Jake
@ngmi.co
techbro with philosophy degree working on global state machines that sync as fast as physics allows
Learning isn't stealing. Or if it is, then we've all stolen.
youtu.be/iv-5mZ_9CPY?...
youtu.be/iv-5mZ_9CPY?...
But how do AI images and videos actually work? | Guest video by Welch Labs
YouTube video by 3Blue1Brown
youtu.be
December 22, 2025 at 9:33 PM
Learning isn't stealing. Or if it is, then we've all stolen.
youtu.be/iv-5mZ_9CPY?...
youtu.be/iv-5mZ_9CPY?...
We should look into getting more AI data centers here in CT. They're funding upgrades to inefficient and dangerous water infrastructure like we just saw in Waterbury.
December 19, 2025 at 8:39 PM
We should look into getting more AI data centers here in CT. They're funding upgrades to inefficient and dangerous water infrastructure like we just saw in Waterbury.
It depends. For example AI is forcing companies to upgrade old an inefficient water infra. This hits close to home for me because a city close to me has been without water for a week as it's century old pipes started failed and losing millions of gallons.
Wish we had a data center here before that.
Wish we had a data center here before that.
December 19, 2025 at 8:36 PM
It depends. For example AI is forcing companies to upgrade old an inefficient water infra. This hits close to home for me because a city close to me has been without water for a week as it's century old pipes started failed and losing millions of gallons.
Wish we had a data center here before that.
Wish we had a data center here before that.
It's just time being a flat circle.
Writing = brain atrophy
AI = brain atrophy
Same things being said after thousands of years. Idk, maybe there's some truth to it. There's those people who say there were hyper advanced civilizations back then that are lost now. Maybe writing was their downfall.
Writing = brain atrophy
AI = brain atrophy
Same things being said after thousands of years. Idk, maybe there's some truth to it. There's those people who say there were hyper advanced civilizations back then that are lost now. Maybe writing was their downfall.
December 19, 2025 at 7:41 PM
It's just time being a flat circle.
Writing = brain atrophy
AI = brain atrophy
Same things being said after thousands of years. Idk, maybe there's some truth to it. There's those people who say there were hyper advanced civilizations back then that are lost now. Maybe writing was their downfall.
Writing = brain atrophy
AI = brain atrophy
Same things being said after thousands of years. Idk, maybe there's some truth to it. There's those people who say there were hyper advanced civilizations back then that are lost now. Maybe writing was their downfall.
Neither was Plato (for writing) fs.blog/an-old-argum...
Plato's Argument Against Writing
The people who invent something new, create a new tool or technology are not necessarily the people who are going to understand what the social impact of those inventions are going to be.
fs.blog
December 19, 2025 at 7:31 PM
Neither was Plato (for writing) fs.blog/an-old-argum...
Broadly to be a computational substrate that attempts prediction (and to uncover what that means sgd.fyi) For what purpose people use that for is up to them.
Just as a pencil is just a tool for marking with graphite, how it’s used and what type of markings depends on the user.
Just as a pencil is just a tool for marking with graphite, how it’s used and what type of markings depends on the user.
Built, or grown?
No, it wasn't… we designed the learning algorithm… But we don't really understand exactly how they do those things.
sgd.fyi
December 19, 2025 at 7:13 PM
Broadly to be a computational substrate that attempts prediction (and to uncover what that means sgd.fyi) For what purpose people use that for is up to them.
Just as a pencil is just a tool for marking with graphite, how it’s used and what type of markings depends on the user.
Just as a pencil is just a tool for marking with graphite, how it’s used and what type of markings depends on the user.
I don't think the person saying that is making an apt analogy. Anyone can prompt, just like anyone can hold a pencil and sketch a cube. But also the ability to easily do either well without practices drops quickly for both more advanced AI art tools and more advanced pencil techniques.
December 19, 2025 at 6:48 PM
I don't think the person saying that is making an apt analogy. Anyone can prompt, just like anyone can hold a pencil and sketch a cube. But also the ability to easily do either well without practices drops quickly for both more advanced AI art tools and more advanced pencil techniques.
But also how many users installed ad blockers after GDPR cookies? This is just my humble opinion but I think users will similarly just use the models that give best results, and if there's a way to avoid speed bumps (like GDPR cookies) they'll do it (even if the model didn't follow the reg)
December 19, 2025 at 6:10 PM
But also how many users installed ad blockers after GDPR cookies? This is just my humble opinion but I think users will similarly just use the models that give best results, and if there's a way to avoid speed bumps (like GDPR cookies) they'll do it (even if the model didn't follow the reg)
Like it's entirely possible we stop OpenAI or Anthropic from accelerating, but also if that happens and Wan or Qwen get 100x better, people will just use those models anyways
December 19, 2025 at 5:02 PM
Like it's entirely possible we stop OpenAI or Anthropic from accelerating, but also if that happens and Wan or Qwen get 100x better, people will just use those models anyways
People are probably going to use the model that works best for them. The US might make it illegal to train on copyrighted content, but it's likely going to be hard for an artist to sue Alibaba or Deepseek or some other China-based AI development co., and if those get better, users will use them
December 19, 2025 at 5:00 PM
People are probably going to use the model that works best for them. The US might make it illegal to train on copyrighted content, but it's likely going to be hard for an artist to sue Alibaba or Deepseek or some other China-based AI development co., and if those get better, users will use them
I'm actually curious why more people aren't mad about translation. Google Translate created the basis architecture that all gen AI uses today. It's trained on work from humans. Takes input, gens output. Replaced countless human translators. Made learning languages less necessary for most users.
December 19, 2025 at 4:34 PM
I'm actually curious why more people aren't mad about translation. Google Translate created the basis architecture that all gen AI uses today. It's trained on work from humans. Takes input, gens output. Replaced countless human translators. Made learning languages less necessary for most users.
I think I fall in this category. I’ve always enjoyed art. My first couple years of college was an industrial design major / art history minor. When I switched to phi I leaned into philosophy of aesthetics.
But how people *make* their art, e.g pencil, illustrator, comfyui, doesn’t matter to me much
But how people *make* their art, e.g pencil, illustrator, comfyui, doesn’t matter to me much
December 19, 2025 at 3:30 PM
I think I fall in this category. I’ve always enjoyed art. My first couple years of college was an industrial design major / art history minor. When I switched to phi I leaned into philosophy of aesthetics.
But how people *make* their art, e.g pencil, illustrator, comfyui, doesn’t matter to me much
But how people *make* their art, e.g pencil, illustrator, comfyui, doesn’t matter to me much