Every time I see a headline like this I’m reminded of the time I heard someone describe the modern state of AI research as equivalent to the practice of alchemy.

Long before anyone knew about atoms, molecules, atomic weights, or electron bonds, there were dudes who would just mix random chemicals together in an attempt to turn lead to gold, or create the elixir of life or whatever. Their methods were haphazard, their objectives impossible, and most probably poisoned themselves in the process, but those early stumbling steps eventually gave rise to the modern science of chemistry and all that came with it.

AI researchers are modern alchemists. They have no idea how anything really works and their experiments result in disaster as often as not. There’s great potential but no clear path to it. We can only hope that we’ll make it out of the alchemy phase before society succumbs to the digital equivalent of mercury poisoning because it’s just so fun to play with.

Copy pasting my reply from the last time this came up:

Check how nearby colleges and universities dispose of used assets. The state school near me maintains a very nice website where they auction off everything from lab equipment to office furniture. It’s also where all their PCs go when they hit ~5 years old and come up in the IT department’s refresh cycle. The only problem in my case is that they tend to auction stuff in bulk. You can get a solid machine for $50 to $100, but only if you’re willing to pay $500 to $1000 for a pallet of 10.

TBH I just use the Feeder app on my phone. Fully self-contained. No account, no server, no middleman of any kind. Just the app.

I’ve been meaning to set up something more elaborate, but this really does work fine, and I like to mention it in these threads for anyone who’s interested in RSS but thinks it’s a big lift to set up. It can be complicated, but it doesn’t have to be. Download an app and start adding publications that interest you. That’s all it takes to get started.

Time to quote Dan Olson again. This was originally written about NFTs, but just replace “crypto” with “AI” and it’s still 100% relevant:

When you drill down into it, you realize that the core of the crypto ecosystem … is a turf war between the wealthy and the ultra-wealthy. Techno fetishists who look at people like Bill Gates and Jeff Bezos, billionaires that have been minted via tech industry doors that have now been shut by market calcification, and are looking for a do-over, looking to synthesize a new market where they can be the one to ascend from a merely wealthy programmer to a hyper wealthy industrialist.

From the incomparable line goes up

I’m reminded of this video about how changes to the construction industry starting in the '50s resulted in the loss of ornamentation in architecture

https://www.youtube.com/watch?v=DBOXF-FION4

I got 8 in the span of about five hours today. Absolutely insane.

Yep, I got a series of those recently claiming that I had unpaid paid tolls. Each messagr came through as a group text with two or three random numbers, which were immediately removed from the group after the text arrived. I’ve been wondering why they started doing this. I assume they’re trying to exploit some kind of loophole in the carriers spam filtering.

I’ve wondered before how large an order would be required to entice a white label manufacturer of robot vacuums into doing a production run of units with Valetudo preinstalled.

I would absolutely buy one if someone could work out a fair business arrangement with the developer and throw the project up on kickstarter.

All the Bardcore covers by Hildegard Von Blingin are amazing.

The sample data shared in the article includes

"c": "ES", // Country code,

ES is usually used for Spain, so it looks like these tests were run from within the EU.

Man, I feel you on the affiliate link fluff. I actually ended up unsubscribing from the Popular Mechanics and Popular Science feeds because the signal to noise ratio was so bad.

The creator of Nunti provided a very good primer on the algorithm design here. Basically, you indicate to the app whether you like or dislike an article and then it does some keyword extraction in the background and tries to show you similar articles in the future. I suppose you might be able to dislike a bunch of the fluff and hope the filter picks up on it, but it isn’t really designed to support the kind of rules that would completely purge a certain type of content from your feed.

Most of the feeds I subscribe to came to me in one of two ways:

1. I enjoyed reading an article posted somewhere else (Lemmy, etc.) so I sought out the feed of that publisher.
2. Sometimes news outlets enter into agreements to republish each others articles. When they do this, the re-publisher will usually include a little blurb at the end giving credit to the original publisher. If a feed I’m already subscribed to has an article re-published from elsewhere then I click through and check out the original source to see if I want to follow them as well.

It can be as simple as just putting an app on your phone. I use feeder which is fine. Pretty bare bones, but in that way it’s easy to learn and use.

I’ve also been meaning to try out an app called Nunti, which I heard about a while ago from this Lemmy post. It claims to be an RSS reader with the added benefit of an (open source and fully local) algorithm to provide some light curation of your feed. It looks interesting, but I haven’t actually tried it out yet because I’m still deciding whether I want any algorithm curating my feed, even one as transparent as Nunti’s. It’s also only available through F-Droid right now, which is a bit of a barrier to entry.

I feel the same way about AI as I felt about the older generation of smartphone voice assistants. The error rate remains high enough that i would never trust it to do anything important without double checking its work. For most tasks, the effort that goes into checking and correcting the output is comparable to the effort I would have spent to just do it myself, so I just do it myself.

If you were actually hoping to buy one but the rounded corners are a dealbreaker, then you may be interested to know that the DIY edition lets you mix and match the older display with the newer motherboards. Looks like opting for the older display even saves you $130 on the purchase price.

Out of curiosity, what software is normally being run on your clusters? Based on my reading, it seems like some companies run clusters for business purposes. E.g. an engineering company might use it for structural analysis of their designs, or a pharmaceutical company might simulate the interactions of new drugs. I assume in those cases they’ve bought a license for some kind of high-end software that’s been specifically written to run in a distributed environment. I also found references to some software libraries that are meant to support writing programs in this environment. I assume those are used more by academics who have a very specific question they want to answer (and may not have funding for commercial software) so they write their own code that’s hyper focused on their area of study.

Is that basically how it works, or have I misunderstood?

This actually came up in my research. Folding@Home is considered a “grid computer” According to Wikipedia:

Grid computing is distinguished from … cluster computing in that grid computers have each node set to perform a different task/application. Grid computers also tend to be more heterogeneous and geographically dispersed (thus not physically coupled) than cluster computers.

The primary performance disadvantage is that the various processors and local storage areas do not have high-speed connections. This arrangement is thus well-suited to applications in which multiple parallel computations can take place independently, without the need to communicate intermediate results between processors.

I’ll have to look a little more into the AI stuff. It was actually my first thought, but I wasn’t sure how far I’d get without GPUs. I think they’re pretty much required for Stablediffusion. I’m pretty sure even LLMs are trained on GPUs, but maybe response generation can be done without one.

I’m not sure what you’d want to run in a homelab that would use even 10 machines, but it could be fun to find out.

Oh yeah, this is absolutely a solution in search of a problem. It all started with the discovery that these old (but not ancient, most of them are intel 7th gen) computers were being auctioned off for like $20 a piece. From there I started trying to work backwards towards something I could do with them.

I was looking at HP mini PCs. The ones that were for sale used 7th gen i5s with a 35W TDP. They’re sold with a 65W power brick so presumably the whole system would never draw more than that. I could run a 16 node cluster flat out on a little over a kW, which is within the rating of a single residential circuit breaker. I certainly wouldn’t want to keep it running all the time, but it’s not like I’d have to get my electric system upgraded if I wanted to set one up and run it for a couple of hours as an experiment.