Netcob

Same - replaced my smaller Synology with a UGREEN, put TrueNAS on it first thing, runs great. The HDD thing was only the final nail in the coffin, but before that, there were plenty of ridiculous "upgrades" that made products worse than in the previous generation. Literally removing features, or continuing to use the same outdated hardware. That's what companies do that don't think they have competition.

For some more recent crimes against society and humanity, I'd also compare it to Stasi. Plenty of people alive today who lived with that.

Around 1 in 30 people was secretly telling on their neighbors. After unification, it was presented as a dark chapter in German history that had finally come to an end. People would get to look into their own "file" to see what and how much had been written about their daily activities. I was a bit young at the time, but I do remember frequent discussions on TV about how to move on from this, and how to make sure it doesn't happen again.

And now we're talking about reading everyone's private messages on a scale that would be the Stasi's wet dream.

I wonder - if the Stasi had been presented as a legitimate way to fight CSAM - would that have been okay?

Once in a while I save ~10 minutes by using AI. About as often as embarrassing myself by having to admit that my primary source was an AI while researching some topic.

The main thing that changed is that the CTO is in more of a "move fast, break things"-mood now (minus the insane silicon valley funding) because he can quickly vibe-code a proof-of-concept, so development gets derailed more often.

Interesting that the bus bandwidth was such a bottleneck, when you could also argue it was VRAM. I never looked into why "Hardware transform&lighting" was introduced so late (I'm guessing it didn't fit on the die next to the rasterization stuff?), but I think hardware t&l combined with storing geometry in VRAM would have reduced bus usage to a minimum.

I struggle with habits a lot, but this is one of the few that work for me. I think the kids call it "habit stacking"?

Basically, I have to brush my teeth and shower before going to bed. It's all one package, and I'm very efficient at it.

So it's at least once a day, more if I was sweating a lot.

Another one: brush teeth before leaving the house for the first time that day. ...that used to work better before working from home.

I've added a lot of other habits since then, using an overcomplicated system of deadlines and rewards.

True, local LLMs might be a better idea for this particular purpose.

Most self-hosted services are generally a better idea, if you know what you're doing.

Imagine typing 80085 into your calculator as a kid, but the number disappears and a finger-wagging animation plays instead.

I don't think LLMs are mature enough as a technology to be blindly used as a dependency, and they might never be.

The big question is, how do you train LLMs that are useful to both humans and services while not embarrassing the company that trained them?

LLMs are pretty good at translating - but if they don't like what they're reading, they simply won't tell you what it says. Which is pretty crazy.

LLMs are pretty good at extracting data and formatting the results as JSON - unless they find the data objectionable, then they'll basically complain to the deserializer. I have to admit that's a little bit funny.

Right now, if you want to build a service and expect any sort of predictability and stability, I think you have to go with some solution that lets you run open-weights models. Some have been de-censored by volunteers, and if you find one that works for you, you can ignore future "upgrades" until you find one that doesn't break anything.

And for that it's really important to write your own tests/benchmarks. Technically the same goes for the big closed LLM services too, but when all of them fail your tests, what will you do?

Maybe some people would prefer that. Not everyone will happily open up to another human, especially when people are the reason for the trauma...

Even if you try to look at the situation from a detached POV I think you can't help but be baffled at how much influence the US is giving up for no apparent reason. It always felt like a house of cards here in Europe - we rely on defense from the US, we store so much of our data there, and almost all of our stuff comes from China or its area of influence.

Now there's a guy at the top of the US who doesn't understand human interactions beyond anger, disgust and mindless loyalty. Or diplomacy beyond simple monetary transactions. He and his rich friend are getting rid of anything they don't understand (or that they don't understand to be of value to them right now), and apparently that's almost everything.

We're living in interesting times.

I wonder what game theorists are doing right now. All I know is that old joke about economists can't be true, about blindly assuming that everyone acts in their own best interest. People do factor in stupidity, right? But there must have been limits, and we're clearly far beyond those.

Probably any hardware that requires you to use some intermediate code. Nowadays probably some TPU or NPU or whatever.

This made me think of what "accelerators" I've come across (just as a regular consumer):

In the late 90s, that's what we called GPUs - "3D accelerators". For a short time, they were cards separate from the actual graphics card, and often would involve routing your VGA cable through that thing. Before it all merged into one device. I was very slightly disappointed as a kid that I narrowly missed that time, but bilinear filtering and higher resolutions and framerates helped me get over the fact I couldn't cram more cool weird tech into my PCI slots.

