much on software and less and less on educating on good design practices and he latest innovations. For those wanting a quick $5 meal maybe 2-minute fast food is OK. For those wanting to invest in a $500K house, a 2-minute design effort not so much, but I’ve witnessed architectural designers whip out a plan that quick. What we need is more H.I to make sure that A.I. is not taking us two steps backwards for every step forward.

One thing absolutely for sure is that A.I. is the current overhyped buzzword that investors will hop on for the next stock market crash. History has a way to repeat itself.

Rick Harrison is President of Rick Harrison Site Design Studio and Neighborhood Innovations. Rick has been instrumental in advancing land planning techniques as well as technology for almost all professions tied to land development.

Photo: Cookie cutter homes, Northwood. by Bernard Spragg, in Public Domain.