monkmartinez

Hear! Hear!!!!

Practical, reliable, affordable and just works technology; The amazing Brother HL-L2320D printer. I can't remember the last time I bought toner. The thing just keeps on printing. I think they make a wireless version, but I can't be bothered to change something that just works the way this printer has. We have fancy printers at work that are also Laser based and they are not as reliable the printer sitting next to me.

Not much? I mean, I set it up initially to use as an LLM search tool for agents. That required me to use a specific version of SearXNG due to a JSON bug. My exact docker command is here:

    docker run -d -p 9017:8080 --name searxng_data -v "${PWD}/searxng_data:/etc/searxng" -e "BASE_URL=http://localhost:9017/" searxng/searxng:2024.8.31-769b2ed03

I run this from the directory where I want to store the data on windows (hence the weird PWD thing). I also use cloudflare tunnels and have the internet facing website, https://sear.mydomain.com, pointed at http//localhost:9017. I set up all my cloudflare tunnels with email authentication. My token lasts a month (but is configurable to whatever time frame you like). My traffic is strictly mine, so it seems to be working ok.

Its basically no maintenance after this setup. I can mess with all the settings (except results format) from the main interface.

Because I can configure exactly which search engines to index, no ads, use it for my local LLM's for search, better image search (maybe not better, but totally different results), a bazzillion tweaks and mods. I don't have to pay Kagi for the same results or keep track of how many searches I have left.

I don't have ads on my SearXNG instance and I find that its comparable to Kagi. Have you tried searching with SearXNG?

I am 100% in agreement, between SearXNG and Google I can find darn near anything. Kagi was ok, it seems to really prefer reddit for the types of things I was searching for. I was also super conscious about the number of searches I had left in my plan. F-that. I type random stuff into the omnibar all day. Leaving Kagi as just enough friction to go to kagi, type a search, etc. that I didn't want to deal with it.

Memory. I don't know the equation, but its very easy to see when you load a 128k context model at 8K vs 80K. The quant I am running would double VRAM requirements when loading 80K.

You could buy access. Buy yourself into a chamber of commerce where you want to live starting with a small business in the area. Keep it small and look for businesses where the subject of the business itself, is interesting to you. Expand from there... angel investing/owning.

Buy a mini-excavator and a truck. Learn how to dig holes. <- I really want to do this when I retire from my main gig. Graduate to bigger holes and bigger equipment.

Develop a product for a hobby you are into. For example, if you like woodworking, design a better ruler. You might just become the next Woodpecker tool company!

Start a vinyl wrap business.

Buy a care home for elderly.

Just a few thoughts that came to mind...

That is a fascinating take and approach to contemplating a language.

Can you think of any other languages that share a duality like Swift? I mainly play in the python ecosystem, but I am looking to branch out and really learn a compiled language. What you wrote about Swift makes sense and would be concerning if I had recently picked it up.

"Yadda, yadda..." regarding picking the right tool for the job aside, I don't want to waste time on a language that can be usurped by a giant corporate borg anytime they see fit.

Correct!

Arizona is dropping the hammer on a bunch of alfalfa farms here. Interesting side note is that most of them were run and owned by Saudi companies or company. There are markedly less cotton farms than when I was a kid. The farms that I have direct knowledge of are very progressive and not afraid of innovation/change which is very encouraging.

Also of note is what Arizona does with it's share of the Colorado river. My city chose to recharge the ground water and I believe the results have been acceptable.

How hard would it be to move to 3nm or 2nm in Arizona if something happened in Taiwan? Genuine question because I don't know much about these plants.

I have heard the water needs are massive, but the reclamation tech is mind blowing. I haven't dug into it.

It is hot here in Arizona, but its generally a dry heat. We have humidity in 15% range for most of the year. Our winters are very, very awesome and generally dry as well.

Exciting times! As a resident of Arizona, it is great to see these plants being built here. The intel plant is absolutely massive and the plant I see most often. The TSMC plant is very north of my location and I've only seen it once.