Then you had sound cards with "audio accelerators" using technologies like EAX. All that eventually migrated back to the CPU I think.

For a while you could by a "physics accelerator" for PhysX, then acquired by nvidia and moved to the GPU using CUDA. I never had one, but one time I kept around an older GPU after upgrading as a dedicated physx processor. Now that's the only way to run older 32bit games with physx turned up, since that's not supported in 5000 series GPUs.

Finally, I got this "Coral TPU", a USB device with 4GB RAM (and I think around 4 TOPS or something?) for very efficient inferencing. There are some open source projects supporting this, like frigate, which lets you process object detection with multiple surveillance camera streams on a raspberry pi. I never really used it though.

And of course now we have NPUs as sub-systems in CPUs and GPUs. I'd love to have a dedicated PCIe card with that, but of course having yet another computer architecture with dozens/hundreds of GB of redundant RAM is kind of a bummer.

If anything, this is just putting the huge disconnect between the ideals and the realities of higher education on display.

I studied CS because I love computers and programming and I wanted an SWE job. The idea was to earn a living doing something I enjoy / am good at, and I needed a degree for that. I didn't mind "learning how to learn", getting to know topics I probably wouldn't have chosen to learn on my own, meeting people pursuing the same goals and so on. That was nice, especially the idea of it, but it wasn't really my goal.

The reality was that during that time my primary worries were getting to lectures on time and passing exams while barely having enough money to afford anything beyond food and shelter. It wasn't me and my teachers engaging in the noble pursuit of science. I was worrying about exams while they were probably worrying about grading while taking a break from worrying about funding and authoring enough papers.

Thanks to some fluke I had a few weeks with almost nothing to do during one semester, so I decided to try and write a 3D engine. That was actually me learning just for the sake of learning: not thanks to the education system, but despite of it.

If AI brings this entire illusion down, maybe that's a good thing.

In a perfect world, we would have had a very broad conversation not just about AI in education, but about the role of education in society.

I don't think it has ever been a purely intellectual pursuit. Especially in societies that tell you that you'll never get a well-paying job unless you get a degree - students feel significantly more pressured to pass tests by any means necessary so they can present a certificate to their future employer than they feel encouraged to learn and grow. For the ones who are actually interested in their field it's probably something in between. But the deal is pretty clearly "Get to a prestigious school, make sure you don't spend any more time than necessary, and if you don't get a degree then it's been a huge waste of time and money".

Of course it's very disrespectful to just generate some AI slop and expect a human to read it and grade it, making them spend more time with it than you did. Of course there's value in training and shaping your mind in ways that will prepare you for not just your job but as a person in general. But the system isn't really geared towards that, and I missed that perspective in the article.

I don't think all that much has changed since the time when almost all scientists were wealthy aristocrats. The work itself is more accessible now because it can also be a job that puts food on the table. But if you want to pursue knowledge for the sake of knowledge, to the point where a degree is just some optional external validation of your efforts, you should be independently wealthy.

The admin API wasn't really official - at least I couldn't find documentation anywhere and had to piece it together from the source code. And regex filters didn't seem to work through that - didn't get an error back, but nothing changed either.

Finally a REST API!

I've been waiting for this - I wanted to play around with blocking distractions on various rules, but controlling pi-hole remotely was a huge pain and often didn't work until now.

Here in Germany, you often come across surnames that end in "-ski", which I assume come from Poland. That is also an example of getting "stuck" with a specific version of a name: in Poland, that ending would indicate that the person is male, while "-ska" would be female, and there's even one when referring to the family that shares that name: "-scy".

So whenever I meet a woman with a "-ski"-name, I wonder how many generations ago that name got stuck with just the male form.

I think most people who care about code have a bit of OCD over these things, but there is a difference between how the code looks vs. how it is structured. I think that's what the author means.

This is really complicated...

- NVIDIA has probably been overvalued for a while

- But what is happening now is clearly a panic

- GPUs are still needed for training and inference, and I guess people will figure that out in the next few days

- So far, all we have is claims from one source. Seems credible enough, and I'm sure there's massive optimization potential in training LLMs, but someone else should replicate the process first before it's taken as a fact

I got my first 3d printer, an MK3S+ a year ago. Pretty late in its lifecycle, but I wanted to spend more time printing than fixing issues.

