Buried in a talk on AI from an artist who is doing cutting-edge video work was the following nugget that entirely sums up the zeitgeist: “The tools are changing so fast that artists can’t keep up with them, let alone master them, before everyone is on to the next.” And while you might think that this concern is only relevant to those who have to stay on the crest of the hype wave, the deeper question resounds with every hacker.
When was the last time you changed PCB layout software or refreshed your operating system? What other tools do you use in your work or your extra-curricular projects, and how long have you been using them? Are you still designing your analog front-ends with LM358s, or have you looked around to see that technology has moved on since the 1970s? “OMG, you’re still using ST32F103s?”
It’s not a simple question, and there are no good answers. Proficiency with a tool, like for instance the audio editor with which I crank out the podcast every week, only comes through practice. And practice simply takes time and effort. When you put your time in on a tool, it really is an investment in that it helps you get better. But what about that newer, better tool out there?
Some of the reluctance to update is certainly sunk-cost fallacy, after all you put so much sweat and tears into the current tool, but there is also a real cost to overcome to learn the new hotness, and that’s no fallacy. If you’re always trying to learn a new way of doing something, you’re never going to get good at doing something, and that’s the lament of our artist friend. Honing your craft requires focus. You won’t know the odd feature set of that next microcontroller as well as you do the old faithful – without sitting down and reading the datasheet and doing a couple finger-stretching projects first.
Striking the optimal balance here is hard. On a per-project basis, staying with your good old tool or swapping to the new hotness is a binary choice, but across your projects, you can do some of each. Maybe it makes sense to budget some of your hacking time into learning new tools? How about ten percent? What do you think?
I still use a dubiously-obtained Adobe Photoshop from 2003 for all my graphic work. I know it very well and it’s totally free to use forever. It can’t save to whatever the latest format hotness is, but it can work in jpgs, pngs, and even animated gifs, and that’s good enough for me. All you suckers paying a subscription to Adobe need to ask yourself — do you really need AI background replacement?
Yes I do. I actually have a job where I use that stuff. And I regularly need like five or six other things from the suite.
At least I can cancel my Autodesk subscription and use Blender, which has finally caught up enough to be a viable (almost) full replacement. I know that one day Autodesk and Adobe will fuse into one big obvious monopoly cartel and nobody will do anything, but that day is not today
Surely 95% of what you use Photoshop to do can be done by a version from 2003, but I get it, if you have a professional need, pay for subscription. I tried to use the Gimp for a month when I was in Costa Rica with a work laptop I couldn’t get 2003 Photoshop to install on, and it was most unpleasant. But I’m a software developer and only need Photoshop for quick and dirty buttons or whatever — when I’m not doing the usual cropping or color balancing on personal photographs.
Oddly enough I did time in the US for drug trafficking and as far as my valuable time, what makes it valuable is that I can do what I want with it. Amazing how that works. Oddly enough I am often times a proponent of buying things that are off the shelf solutions when they make sense. We all have our own decisions to make what it makes sense to farm out and what to do yourself.
Also consider cars, I drive cars, trucks, construction equipment and motorcycles. If felt a repair was out of my experience beyond my abilities to safely carry out, do you think I would want me driving around in it? I am sure most people feel the same way. I won’t disagree with you that many people are not too concerned about the rest of the world, but in this case, their ass is on the line too.
You make an excellent point about the value of personal time, something I must price extremely cheaply in my brain, because I always want to do EVERYTHING, even if I don’t know how to do it. I’ll watch a YouTube video and then just do it. Usually this works out okay for me, but wouldn’t my time be better spent doing something I am extremely good at instead of things that are going to take me twice as long. But then I call a plumber to install a sensor on water heater at a rental property and he quotes a figure of $1200. So I buy the part and spend 20 minutes doing it myself — the money I just saved is much more than I could earn doing what I do professionally.
Oddly enough, I use GIMP for all my photo manipulation. I’ve been using it since about 1996.
