Highlander is basically Wolverine
Posts by Gregg Furstenwerth
I still have my 2013 Tesla. It’s pretty good, has issues but still runs.
Why the hell are the tires spinning mine doesn’t do that the model s. It has awesome traction control. The X I use to have was awesome too.
Does it do the exact same thing? You bet.
It’s no wonder everything is so bloated and untrustworthy. Go boil a pot of pasta and throw it on your wife. That’s basically the same thing we do with all this complexity and expect them to use it.
“Fuck Best Practices: A Chainsaws Guide to Writing Code That Actually Works”
I’m thinking of writing a blog post titled this.
One file of code rendered from 400 lines monster to 63 lines. Same functionality. Imagine this across tons of files. We are talking thousands of lines of chainsaw.
When it’s gone it’s not coming back, the loss will take longer than we can imagine to recover to some new normal. Something that looks like this. Graveyard with algae.
The hope is that not everything is dying, that some survived. Then the following compounds down losses but hope clings as some survives. It is doing it in Hawaii where those photos were taken. In the Florida keys it did so to functionally extinct.
In 2013 there was these early phase bleaching photos and people were like it’s so pretty. Well this is what it looked like before things went seriously off the rails. It starts this way with hope.
Will miss it. Let’s see what the ocean becomes now.
Good deal. Not like I didn’t sound the alarm for a while now. Nobody seems to care. Just the way it goes.
Bye bye coral.
www.theguardian.com/environment/...
It’s clear what the limitations are. What the requirements are. What the scale issues are and what can be done with what hardware.
I’m curious why this is never spoken of? It’s a crucial distinction that explains quite a substantive problem and what solutions we don’t have.
@vtanger.bsky.social reading your most recent piece I always ask myself reading these. Why no one actually specifies what the actual hardware requirements are. It’s not like it’s difficult to determine. I’ve written my own AI inference engine.
futurism.com/fixing-hallu...
@vtanger.bsky.social you think the seahorse thing is weird.
Well the interesting thing is as you change angles of devices between devices through water they can get signals through. If you’re trying to send say Bluetooth through water.
Now I know people will say this is impossible. However run the tests yourself you’ll see you can send it minimal distance.
Well the nuance is how it disperses and reduces for sure. May take a ton to completely cut. However water does refraction and reflects. So low level is gonna be all but blocked. It will reflect it and distort it. Even with minimal amount. Its signal attenuation is going to be wasted.
6 meters is general deco. So yeah it’s deeper than 10ft most cases especially with equipment I use designed to go deep and do deco.
Call it a hammer cause it sometimes builds something other times it breaks it 🤣
The word tools annoys me as they’ve vaporized that words meaning with the AI tools lol
This is the problem is that it’s said to be a replacement vs assistant.
Computers are assistant not replacement for humans.
The only way to do it is with sound but then propagation of sound waves is not linear. Look up “sound is lazy” related stuff underwater and that even gets weird. The entire thing is a disaster to solve. It’s unlikely to be solved ever. Water is weird.
Imagine that essentially a building covered in 2 inches thick water beds would eliminate most bit flips. Pretty wild.
You can easily test this yourself doing RF and how far it penetrates water. I’ve done these experiments for doing links underwater.
Yeah water blocks the entire EM spectrum practically. Depends how much is required but like UV makes it deep but it’s like the equivalent of skin flakes to the depth of the ocean.
Dead short. In fact you could encapsulate a building with 2 inches of water barrier and block most of this.
With the invariant nature of this you cannot control for this sanely. The good news is that it’s generally rare. However as much as that is a problem. That’s why you always have analog backups. They can’t bit flip and are bound by laws of physics.
Ohh yes I’m very aware of all this.
However that nuance is less of a concern in the timeframes I’m speaking of in this.
Remember gamma and X-rays don’t penetrate water 🤣
The coolest thing I have made is a life support system that runs my rebreather. That was immense challenge to make something that never crashes. It can’t crash, it can’t fail, it can’t break.
That’s quite large, what I’m writing is about 50k lines of C currently. As much as people dump on it these days. I really like the language for its speed and hardware interaction.
Nothing huge but it’s compact C. I won’t use it unless I can do static memory though.
Easier to just use tailwind lol
Well the larger issue I see is AI increasing productivity and instead of making lives better. Reducing workload it will increase it. If you can be 10 times more protective, then you must work that much harder and keep at that level. Expecting people to take on that workload is recipe for burnout.
Same to you. Thanks for the reply.