Even before AI, our industry was already getting battered

Just a though dump here guys - share whatever you think even if it’s not directly related to what I’ve posted :slight_smile:

I don’t do as much rendering/visuals/artistic stuff these days (partly because AI) - I think how AI affects that realm is at least for now, worlds different versus “spec’d” design work

For those of you who are genuinely busy right now (envious!!) - instead of reading this post/rant, post your strategy for navigating the AI boom.

I do 2D CAD, 3D modeling, and programming - I’m not much of a ‘real’ programmer (getting better and better) - but I do enough reading to know more about the software industry than my programming skill level would infer.

I’m not worried about AI replacing the work that I actually do (not yet):

If AI can model something correctly, it was probably incredibly easy to model in the first place (excluding art/creative stuff). We see people spitting out entire buildings. But are you actually going to use that? Are you actually going to a client without checking it? Or worse yet (for us), send it to the next person in line and expect them to find all the flaws?

Checking is the hardest part of design and drafting. In the past, these people were the highest paid in the industry. Something changed and people just checked less and less. And the trend where a person could save a few minutes could mean costing someone further down the line a few extra hours.

Many of use have already dealt with this: Ever take on a job where someone has “already done most the work” - All too common is being handed a client’s Sketchup model. Not to dunk on Sketchup (lots of great work comes from it) but the learning curve is very favorable if you don’t have a clue.

Whether we know it or not, we’ve actually been facing many different issues in our industry the last several years.

  • False marketing claims about products which, in turn, raise expectations - Some years back they claimed BIM was something like 10x faster than CAD. It wasn’t. In the office I worked at it was in fact about 4x slower. A few of us worked long hours with no extra pay to help out the BIM manager. The BIM manager was actually part of the problem - they were a phony and set up the projects poorly - and little did we know that they were actually throwing us under the bus behind our backs.
  • Perpetual licensing and the relative variability of software costs - Perpetual licenses probably hurt freelancers the most, especially if you live in a country where they can actually enforce software piracy. You now pay so much for software that you can’t upgrade your PC. It’s created an artificial ceiling for us.
  • Outsourcing - More affecting those of us who freelance, outsourcing can equate to fair competition. Agencies often spam and swarm potential clients to the extent where it makes it significantly harder for you to match up with the best clients, and vice-versa. Agencies (and individuals) may also be posing as clients themselves… hence my next point…
  • RENT SEEKING - In upper-case because this might be hurting us more than AI right now. For 2D CAD related jobs is happening, but as of now I’m not quite sure the extent. I’ve found that with Rhino we often make our own opportunities and are somewhat immune (but not entirely). Rent seeking in software development is off the charts right now. Basically we’d be dealing with a intermediary who not only contributes zero value to the project, but degrades it; they have to limit information should the client or person actually doing the work discover what’s up… This is actually what I’m most curious about right now and an example of AI having a negative effect on you in a way you maybe hadn’t even though of.

Add to all the above, economic conditions, getting older and slower (at least for me), huge growth in real legitimate talent/competition, and whoever knows what else - we’ve already been battling it for years.
Not only relating to CAD, people are showing less and less respect to those that do the actual work. Too many people want to manage and be a ‘boss’ and they want it to be effortless. Sometimes they’re even willing to sacrifice huge amounts of profit for that.

My 2 cents regarding AI:

When I’m smart about how I use AI it helps me quite a bit. It’s naturally not hard to go overboard. Case in point: Letting the AI write too much code, and losing my understanding of what the code is actually doing. For CAD and 3D modeling, AI CAN’T do what I do. It can ostensibly appear to do what I do, but checking and fixing the work would take longer than just modeling it myself.

There’s a different between relying on AI or using AI to assist your work. I’m hoping I remain more towards the later. Smart use of AI makes me more competitive and in theory allows me to offer better value to clients, but at the same time, one way or another, AI is going to make it harder to match up with those clients.

I guess my big point here is that we’re already experiencing (or have experienced) several key shake-ups. In addition to finding ways to prevent myself from becoming irrelevant, I’m also finding ways to navigate around all that non-AI related stuff.

“our industry” - which industry sector are you referring to?

Bad term to include in the title! Everything ounce of energy I had went into the post and the title was an afterthought… I guess…

Maybe our skillset… or our services might have been better.

Different industries require different skills :star_struck:

Of course. But there are huge amounts of overlap, and with Rhino being so versatile…

Architectural modeling has made me way better at modeling something like a vacuum for example - not good enough to make money at industrial design, but much better at it compared to when I only used AutoCAD. If I really really wanted I could potentially move my focus towards that area.

Think of setting up your layouts - whether I’m making a PDF for that vacuum or an interior elevation, both tasks also have a lot of similarities.

Most of what we do in Rhino bears some similarity and likewise, AI has the potential to affect each and every industry relavent to Rhino users.

The probability of the vacuum designer losing work to AI is similar to me losing work to AI.

Well, industrial design is not about “modelling a vacuum cleaner”; there are many brand related, user-centric, sales region related, supplier component related, and technical as well as production parameters that influence a product’s development process, even that of basic consumer products like domestic appliances, particularly that of innovative ones.

I’m curious what an industrial designer would be experiencing I guess. Especially if they’re a freelancer. In Architecture, AI is definitely having an effect on rendering for example. Although it seems unrelated, excessive “lead generation” is being driven by AI. Lots of things are nipping away all at once. I think the modeling skills themselves aren’t replaceable, and product visualization isn’t the same as architectural renders. But there’s lots of different things syphoning value away from what would otherwise be their compensation for their talent.

I would recommend Cory Doctorows latest book about centaurs. It touches many points you mentioned.

Depending on how you use AI can make you go further or stop you completely.

Most fear mongering of AI boom is oriented towards investors, to make them think that AI can capture all artists, designers, coders, drivers, doctors… market. So their suffering is just advertisement for investors. While in reality AI usage just makes it more expensive to produce quality work and makes people lose skills.

AI is designed to create as statistically plausible results as possible, so mistakes are that much harder to find. Starting from scratch most of the time will be more productive. Or adapting old fleshed out design for new project.

What I find usefull is for AI to search old email chains for decisions made years ago; to clean up ifc models for import as references; creating nicely excell spreadsheet or websites to share project state, create addons for software as personally taylored drafting helpers…

If some AI work is good, its only because most of the work was done by person. Even most detailed prompt will not have enough context to make each pixel, line, shape intentional.