We have a licensed copy of Photoshop at work. My attempts to use it were most unpleasant. Nothing is where I expect it. The names of the menu entries are all funky. Nothing matches what I expect.
It is in very large part just what you are used to – but that is a VERY large part of it.
100% the same. I have all the latest and greatest Adobe apps through work, but all they do is clogging disk space, because I know my way around Gimp, Inkscape and Kdenlive, and Adobe feels incredibly nonsensical and hard to use.
Yup, background removal saves hours, literally.
But also the other tools in the suite move on, and INDD you do have to keep updating to be able to exchange files.
Can’t agree more. I still use an older version of photoshop. Learned it well, I always get the job done efficiently. Blocked it from calling home and never had any issues. Never had to waste time when opening the app that it would block it’s use until I upgraded which would interrupt my workflow. Would never upgrade to a subscription model. Getting milked for using an app is pure profiteering without any real benifits. Keeping my computer apps isolated from calling home has kept my workflow fluid. Obviously not all apps can be blocked if specific internet use is required or a major fix is necessary for repairing annoying bugs. Online file conversion or separate apps resolve compatibility issues when working with newer photoshop formats. Being technically savvy helps but colleagues who aren’t succumb to purchasing the subscription model. Unfortunately.
I think I look at it from the other direction. There are solved problems and unsolved problems. Solved problems, for me, are the CAD software I use and the Circuit design/layout software. I’d still be on Eagle if it weren’t for that damn company.
And then there are unsolved problems, like the actual project I’m working on and how to get this part to interact with that part in just the way I want.
On the other hand, I completely abandoned hand written code once I tried AI. Now I just ask the Great Intelligence to write the great majority of my code for me and then the only problem to solve is tweaking it to be exactly what I want.
So no. and yes. If the tool is compelling enough it’ll get into the mix.
What kind of code do you write? Do you have a github repo or something we can see the code AI generated? I’m curious because the time I tried to use it to write embedded systems code it wasn’t so helpful (of course I was new to using it too).
Almost all the code I need is simple Arduino stuff. The last one was an ESPNow implementation and a Python-based UI to match. Both of which, aside from simple changes, it nailed easily.
Claude.ai works well for me.
I haven’t written real code since SQL was in its infancy :-)
On a smaller level, a retired co-worker found some 8mm home movies and wanted my help in putting them on DVDs for Christmas gifts. I used the Windows movie maker from Windows 7 and 8 on 2 different PCs to do the deed. I found the video stabilizer from 8 works a lot like it used to on YouTube before they removed it. Whatever codec is needed is present on Windows 7 so I used it to do the burning. I know there are way better ways, but I’m working with what I’ve got and my co-worker was able to hand them out as priceless gifts to his family.
This, a little awkward but it works and is pretty fast so why not.
On the other hand, learning new tools is usually a net good in any case but mostly rewarding when the new tool is actually user friendly. Doing that complicated and highly accurate interrupt routine on that atmel 8-bit might feel at home. But on the other hand, doing the same thing on the pi2030(? Numbers?) might be simpler to the point that you’ll spend an hour either case? And if you’re putting a 100 hours into something then 1hr here or there does not really matter. You might even save time doing both.
Though myself I usually go for the cheapest option “this should really be doable for less than 5$..” and then never really get done :) mostly because preoccupied with life in general. But doing something cost effective usually also takes more time so.. that doesn’t help getting done.
Tools change. Good design sense accumulates.
in 2007 i bought a license of zbrush v3.1 for $489
For 15 years Pixologic gave me free updates.
Maxon gave me one final update, zbrush 2022.0.8 after they purchased it from pixologic.
The upgrade to Zbrush 2023, the last perpetual license offered, was priced at $659
Thats $170 more than I originally paid for an update, not a new license.
Now Zbrush costs $400 a year, Thats ONE year for only $89 less than i paid for my first 15 years
so Ill be using zbrush 2022 forever, or at least until Maxon sells it to someone else and they start offering a reasonably priced upgrade or perpetual license.
