However I also think that Agentic AI is very much not an out-of-the-box solution, local or otherwise, and it takes a high level of technical knowledge to create an effective AI agent. And there's a problem now where most orchestration is fixed on what models are used for what tasks with no ability to weight constraints like cost, speed, and security.

The official URL is "hugging-face" and the technical page lists the Unicode Name as "Hugging Face" while calling it "Smiling Face with Open Hands Emoji". The original proposal was called "Gmail HUG FACE".

https://www.unicode.org/L2/L2014/14174r-emoji-additions.pdf

It seems like they're removing POP/IMAP support from the web client?

That's not the case in the US.

The big box hardware stores have a large track saw vertically mounted that can cut full sheets of wood, but there usually isn't someone working who is trained to operate it. They also have a rickety cart with a hand saw and tape measure attached with too short cables that you can try to use to make your cuts of dimensional lumber.

I want to say that Apple invented that particular design language, and prior to the original iPod nothing resembled that UI but I'm not sure.

It seems to me that the OP was either too young or not alive when the OG iPod came out, and so they're looking at it like a modern man trying to figure out caveman tools.

There were a couple things Apple had in their favor but it certainly didn't take 10 years for folks to catch up. They mostly one because they captured people's hearts and minds.

The key things Apple did right was a high speed connection and owning their software. The Creative Nomad Jukebox used the same HDD as the iPod and beat Apple to market by a year, but Creative didn't own the software. The iRiver H series of players showed up around 2004 and were comparable to iPods in terms of quality and capability, though their aesthetic wasn't polished minimalism. Their UI wasn't as good but it was still very good. Their crux was Windows and Mac support wasn't great. I had an iRiver in 2007 and was running an opensource (rockbox?) firmware to get around some of those issues.

Apple has always been good at innovating around the problem which is usually the UI or software at large. This comes in part because of their vertical integration and being able to make minor tweaks to other parts of the ecosystem (e.g. desktop OS) to dramatically improve the experience of their product. The iPods are another great example where they made minor tweaks to iOS Bluetooth that magically made the experience of using wireless headphones dramatically better.

It isn't about improving things or fixing the problem, it is about mitigating cost.

Public sex, shooting up drugs, and vandalism still occur; but the owner/operator of the bathroom no longer has to maintain the bathroom or deal with individuals complaining about the poor state of the bathroom as a result of the behavior.

When the iPod came out you largely had two options for carrying your music collection on the go. You either carried a binder of CDs, or you had some niche player like Mini-Disc or an MP3 player. Both alternatives were expensive and had limitations similar to a CD in terms of number of tracks you could carry.

I had an MP3 player on either side of 2000 that was slightly smaller than a deck of playing cards that could use Smart Media flash memory cards. The largest card at the time was either 16 or 32mb and was enough to hold 1 album at near CD quality.

Creative's Muvo was a weird form factor that was larger than an iPod. It had a horrid interface both on device and for loading music. It's only grace was that it was slightly cheaper than an iPod and didn't need a Mac with FireWire. Although iircc this was pre USB 2.0 so not having FireWire would mean loading music took forever and a day.

The iPod allowed you to carry most, if not all, of your music collection in a package slightly larger than a deck of playing cards. And it had a fantastic interface for navigating music on the device.

This was at a time before most people had laptops and if you had a PC it was at home and used sparingly. The iPod was such a compelling mobile computing device that it drove adoption of the iMac. Apple would eventually release iTunes for Windows and USB support but that was many years later.

Very rarely I'll come across a page where valuable information is hidden in those areas and it takes someone else showing me to actually see it.

Personally, I don't see the value in idle timeouts or time based eviction. It doesn't improve model load performance but does mean you're more likely to incur a load penalty for any request.

I generally use the same two models for my agentic workloads and they sit comfortably next to one another in the 32GB of VRAM on my GPU. If I need to load a larger model, llama-swap can easily eject both models if necessary.

The gulf is quickly shrinking though and it seems like the need for llama-swap will disappear soon.

https://github.com/mostlygeek/llama-swap

https://github.com/cptskippy/battlemage-llm-gateway

It's designed so that you can re-run the scripts to pull the latest updates. When Muse Glimmer was released the other day I just ran the 02 script to build the latest version of llama.cpp with support for it.

What I meant about configuration is that the onus is on you to manage all of that. I know some people love that level of control but most people it's a burden. Containers are part of your profile and move with you if you're using Firefox Sync. You don't have to copy or maintain multiple profiles or Sync accounts.

The A3B models are super fast but I found the A3B Q4 model ran in circles a lot and ended up taking longer to complete tasks that 27B Q6 because it kept having to redo/rethink/fix something.

I was writing extensive prompts to rein it in and it would still ignore basic directives like "never force push on the repo, ask me instead". I ended up switching back to 27B after about a week of frustration and lost productivity.