I can only hope this brings more high tech jobs to the state and associated wages. A rising tide lifts all boats, que no?

I hope you are not a sole, and have the complete human embodiment of a soul. Otherwise, yes, I agree. My one off projects use more and more cdn loaded information. I love pico, bulma and htmx.

Openrouter... you get all the models and its not as expensive as you would think. I spent $3 with aider the other day in like the blink of an eye with Anthropic. I am working on a FASTHTML thingy and loaded all the docs, plus a few huge replicate api files into the vector database. Most of my back and forth usage averaged about $0.02 for each turn with Claude 3.5 Sonnet. To give you an idea: My context + prompt were around 18000 tokens with completions around 1500 tokens.

Which would make the cost of each mirror higher, right? Additional processing will naturally increase the cost per unit.

By the way, super cool project and thank you for sharing. My experiments with concentrating sun power when I was a child were directly related to the spontaneous combustion of insects. Still making amends for the number of ant hills my brother and I cooked with the sun.

TFA didn't address the one thing that, in my experience, has been the raison d'être (at least in my city) for large employers to generate lots of emergency medical calls.

SICK LEAVE POLICY

We get a ton of calls to certain call centers in our area while other large centers generate very few. When management isn't hawking over the patient, I ask them what the sick leave policy looks like. Generally, they can only take a limited number of sick days without triggering some kind of adverse metric that will impact their employment. These metrics can be measured in points, stars or some other arbitrary method to keep track excused vs. unexcused absences.

If you call out sick, that would score the highest in terms of adverse action record keeping. If you "go down" at work, you have more options and the metrics do not generally apply. You can also file a workers compensation claim which you can not generally do from home with a cold. Whether or not the claim is accepted by the insurance agency is another matter entirely. However, it is not a stretch to see why a bad sick leave policy would lead to employees figuring out ways to circumvent the policy without damaging their employment status.

These types of abuses, again in the states, are so prevalent that we joke about it at the station. The EMS system is abused from top-to-bottom and there is nothing to hold anyone accountable. It is simply going to break someday and that will be that. I can say that it seems the day when there are no ambulances available is drawing closer and closer... It already happens a few times a day in my city.

Thank you for using discourse! It is sooooo much better than discord. I really, really can't stand discord. Searching for an issue or topic is like wading into the ocean looking for a bluegill.

You can't. Sorry.

Unless...

You have a couple hundred $k sitting around collecting dust... then all you need is a DGX or HGX level of vRAM, the power to run it, the power to keep it cool, and place for it to sit.

Why are (some) Europeans surprised when they are not included in tech product débuts? My lay understanding could best be described as; EU law is incredibly business unfriendly and takes a heroic effort in time and money to implement the myriad of requirements therein. Am I wrong?

I chart codes on my EPCR, in the PT's house, almost everyday with one hand. Not joking about the one hand either.

Its easier, faster, and more accurate than writing in my experience. We have a page solely dedicated to codes and the most common interventions. Got IO? I press a button and its documented with timestamp. Pushing EPI, button press with timestamp. Dropping an I-Gel or Intubating, button press... you get the idea.

The details of the interventions can be documented later along with the narrative, but the bulk of the work was captured real-time. We can also sync with our monitors and show depth of compressions, rate of compressions and rhythms associated with the continuous chest compression style CPR we do for my agency.

Going back to paper for codes would be ludicrous for my department. The data would be shit for a start. Hand writing is often shit and made worse under the stress of screaming bystanders. Depending on whether we achieved ROSC or not would increase the likelihood of losing paper in the shuffle

This borked our dispatch/911 call center then as well. However, it wasn't as bad as this one. This outage put our entire public safety system into the stone age and with that we were at stone age efficiency.

Just completed a standing 24 due to this outage. My B-Shift brothers and sisters had to monitor the radios all night for their units to be called for emergencies. I heard every dispatch that went out.