I stuck with Rhino 4 from 2007 until 2020 when I scraped part of a stimulus check into a rhino 7 upgrade. Im still learning to use the new bells and whistles that brought me. At $600 to upgrade It will probably be another 3 versions until I upgrade again. Rhino 10? maybe? LOL
Ive still got a set of socket wrenches I bought in high school (1989) to work on my first car.
If a tool still gets the job done why replace it?
Wait until you see how much literally everything else has increased in price since 2007
I keep up to date with Rhino since I use it heavily, and it always has areas for improvement, which McNeel do address, if not on the timetable I’d prefer.
That being the case, subscription pricing wouldn’t make much of a difference financially (and the same was true with pre-CC Adobe products). The bigger problem with a subscription model is that the software ends up as a perpetual beta, because developers aren’t working towards a specific release that has to justify a specific price. That’s surely a big part of Adobe’s quality control going from hero to zero in just a few years.
The problem is, Rhino is going the same way. They increasingly rush out the latest version before it’s release-quality, and then they’ll reply to bug reports by saying they’re focused on fixing it in the next release.
It seems like an iron law of software that after the first couple of years, quality only decreases.
I disagree about rhino rushing out the latest versions.
Rhino has always had their WIP next version available to existing users, and has always relied on its user base for debugging and guidance.
No, They just dont roll out incremental PAID updates for every bug fix, but the WIP version is updated weekly.
At this point theyre doing an official update every 2-3 years, so thats 2-3 years of development with debugging and community feedback. Not bad for a small employee owned company.
And as for subscription pricing not making that much of a difference???? I just showed that under Pixelogic I paid $489 and got 15 years (180 months) of use with free updates.
While under maxon that would get me just shy of 10 months use.
With a release schedule of 2-3 years between versions Rhino would have to price their subscription at $16-$25 for it to “not make much of a difference financially” to always keep up to date. But given my tendency to skip 2 versions before making the 3rd upgrade they would have to charge $6-8/mo for it to “not make that much of a difference” to me.
Similarly, I BOUGHT adobe Substance Painter, Designer, and Sampler on steam for $450. That trio costs $240/yr to rent directly from adobe. This July I break even with subscription cost and will enjoy free use from then on.
Subscription pricing DOES make a BIG difference. If it didnt software companies wouldnt be switching from BUY to RENT models of distribution.
I prefer to spend my time on the creation side (projects) and invest in learning a new tool/flow only when the existing one no longer works for me. I also ride OS versions a long time… I understand this puts me out of the cutting edge but I am still making cool/useful devices and being very productive. As I’ve gotten older I also understand I don’t have to follow every new thing. I’ve seen a lot of “new” things come and then die.
For example, I can acknowledge how cool Kicad is and understand many of the advantages of using it but my old version of Eagle still suffices and I’m very productive with it which allows me to complete projects faster. Someday I’ll have to move off of it and I will without a second thought (and I bet that Kicad is even better then).
I persisted with Altium for way too long before gritting my teeth and moving to KiCAD. So glad I did. Once learned I actually found it better – but personal preference: they are both effective tools, just that one unavoidably costs and the other is “free” (I donate to KiCAD because I use it for work).
I use an image viewer that was last updated 21 years ago. Apart from not understanding webp and non-English character filenames it works perfectly.
Bruce Lee supposedly said something like;
“I would rather face an opponent who practiced 1000 punches once,
than face someone who practiced one punch a thousand times.”
I tell my kids to learn a new thing every morning, and nothing worse will happen to you all day long. That said some of my favorite tools (hand planes and chisels) I got from grandad and he got from his dad. They are more than a hundred years old and still working just fine. They sit on the bench next to a DeWalt cordless drill that I can hardly do without. Use the best tools you can, learn from them every day.
I wrestle with this question all the time as a professional educator. Is it better to do ” the thing” the way you can best do it? Or spend time (the most valuable resource) learning some other way. For example there is a task I can do in about 20 seconds about 90, maybe 95% of the time. The remainder of the time it takes like 10 minutes if I mess up the first pass. There is another way to do it that takes like 2 minutes with about a 98% success rate. Which is better? Still don’t know!
