Pinned Tweet
This is unsurprising to anyone who watched the Netflix doc, which is still available btw
netflix.com/title/80224905
Jaron retweeted
Myabe this'll fix the quality but dayum that edit is fire!
#Wutheringwaves #WuWatwt
Jaron retweeted
是的，我一开始也以为美国人被什么阴谋论洗脑了。结果一看，数据中心建你妈居民区脸上，方圆一公里都有让人睡不着的噪音。中国的数据中心一口气建到新疆去散热，城市里找不到这玩意。美国人要是知道中国人有多娇气，可能会生气的。
美国人民还是太能吃苦了
红脖子们抵制数据中心建设，一开始以为是知识水平不够，不识发展先进生产力大局。这两天网上看到图片，直接把数据中心建到人家门口，抛开电费上涨和没带动当地就业不谈，自家好好的美式大house成了噪声房，这你受得了？
数据中心配备燃气轮机，24小时开机，厂区边界噪音达到80分贝，相当于飞机在你头上天天飞。这种噪音低频共振，和车流那种随机噪音完全不一样，夜间对睡眠的破坏力远高于普通噪音，很多居民头疼恶心失眠。
此外燃气轮机排泄氮氧化物、一氧化碳和甲醛等等有害气体，居民反馈闻到类似腐烂白菜的刺鼻气味。
为什么上燃气轮机？因为电网基建不行，新建电网至少两年起步，耽误数据中心赚钱，于是申报成临时应急发电机，监管和环评标准宽松很多，直接燃气轮机进场，先通电跑算力赚钱。这类项目大量放在低收入和少数族裔社区，地价便宜。
这下真的是制度劣势了
Jaron retweeted
Was literally explaining the, "society falls when old men don't plant trees they will never sit under" to my neighbor (boomer)
He said he didn't get it and why would he plant trees if won't be around to see them
Jaron retweeted
It actually IS the role of government to provide for you. Living in a society is a trade off of relinquishing some of your personal sovereignty for stability and material comfort. If the government doesn't provide you with ANYTHING you're just being oppressed
Jaron retweeted
Every billionaire's "started in a garage" story conveniently skips the parental loans, elite connections, and safety nets. Bezos got $300k from his parents. Gates' mom sat on IBM's board. Musk's family owned an emerald mine. The garage is just aesthetics; the capital was inherited. Stop calling it self-made when it's generationally funded.
Jaron retweeted
After you shower and moisturize, you don't put your dirty underwear back on, do you?
This post is about Hakeem Jeffries and Chuck Schumar.
Jaron retweeted
Sad to see the Cybertruck going away when it was a useful way to learn who the dumbest, most gullible people were in your parents’ neighborhood
Jaron retweeted
I don’t know who needs to hear this but if the market is wildly selling off AI stocks because of the election results then it’s safe to assume the value wasn’t in the product but in the administrations willingness to let Silicon Valley fellate each other on your dime.
Jaron retweeted
ChatGPT has now a big problem.
Researchers at Oxford exposed a massive threat to large language models.”
It’s called "Model Collapse."
And it explains why the internet is about to become a toxic waste dump for Large Language Models.
Right now, AI companies are scraping the entire web to train their next-generation models.
But there’s a massive problem.
The internet is no longer just human data. It is rapidly filling up with AI-generated text, images, and code.
What happens when an AI trains on data generated by another AI?
A new paper in Nature just gave us the answer. And it is terrifying.
When AI models are recursively trained on AI-generated data, they suffer from irreversible brain damage.
The researchers found that the model starts forgetting the "tails" of the original data distribution. The rare, weird, unique, and complex human nuances completely disappear.
It's like making a photocopy of a photocopy.
Every generation loses resolution. The output gets flatter, blander, and more generic, until the model completely collapses in on itself.
For years, we thought the biggest bottleneck for AI was compute, getting enough H100 GPUs from Nvidia.
This paper proves the real bottleneck is us.
If we keep flooding the internet with synthetic AI slop, we will poison the very data pool required to build smarter models.
The ultimate irony? We thought AI was going to make human creators obsolete.
Instead, this research proves that verified, genuine human data is about to become the most valuable commodity in the tech industry.
Cyberpunk2077 was supposed to be fiction, guys.
ChatGPT has now a big problem.
Researchers at Oxford exposed a massive threat to large language models.”
It’s called "Model Collapse."
And it explains why the internet is about to become a toxic waste dump for Large Language Models.
Right now, AI companies are scraping the entire web to train their next-generation models.
But there’s a massive problem.
The internet is no longer just human data. It is rapidly filling up with AI-generated text, images, and code.
What happens when an AI trains on data generated by another AI?
A new paper in Nature just gave us the answer. And it is terrifying.
When AI models are recursively trained on AI-generated data, they suffer from irreversible brain damage.
The researchers found that the model starts forgetting the "tails" of the original data distribution. The rare, weird, unique, and complex human nuances completely disappear.
It's like making a photocopy of a photocopy.
Every generation loses resolution. The output gets flatter, blander, and more generic, until the model completely collapses in on itself.
For years, we thought the biggest bottleneck for AI was compute, getting enough H100 GPUs from Nvidia.
This paper proves the real bottleneck is us.
If we keep flooding the internet with synthetic AI slop, we will poison the very data pool required to build smarter models.
The ultimate irony? We thought AI was going to make human creators obsolete.
Instead, this research proves that verified, genuine human data is about to become the most valuable commodity in the tech industry.