We were back in the 1960's with paper and pen for everything, no updates on nature of call, no address information, nothing... find out when you show up and hope the scene is secure. It was wild as it was coupled to a relatively intense Monsoon storm.

You need to document everything during a code arrest. All interventions, vitals and other pertinent information must be logged for various reasons. Paper and pen work but they are very difficult to audit and/or keep track of. Electronic reporting is the standard and deviating from the standard is generally a recipe for a myriad of problems.

We took 911 calls all night, I was up listening to the radio all night for my unit to be called. The problem was the dispatching software didn't work so we used paper and pen. Glory Days!!!!

You are right of course.... IF your metric for manageable/useable is measured only tokens per second (tok/s).

If your metric is quality of output, time, money and tok/s, there is no comparison; Local models just aren't there yet.

Yes, I am with you 100% and keep several LLaMA's on my workstation for that reason. I use Openrouter for everything else. Everything that isn't sensitive goes to one of the big kid models because they are just sooooo much better. LLaMA 400b might be the start of running with the big kids, but I know we are not close with the current available models.

llama3 at 8billion params is weak sauce for anything serious, it just isn't in the same galaxy as Sonnet 3.5 or GPT-4o. The smaller and faster models like Phi are even worse. Once you progress past asking trivial questions to a point where you need to trust the output a bit more, its not worth effort in time, money and/or sweat effort to run a local model to do it.

A novice isn't going to know what they need because they don't know what they don't know. Try asking a question to LLaMA 3 at 8 billion and the same question to LLaMA 3 at 70 billion. There is a night and day difference. Sonnet, Opus and GPT-4o run circles around LLaMA 3 70b. To run LLaMA at 70 billion you need serious horse power as well, likely thousands of dollars in hardware investment. I say it again... the calculus in time, money, and effort isn't favorable to running open models on your own hardware once you pass the novice stage.

I am not ungrateful that the LLaMA's are available for many different reasons, but there is no comparison between quality of output, time, money and effort. The API's are a bargain when you really break down what it takes to run a serious model.

We don't have any more ability to test a CO detector than you! Replace the batteries, press the test button and/or buy a new unit if the old one fails. Please don't call 911 unless you have an actual emergency.

Nothing is fair. I try to impart this lesson with my children everyday. They see it, viscerally, when they go visit their grandparents in Colombia. There are really poor people in this world who never had a chance. You have a chance and have learned an important lesson.

A new member of my crew immigrated here from Guatemala 6 years ago. He is now 24 and has learned English to the point where there is no accent. My wife has been a citizen for 18 years and still has a very strong accent. My Guate buddy took his opportunity by the horns and hasn't let go. He is amazing and has used the system(s) to the best of his ability.

He joined the USAF to get ahead and it has proved to be one of the best moves he made. He recently finished paramedic school paid for by the GI Bill. This is a story as old as time, and my grandfather did the same thing after swimming across the rio grande in 1934, except is was called the US Army Air Corps.

I say all that to say, you can lament your current station in life all you want. Just know that nothing will change if you sit in the corner all pissed off. The other option is to recognize that you are in control of how you respond to adversity. There are two options when one feels they have been slighted; 1. Get angry, stay angry, and pout. 2. Brush it off and have another go in a different way.

Some ideas that you may or may not like:

You can still join a military guard or reserve unit. That will unlock a LOT of benefits and you might actually like it. There are always government jobs that pay ok, but offer tuition reimbursement, pensions and other benefits in lieu of high pay. Apprenticeships with the trades (electrician, plumber, HVAC) can lead to making a LOT of money and owning your own business.

After all is said and done, there are more than one way to skin a cat. (I don't condone cat violence)

Let me ask you a question; Why doesn't Singapore deal with this type of crime?

We, as a society, have stopped demanding people to take responsibility for themselves. We don't hold anyone accountable for their actions. We ostracize the basic enforcement of accountability and those that are charged with upholding law. We are soft AF as a society.