Fortunately for the students kind of the whole point of being a learner is you are given the valuable time (plus as student compared to a professional your time is worth way less) to practice and learn new stuff.
For my part when working and not teaching I value my time at $200/hr. Somewhat arbitrary but hellos the mental calculus- if it takes me 10 (or 40hrs) to become proficient or better save it make me the equivalent $. Unless it’s just fun. Then it’s a hobby and is expected to cost me money anyway!
My take is simple. If AI tools of the future are so good, they are so easy at making expertise in a task obsolete, then whats the issue? Current studies strongly indicate new AI fancy tools and applications reduce critical thinking in their users. I think critical thinking is more important than a tool. if you always design with ancient components and for some reason someone said you had to do something else, something tells me you would be able to adapt with a tiny bit of friction and depending on your personality a grumble or two.
Some of the premise of this article is flawed as well. Changing tools regularly can help deepen someones understanding of their subject matter. Seeing the same material from different angles is a brilliant way to reinforce learning and is highly recommended by educators. I haven’t been around as long as some, but most of these tools aren’t all that different. It’s usually a matter of learning which levers to pull to accomplish a task.
Hmm I’m curious about these studies you reference regarding reducing critical thinking. My personal experience has been that you can’t just turn your brain off and let the ai lead you along, otherwise you end up completely in the wrong neighborhood. It’s a willful leading the ai towards a certain conclusion while the nitty gritty steps get done for you
The Master builds his own tools.
Hardware-wise I am slowly rebuilding my house ecologicly and selfsufficiëntly. I do everything myself, even if it takes me every winter to learn new skills that I apply in the summer.
Software-wise I slowly switched to 100% open source in the last decade. Examples: from MS Office to LibreOffice, WSFTP to Filezilla, PaintShopPro to GIMP and Inkscape, SoundForge to Audacity (still cheating sometimes), Winzip to 7zip, Premiere to Shotcut, etc etc. It wasn’t always easy but worth the invested time in the long term. Now running several (identical) unix workstations and windows laptops with the same stack (using apt and choco). With gratitude to the open source community!
Check out Ardour for sound.
for technical / engineering stuff, i think the ‘solved vs unsolved problems’ sums it up. sometimes churn is just churn, and sometimes churn lets you explore a new problem space. and i’m immensely interested in those sorts of trade offs but each one i make with full particularity.
for art stuff, i think it’s a more interesting problem. i’ve applied my not-invented-here disease to musical instruments a couple times, trying to build relatively exotic instruments that have crossed my path. and when i build something new, i am always impressed by how unplayable it is. and that is obviously because i don’t know how to build an instrument. but also, i buy a real instrument that has been refined over hundreds of years, and it is nearly impossible to play too! people develop full mastery over something unplayable like a flute where a new player can barely even make a sound let alone control it!
so when we have tools that have an artistic aspect, i think it is a weird place to be where people are sometimes using the new tools before mastering the old ones. we’re facing an era where some tools are never mastered, and we see a flood of newbie art because the old tool is so perceptibly dated, even before anyone masters it. a neat problem to have, really.
I’d use SuperCalc3 under DosBox if I could find a copy. For the early 90s , butt-kicking math and graphing functions
Indeed.
https://archive.org/details/008_supercalc-3-msdos-user-copy
You failed to ask the only important question: are the new tools more useful?
Novelty doesn’t guarantee utility. There’s no guarantee the people who make a new version of a tool even know how to use it.
Ask a hundred design students to reinvent the screwdriver. How many will look like something out of a Tim Burton movie, and how many will have a hexagonal handle that fits in an end wrench like the old Craftsman classic? Which one would you rather use on a screw that’s rusted in place?
The same is a hundred times more true in software. There are graphics packages built and sold by people who have no idea how to send a job to a printing house. There are CAD packages perfectly happy to draw hollow ball bearings with precision ground internal and external surfaces, or fluid paths that are joined inside a block by an H-shaped hole.. and all of them seem to assume that flat, square, parallel and exact dimensions just happen. Meanwhile, machinists call fillets “the most expensive button in CAD”.
