These "AI" tools are simply the next level of "autocomplete"

We started with basics like tab completion of symbols
Eventually moving to full code analysis to figure out what symbols could be tab completed
Then we moved on to function/method parameter completion
From there we got code sniplets that an IDE could inject into our code
And now we have those sniplets being dynamically modified based on parameters that we're giving it

Its never been complete, its never been perfect, but if we look at "AI" in the way I just described and use it accordingly, it most certainly does speed up development.

I recently used ChatGPT to write some code for an RP2040 to create an emulated network card for the Apple IIe. It knew nothing of the actual protocol or bus, but I was able to feed it enough information on the spec to get about 80-90% there. From that, I just filled in the rest of the code to ensure it worked correctly.

For a notice, AI most certainly can hamper the troubleshooting and learning aspects of programming. For a seasoned professional, it definitely can speed up the process of producing specific types of library code.

It really isn't at the point where it is making full applications, not even close. But for algos/libraries it is doing a pretty damn good job right now.

the "P" in HIPAA is "portability" - the law actually mandates this level of portability and sharing among medical organizations, not prevents it.

I do photography as a side thing. I've got about 10TB worth of content between RAW and PSD/PSB files. That's just one part of things. Other family members do the same, or record videos. We also sync our phones automatically to our self-hosted solutions through apps like Nextcloud. We're also going back historically and scanning in film/slides dating as far back as a century ago. Its a family archive, something better than a traditional print photo album in that we have searchability, portability, and sharability. The archive isn't just for one person, its for a whole group of people to enjoy whenever they want.

I'm running a VaultWarden cluster replicated across multiple sites, with BitWarden front-end. BitWarden themselves offers a paid hosting solution, or you can do it yourself for "free" (cost of time/hardware/resources). It integrates perfectly to every major browser, including Firefox (typing on that right now). It works beautifully on Android too to provide in-app passwords. No copy/pasting, its full on integration on all platforms. I have the browser client running on Windows, MacOS, and FreeBSD. For the rare time I want the web interface, its available, otherwise everything works great from just the browser plugins.

So shit like this pisses me off for a different reason.

Why?

Because Microsoft ALREADY did this in 2007, nearly TWO DECADES AGO, and then canned the tech and threw it away. We ALREADY had a full 3D model of Notre-Dame. It was cool. It was awesome. It was interactive. And it was entirely dynamically generated just from photographs on Flickr.

The technology was called PhotoSynth. But you already KNOW they're gonna shove some fucked up bullshit "AI" marketing behind this, its going to take longer to produce results, and the results are going to be shittier than we already had.

https://www.ted.com/talks/blai...

Articles like this always limit it to "for Linux" - but its just a general F/OSS piece of software which works on a number of platforms, including Windows itself, FreeBSD, and more.

I've been experimenting more and more lately w/ FreeBSD as a desktop OS on a number of machines.

You know what FreeBSD has natively and used by default? OpenZFS.

Do you know how many years worth of ZFS snapshots I have on some of my machines? Many MANY years worth. They're created when -I- want them. They're destroyed when -I- want that to happen. I can inspect their state and contents at will. I'm in total and complete control over MY storage.

The SOC is quite a bit more powerful than anything on any phones. Its roughly on par with the Nvidia RTX 3050.

Totally! I'm not worried about the opinions of a few people that only read a release note and think they know the entire ecosystem. I fully understand that wifi has lagged behind in FreeBSD, but other areas such as storage management and security are a decade ahead of Linux (inb4 haters try to argue with me because they don't even know cool shit going on w/ FreeBSD like Capsicum as one of many examples)

Small BSD based routers already exist: look into devices that ship w/ pfSense or OPNsense. Devices that do routing and devices that do wifi access points don't have to be the same device, which is how my networks are setup (quite nice if you want a solid mesh wifi setup by the way!)

The note there about Wi-Fi doesn't convey just how large of an improvement that actually is.

From testing tonight, I went from ~10-15mbps on Wi-Fi on 14.2 to about 250mbps on 14.3.

This is a massive overhaul of the entire Wi-Fi subsystem within FreeBSD, and its continuing to be improved upon even further on FreeBSD 15 in development right now.

Yeah, this isn't breaking any benchmarks with its current performance, but it certainly went from "absolutely sluggish and rubbish" to "perfectly usable for daily tasks" in a single revision.

Sure, this sounds like a great idea up front, but the requirements hurt the ecosystem for bespoke designs.

The FrameWork 16 for instance has 6x USB-C ports on it for its modular connectivity. By Microsoft's new requirement, ALL 6 ports must have charging capabilities, video capabilities, and because they're above the 40gbps threshold, support PCIe connectivity too.

This sounds -great- on the surface, but what is the level of complexity and monetary cost to have SIX charging ports on a single laptop? Where is the practicality in that anyways? Additionally, the cost of the muxes to push video to all of those ports, and PCIe to all of those ports.

Does your desktop have 6 display output ports? Probably not.

Thanks to these requirements, the FrameWork 16 is essentially dead in the water for Microsoft Windows certification, along with any other vendor that wants more than 4x ports on their devices.

We're either going to have a limited number of ports, or regress back to the slop of USB-A, USB-C, (mini)DP, HDMI, and whatever slop a company wants to accommodate more connectivity, instead of a single port that has a few optional specs turned on/off depending how its wired internally.

Yeah, I totally get it and would love to have a world where there is truly a single port. But even with this new spec requirement, it still isn't standardized entirely. There is the above 40gbps spec and below 40gbps spec with different requirements for each. There is a difference between "has video" and "has GOOD video", as well as "has PCIe/Thunderbolt" and "doesn't have PCIe/Thunderbolt"

Vendors having the choice between driving up per-port costs for all this optional stuff that almost nobody will ever use vs not having the ports at all? Welcome to the Apple world of only having 2 USB-C ports.

Don't worry, AI workers will replace the normal workers, and these AI workers will be the ones to create the AI morality filters. And since these are all cloud computer based infrastructure, the AI will give it a cunning pun of a name, something to do with networks in the sky.

Instagram has allowed non-square photos for years. And their UI browser even allowed a non-square layout of photos on a user's profile for at least a year as well. What's changed? They formally announced a feature that's been there forever now?

Proof that nobody even bothered to read the fucking summary.

"it makes much more sense for the state to invest in, or collaborate with, decentralized social media networks based on free and interoperable software"

Yeah, the government investing in this would so totally be "big brother watching" heh....