One of the issues I’ve seen myself and have downvoted in this community is people just posting dumb shit. They don’t actually have that opinion and if they do, they haven’t given it any thought. Nothing wrong with having an unpopular opinion if they can actually defend their opinion. But just straight up posting something unreasonable under the guise of it being “unpopular” is downvote worthy.

The Netherlands have this special place called Westland also known as the Glass City. It’s where they grow all of their tulips or whatever in greenhouses. The entire area is filled with back to back greenhouses that blast huge amounts of lights down to make the flowers and fruits and so forth grow all of the time. It’s actually clearly visible from space, being super bright in an already very densely populated country.

Which has lead to a whole new issue popping up where the AI chooses not to use the tool, because it doesn’t trust it. Now this is anthropomorphizing a lot, in reality it’s a reward misalignment issue, but still. It’s called tool aversion and can be an issue. The LLM would even make up a reason the tool won’t work, or tell you it did use the tool when it didn’t actually.

In order to fix this a technique called cryptographic receipts is used. This adds an expected hash output based on the tool and the input. The tool outputs the correct hash, but if the LLM didn’t call the tool and made up some BS, the hash isn’t there. On the router level the missing or incorrect hash is easily detected and an exception raised in order to not present the result to the user (and hopefully correct the issue).

Another whole issue with using tools for certain stuff is it requires the situation to be properly evaluated for the tool to be called in the first place. So a user might hear the marketing say: Our old AI used to mess up maths, but our latest and greatest model is super duper smart and can do maths. The user then goes to verify this is the case with some simple tests, and sees it’s correct. Then follows up with some harder math problems and the output is still correct. However the user doesn’t know and isn’t informed the AI used the tool, the user is lead to believe the AI is smart and can do maths now. So in situations where the detection doesn’t work, or the LLM doesn’t use the tool for whatever reason, the user is fed a nonsense response. Which the user will fully trust, given what he knows and tested himself.

It’s so annoying these AI companies continually do shit like this. Lie to users, keep them in the dark and overpromise. Users relate the AI to human intelligence. And for humans, if you can do hard math problems, the circumstances don’t really matter. We grasp the basic concepts and can execute to get a result and know how to double check. For these AI tools, it might get a PhD level math question totally right, faster and better than a human. And then get a simple calculus problem, someone in primary school could solve, wrong. For users this makes no sense.

As someone who has seen printers do these kinds of numbers for a while, it’s going to be rough. There’s three basic options, first of all you can go with the hyper expensive enterprise level printers. They are space age futuristic, you can’t touch them, there is a phone number that warps an engineer to your location to fix it if it breaks. But it doesn’t break, it just works and keeps on working. Engineers still show up every two weeks and do magic shit to it. There’s always an even number of units installed so there will always be a backup machine right around the corner. Big downside, it costs mucho monies, like shockingly much.

Then there’s the professional line of printers. Pretty decent, you can do a lot of fixes yourself. It will break down every now and again and need to be replaced once a year. But it’s much better in terms of costs. Still hurts to put a machine down in the back of the shops with a “parts” label on it, but it’s a solid option.

My favorite option by far is the jank option. Buying the absolute cheapest printer you can get. We are talking about laser printers (don’t even think of doing volume on anything other than laser), so cheap means bad. Brother is by far the best choice, they cut cost where possible, but don’t go to far. These things are then rode hard and put away wet. They start making weird noises, they start emitting dust, their print quality suffers, but they keep on trucking. You can tell the people using it some neat tricks, like hit it pretty hard here when it makes this kind of noise. If it gives out this error, reset the thing using this secret special code so it ignores all the errors and keeps going. The little rubber paper pickup ring goes first, so order a box of thousands of those from China for a few bucks and you get pretty fast at replacing those. And the best part, just buy like a shit ton of the printers. You can get like 10 for the price of one decent printer. Just have cold stand-by units at every station and switching over is 10 secs of work if the infra is configured correctly.

Pretty wasteful, so maybe not something one would do these days. But back in the day, I’ve seen entire logistics hubs running on those cheap little lasers. Churning out hundreds of labels and packing slips per hour each, without hardly any costs or maintenance hours. And basically zero downtime.

Fake news, did Riker put you up to this slander? If Picard hates children so much, why does the whole ship celebrate Captain Picard day? A celebration specifically aimed at children? Curious…

I’d like and subscribe

There is this super cool video series debunking some of the horse shit Destin has been saying about creationism in favor of evolution. Would recommend: https://youtube.com/playlist?list=PLInNVsmlBUlSjLSj9yGEKphF0RYRYBlXg

Anyone working for the US government might get in trouble if they see their party members being wounded like that.

I got me a hundred gigabytes of RAM I never feed trolls and I don’t read spam

They’ve been an imminent threat for decades now!

Yeah these newer systems are crazy. The agent spawns a dozen subagents that all do some figuring out on the code base and the user request. Then those results get collated, then passed along to a new set of subagents that make the actual changes. Then there are agents that check stuff and tell the subagents to redo stuff or make changes. And then it gets a final check like unit tests, compilation etc. And then it’s marked as done for the user. The amount of tokens this burns is crazy, but it gets them better results in the benchmarks, so it gets marketed as an improvement. In reality it’s still fucking up all the damned time.

Coding with AI is like coding with a junior dev, who didn’t pay attention in school, is high right now, doesn’t learn and only listens half of the time. It fools people into thinking it’s better, because it shits out code super fast. But the cognitive load is actually higher, because checking the code is much harder than coming up with it yourself. It’s slower by far. If you are actually going faster, the quality is lacking.

First time I went down into Siofra I was very low level and just started the game really. Those sniper bastards could one-shot me from across the map. At least at Anor Londo you’re pretty high level usually, but they fixed that by making you walk the plank while shooting at you.

The book was one of the first pieces of media I actually found to be scary in something like two decades. The idea of someone or something being out there or even in the same room. You look at it, see the threat and then disregard it completely. That’s some straight up nightmare fuel.

The premise falls apart if you think about it for even a second, but still an excellent base for the story.

Goddamnit Donut

Written by AI

Indeed

Ah change.org, the website most famously known for not changing anything ever.

This isn’t installed by a professional, they obviously did this themselves. Which is fine really. The internet can get so uptight about electrical safety due to the amount of Americans online. In Europe something like this is absolutely fine and normal

In Europe ground fault interruption is mandatory, as well as breakers that trip at a relatively low current at a fast speed. All of the wires are also at least double insulated with heat resistant plastic. The appliance is more likely to get replaced due to old age rather than the wire wearing through. And even if it does, the ground fault interruptor kicks in right away and prevents anything bad from happening. The owner investigates, buys a new cord for a few euro and easily replaces it.