And it definitely worked! I got the kit and built it within 10h or so (very enjoyable time actually, like building LEGO as a kid) and have printed lots of stuff ever since. During that entire year I only had a clogged extruder one time and had to take that apart a bit. Any other issues I've had were either due to bad filaments or my own errors (not taking long overhangs or low adhesion seriously while slicing).

And all this time I have been using it completely offline with OctoPrint on an RPi.

When I see someone use that word, it's almost always a clear knee-jerk reaction to:

- Media featuring women who aren't exclusively an attractive love interest or a very minor character

- Media featuring non-white or non-heterosexual people in major roles

I find it difficult to have any rational discussion of this topic because it always gets drowned out by overwhelming racism and sexism. You can't talk about overcorrection or virtue signalling without an army of angry white guys present who watch hours of "(some guy) DESTROYS wokeness with FACTS and LOGIC" on youtube every day.

That said, compared to disinformation and identity politics, this is a non-issue. It's a convenient topic to focus the anger and time of straight white men so they don't notice how billionaires and opportunistic politicians are taking their futures away while pretending to care about "free speech". This will have huge consequences not just for victims of racism and sexism, but literally everyone who isn't filthy rich.

I always wondered - why do we usually need to build software for a target architecture/system on the same architecture/system? Unless the build includes tests, it doesn't need to be run?

I get that Apple has its own BS reasons for that, but Linux?

Not to be confused with the "lemonscape", a hallucinated world you enter when you've eaten too many lemons.

A few days ago I made a 50x timelapse of a drive using my dashcam and... oddly enough, that looked exactly like that pseudo 3D effect. It looks like the road is just bending left and right, whipping around out of the blue, while everything else is stationary.

I love the ingenuity, I also love youtube documentaries on this topic, but... as a child (in the 90s), I did not like these pseudo 3D racing games at all!

Mode 7 on the SNES was usually fine. I don't know how accurately it rendered a single flat surface in 3D, but it felt real enough and responsive enough. Except for the very rare cases where they simulated non-flat surfaces (Speed Racer, Super Off-Road), even though that was technically much more impressive.

The effect just didn't work for me - it didn't feel like turning, it just felt like what it was: The game displaying a "left turn" animation and telling you that your car will now start drifting to the right if you don't press left. And that felt more like playing a Game&Watch toy.

How many of these people didn't understand that anyone could see their videos?

It might be a bit difficult for the highly technical HN crowd to grasp how little many people understand technology. Not changing the title is already a big clue. Since it was a feature built-in to a native app, people might have thought their videos would not be public or only shared with friends, and lots of them might not even have understood what they were doing at all.

I was experimenting with Python for some personal data-based project. My goal was to play around with some statistics functions, maybe train some models and so on. Unfortunately most of my data needs a lot of preprocessing before it will be of any use to numpy&co. Mostly just picking out the right objects from a stream of compressed JSON data and turning them into time series. All of that is easy to do in Python, which is what I'd expect for the language used this much in data science (or so I've heard).

But then I do have a lot of data, there's a lot of trial&error there for me so I'm not doing these tasks only once, and I would have appreciated a speedup of up to 16x. I don't know about "high performance", but that's the difference between a short coffee break and going to lunch.

And if I was a working on an actual data-oriented workstation, I would be using only one of possibly 100+ cores.

That just seems silly to me.

Yes please. I am currently driving a pre-touchscreen car and I hope I can just skip that era entirely.

Touchscreens are a menace. The most dangerous moments I have in my car are when I'm trying to skip the ads in my podcasts. Which got way worse since google removed default media buttons from maps. I bet that decision has an actual body count.

I live in Germany, and I've developed a similar rule with doctors in general. But here I can use an even better rule of thumb - be extremely suspicious of anything that isn't covered by public health insurance. I'm sure other people have had different experiences, but I'm 40 and so far almost every time I had to pay out of pocket it turned out to be some controversial or even pseudoscientific BS. The last one was some weird back pain therapy - I wasted my time and money on that until I read some paper on it. Its conclusion was actually that the therapy "probably works", but when I actually read it, the data ranged from "inconclusive" to "not effective". And that was the paper the company selling machines for that therapy was linking to.

What worked in the end was - surprise - lifestyle changes.

Just for yourself, write a short script for a youtube video. Try to make it engaging, even if it's just a paragraph. Then put your phone up on your desk, put your script next to it. Read the script and record yourself.

Also go out and meet more people.

My vacuum has often lost various parts around the apartment whenever it got caught on something, but so far it has never shat on the floor. It is quite a fancy one so that's expected.