I partly disagree, software still sucks and it's a great time to build.
The #1 OS is slow and crashes all the time. The #1 email client takes 10s to load on my mother's laptop. Most popular products are slow, buggy, filled with spam, & filled with dark patterns. Enshittification won. FAANGs are the new IBM. Let's build better stuff.
Unfortunately, we've reached the point where a generation of new adults take janky technology as normal, both as consumers and producers.
Their entire life was in an environment where nothing was stable or cohesive or efficient and everything was either "free" or rented. They don't recognize what they're missing or why it might matter.
So as consumers they don't know to care when you build better stuff, and as producers they don't even know what it means to build better stuff. And soon these people will graduate into leadership and management with the same understanding of the world.
Surely, there's plenty of opportunity for the rest of us to keep quietly rescuing these janky projects from disaster, shoring them up as their sloppy compromises overtake them, but unfortunately it's very possible that it'll be a long while still before a strong and viable demand for "better stuff" returns.
> Surely, there's plenty of opportunity for the rest of us to keep quietly rescuing these janky projects
Once our generation (the rest of us) is in the ground, there will be nobody alive that even remembers that software can be made with high quality. Nobody who's ever seen really fast performing software, software that didn't crash unexpectedly, software that didn't eat your battery and storage space, software that wasn't exploitable by a 14 year old in their basement, software that didn't leak personal information all over the world. No developers who have counted CPU cycles or took the time and effort to keep a for loop in a single page of memory. Neither developers nor consumers will even believe that software can be great.
>Once our generation (the rest of us) is in the ground, there will be nobody alive that even remembers that software can be made with high quality.
Please tell me, when was this mythical era? Most software has sucked and been balls forever. Windows, why is my screen blue? MacOS, why do I have a bomb icon? Linux/Unix, why don't I have desktop apps and why doesn't my sound card work?
Almost all corporate/sold software is laden with too many untested features, and always has been. Fast purpose built software is uncommon as your average user wants features.
The previous startup waves were built on top of expanded Internet access, mobile phones, and virtualization. This allowed low capitalized startups to create 10x better software. What you’re describing is marginally better software, you’re not going to beat Gmail by making it 10s faster. No one really cares about dark patterns and spam.
There will eventually be new waves that create opportunities for startups again, LLMs are like that in some cases, but I’d argue that mobile phones were by far a larger disruptive innovation than LLMs so far.
When Apple was a startup they were riding the personal computing wave, when they created the iPod and iPhone they were no longer a startup. Sure, if you have a billion dollars and thousands of engineers you can create your own waves.
I don’t think that ignoring these foundational innovations or trying to work against trends is generally good advice for startups though, it’s easy to think you’re really clever and smart and different but meanwhile the team that did “Airbnb for dogs” has sold for $1B
You'll have a much harder time breaking into the market with things like a new OS because you have so much more ground to cover today than you would had you started back in the 90s. It's not enough to have a half-decent GUI and a web browser: you'll never get any corporate customers without a significant amount of security hardening, and you'll have to have an extremely compelling reason for them to switch even their servers over (which are run by IT people, who are much easier to train on new technologies than the rest of the workforce), much less their desktop fleets.
Consider how many businesses are built around Excel and either cannot re-create their workflows in LibreOffice Calc/Google Sheets or don't see the value proposition in doing so. You can argue till you're blue in the face that business critical processes shouldn't rely on Excel, and you'd be right, but good luck convincing the people who matter most that they need to change something that still works well enough (especially if they're not privy to the behind-the-scenes work required to integrate other systems with Excel). It's not like Excel is the only thing keeping them hooked on Office.
When it comes to other, less ambitious projects, like email software (not even hosting, just an email client; hosting is its own can of worms), you're competing against companies that either give it away for free or include it in a bundle of other applications businesses really want, like O365. I pay for Shortwave because I loved Inbox by Google that much, but it's a very niche product. I doubt the average person loved Inbox enough to trust a third party with their emails so they can get that experience back, much less pay said third party when the Gmail web interface is good enough for most tasks.
There's a reason we can describe companies like MS as being entrenched: there are no real competitors left. Enshittification works because people aren't inclined to switch without significant upsides. If your selling point is not being actively hostile to your users, why should I believe you will resist doing the same when you IPO or get bought out?
Yeah, the “it was easy back then” sounds bitter and inversely correlates success with problem difficulty.
It is getting a little tiring to hear that building software is a superhuman endeavor and that almost nothing new can be built and nothing can be replaced.
If anything, this attitude plays right into what the people criticized in the article want everyone to think. “Just give up, there won’t ever be another Marc Andreesen or Mark Zuckeberg”
What usually happens next is that 99% of that "better-ness" value then gets clawed back by the creator and monetized, and then users are left with something roughly comparable to the product it was supposed to supplant. E.g. "we're gaining users left and right, how many ads do you think they will tolerate without abandoning us?"
I meant popularity. My brand new Win11 laptop freezes or crashes all the time. My relatives laptops too. It's slower every year, even on high-end hardware.
"New" Windows laptops have always crashed unless you carefully purchase your units carefully and generally remove any manufacture trash installed with them. Consumer laptops are especially bad. Meanwhile I've purchased and supported thousands of business grade laptops with a much lower rate of incidents, but they are typically more conservative with the cutting edge hardware and have better bios/uefi support.
Users have inertia, it takes years for superior products to win.
VC capital is useful for speed & innovation, but most of the time leads to bloat & rent-seeking. Does DocuSign need 7,000 employees? My journey of bootstrapped entrepreneurship has been much more sustainable and respectful to my users.
I am kind of with you, I do think the notion of a lone developer not collaborating is going to be the norm with AI. We just don't need to collaborate anymore.
The #1 OS is slow and crashes all the time. The #1 email client takes 10s to load on my mother's laptop. Most popular products are slow, buggy, filled with spam, & filled with dark patterns. Enshittification won. FAANGs are the new IBM. Let's build better stuff.