Tools that last do so because they’re useful. We replace them when we find something more useful. Novelty is more likely to be several steps back than a step forward.
KiCAD didn’t replace Eagle because it was new. Eagle was always a hot mess, then someone at Autocad said, “what are the most idiotic license terms we can imagine for this user base?”. As soon as KiCAD stopped being even worse, people started to jump ship.
Good points all around!
I was assuming that you knew that the new tool had a feature that you actually wanted. But your point is something deeper — that you can be fooled into something worse by its shiny-newness. (I’m so reluctant to try out new tools, that’s not my problem…)
The “most expensive button” is pretty funny.
Every version of Micro$oft Word since Word v2.
“Um, You might’ve noticed my underwear has a–has a hole in it.
it’s uh, You know, i–i…
I–I don’t see any reason to–to throw it out. It’s…
The waist is still fine.
You know, you can see–see It’s still real stretchy.”
-Peter griffin
I’m a fan of learning new tools if possible. Even if it’s just a new and better way of doing something within the current software, such as using the new features in autodesk civil3d for my day job. Because even though you may feel comfortable with the way you do things, with the pace things move today others may start to see the holes in your underwear that you think is fine but shows you are behind the curve.
It seems like there are at least a couple of important variables related to scale.
If you are just doing your own thing for your own amusement your freedom to use whatever idiosyncratic methods amuse you(whether that means something that hasn’t changed in absolute ages, or a certain amount of not-really-justifiable adoption of new tools because the project is, in part, an excuse to futz with new tools) is basically unlimited. If you are working with others, or pulling from others work, or doing things at least in part for them to be useful to others you now have to consider either finding some mutually acceptable overlap or the amount of time and effort involved in porting things between different ways of working.
Similar thing for scale: if you are doing once-offs have a lot of freedom to continue using some part for a decade or more after the vendor starts issuing dire warnings about using it for new designs; or use something that would normally be unaffordable because hey this particular model of obsolete thin client was actually implemented in a quiet beefy FPGA because a dead startup needed a minimum viable product well before they had an ASIC sorted out. If you need to build a thousand of something; though, you probably actually need to care about what’s current and what’s cost-effective now.
To put this in perspective, some artists use brush and paint to do portraits just like medieval times instead of using modern digital cameras.
The tools don’t matter if the result achieved suits its market or purpose.
Sometimes the new tool doesn’t produce the same result. Carpenters have (mostly) replaced planes, chisels and scrapers with sanders. The planes and scrapers can get a very nice shiny surface on the wood, but you need skill to get there. The sander can be handed to anyone and they can a smooth surface down to a line easily. Almost as fast as the skilled person w/ the edge tools.
Agreed! At least in the carpentry case, hand-planed is a different quality.
But if I use Platformio or makefiles, nobody else knows. :)
There seems to be a fundamental misunderstanding here.
As long as the tool you are using does the job, and does it well, you don’t change tools.
How often should a mechanic throw their wrenches out and buy new ones?
The DESIRE to have the new-shiny is part of the problem, and is an infection that is intentionally spread by those who want your money.
Why would I WANT to learn an entirely new interface, for my mostly-the-same phone every 18 months, just because some UX designer needs to justify their job to a company that needs to make their product number go up?
My phone is a tool. The more time I have to spend learning where the same setting got put THIS version, the less time I have to do other things.
Oh no. I’ll miss out on icons with rounded corners or something…
Having a clean UI that is useful is great!
But is one step (or more) less important than being able to just use the tool.
But ART?
That is about exploring.
The job of an artist is to push the boundaries and express thoughts.
The problem we have here is both kinds of “work” being treated he same way.
And the “art” way is sexy and new, so that’s what a company is going to push.
Also. Generative “AI” is theft.
If you can’t replace the term AI in your statement with “a search engine” without it feeling like you are stealing someone else’s work, it’s because you are stealing someone else’s work.