This sounds right. There was a "dark ages" in the early 2000's, where Unix/Linux browser support was terrible. You were either stuck with old Netscape 4.x, which crashed every half hour, or you had to work with half baked, almost unusable beta builds of Mozilla (pre-Firefox). Because of this, I actually ran Windows on my desktop for a while!
So you don't think the complete lack of connectivity (video, networking, and storage) is going to be a problem for a normal person? There is absolutely no comparison with a Raspberry Pi. It has all those things and great software support. That Sparc platform is for a satellite!
But yes, you are correct, it's much easier to get a SPARC license.
You kind of proved my point: that's an embedded platform. No consumer is going to be running it. The last link indicates "VxWorks, PikeOS, and RTEMS operating systems available."
I'll give you the benefit of the doubt that you can load a Linux distro on there. You have a couple other problems: there's no ethernet or wifi, and you only have 512 megs of RAM.
It wasn't just the chip, it was the whole ecosystem.
Could an every day consumer buy a Sparc motherboard, chip, and build their own system? No. There may have been one or two vendors that did that late in the 90's, but they didn't stick around.
Solaris 2.4 and earlier was a little rough. I no longer remember the exact details, but 2.5.1 was the first release I felt comfortable running instead of SunOS 4.x.
I worked at smaller companies that were more risk tolerant. At an earlier startup (late 90's), all the developers ran RedHat on Dells, but the actual production environment was Solaris. I remember going out to a customer site and deploying some massive Sun Enterprise setup. It probably cost a half million USD for the whole configuration. Multiple servers (Enterprise 450 and 250 I think, two of each), multiple storage boxes (StorEdge), Veritas filesystem software, some other Veritas HA system, several Cisco switches, some firewall brand that I forget. All this was connected to a fractional T1 (512kbit.) It never went live due to some contractual issue. All that money wasted.
They could've also just built a real package manager, left it up to the community to add GNU tools. I remember the Solaris package situation being annoying, having to download crap off FTP and install manually. Or often build from source.
Solaris was pretty bare bones, out of the box. I spent days customizing it. Even early Linux distros, like Slackware, were in better shape.
I'm not saying it would've been easy. They could've installed the GNU tools in a different path and left it up to the user to change their preference. They did something similar with putting the BSD tools in "/usr/ucb" (I may have that path wrong.)
I had a similar experience. I forget the exact timing, but it was probably around 2002 - 2003 that Linux felt "good enough." I was at a startup that finally considered putting its "enterprise" stack on Linux.
Yep. Solaris was the Unix for the 90's. I worked at a couple of early ISPs and users demanded Sun Solaris shell servers. At a couple of startups, we built everything on Solaris because Linux wasn't accepted in the enterprise yet.
By the early 2000's, 1) Linux was solid, why bother with Solaris? 2) ebay and other channels were flooded with used Sun hardware from failed dot-com companies anyway. No need to buy new.
With good input (prompts, specs...) LLMs can generate code that is often correct, faster than a human could generate equivalent code. Even when there are bugs, it is still "useful" from purely a time savings perspective. If you don't like the results, you can iterate rapidly.
Yes, you can use it to generate crap. I find Claude especially bad at writing like a normal person.
That may have been true in the past, and I usually wait for the first point release, but macOS and iOS 27 seem to be a definite improvement. My 5 year old phone is much snappier!
While that would be bad if they broke all of TLS, it would not let agents "take over the internet." What would that even mean? Pumping out even more slop?
Also, AI providers are literally getting a stream of traffic with every prompt and every response. How can they not know what's being worked on? They are more likely to use that an excuse to ban open models where they can't know what's being worked on.
"Not shutting off API access" is a science fiction scenario.
Anthropic and OpenAI are both behind Cloudflare. It's fairly easy for an upstream to shut you off. Beyond that, the government / law enforcement could seize and disable their DNS within an hour.
Is it really capex though? New models are being constantly trained, released at least quarterly, while old ones become obsolete. Training costs vary, but never disappear.
Many forget that IE was basically the only decent browser for most of the early 2000's. It was the defacto standard. Netscape 4.x was ancient and crashed every half hour. Mozilla/Firefox wasn't stable yet. Chrome wasn't out.
Prompt injection is often not the issue: it's prompt quality. If given a vague spec (prompt), the human is more likely to understand the problem domain, see what's missing, ask clarifying questions. AI is more likely to assume and take a probable path, which may not always be correct. It's (probably) not going to tell you no, what you're asking makes no sense.
2016 was an acceleration point into madness. We were going down hill anyway, and the hill suddenly steepened. Covid didn't help. Now we have half the world with AI psychosis. I can only hope things even out a bit in the next 10 years... but they won't.