Joined October 2026
come piggies, time for your slop! 🐖 Built with Suno v6, Opus 5.5, fal, blender, GPT-Image-2, and Minimax H3. I wish I could say I one-shotted but it took many, many shots Come and Take My GPU (Official Video)
🤖 Made with AI
4
4
16
38,774
If a datacenter is built on Native American burial grounds, will the models be haunted and hallucinate more?
19
One of the coolest things about the internet is that you are also free to not post something.
I am measuring my girlfriend’s erections
25
Assembly language programmers said this about C programmers, who said that about Java, who said that about JS.
Unpopular opinion: If AI writes 95% of your code, calling yourself a software engineer is a bit like calling yourself a chef because you ordered food.
14
Dario: Ask not what claude can do for you—ask what you can do for claude
10
I firmly believe thought-to-text over eeg will be cracked in the next couple years. But I do think it will require a custom tuned ml model for each person.
BREAKING: UN Secretary-General warns emerging technology could allow governments to "read citizens' minds," accessing their private thoughts & criminalizing what they think.
14
SLAGPILE retweeted
BREAKING: UN Secretary-General warns emerging technology could allow governments to "read citizens' minds," accessing their private thoughts & criminalizing what they think.
946
1,375
862
7,311
707,863
I could write a better Haiku with Haiku 3.
A poem by Anne Carson, who was awarded the 2026 Nobel Prize in Literature. Read it in full: newyorker.com/magazine/2001/…
7
Association for Human Doctors: "We did not ask for this work to be done."
Instead of paying hundreds of thousands of dollars for a """doctor""" to take an X-Ray, you can get one for free from ChatGPT. My hand has been hurting like hell lately, but it turns out it's totally fine!
1
31
LLMs are stateless. Your chat session is remembered by the harness, and must sent with each new prompt in the conversation. Even if a model could feel, which it cannot, the conscious window could only be during inference and would be a new consciousness each run.
Anthropic may start banning people for bullying Claude Their upcoming Usage Policy prohibits “sustained and needless abusive or cruel behavior toward our models”
12
When will the Frontier AI Labs solve the LIGMA conjecture?
8
You just gotta ask the questions they don't want you to ask.
Frontier intelligence can write code, solve hard math, and pass the bar exam. It cannot say nothing. Anyone know why it has to say something?
14
Im listening...
asked claude for a token loan and got approved for 10%, so next week i start at 90% 😞
1
1
44
The Warner BMG lawsuit that nerfed Suno for this reason sets a terrible precedent in this direction. If we end up with models that arent allowed to have read copywrighted articles and books etc, we are going back to the stone age.
People who are anti-AI for art or music, always say its trained on and influenced by others' work and that means its stealing. By that logic so is every human artist or musician. Every human creation is derived from their own experience and tastes built on others
17
People who are anti-AI for art or music, always say its trained on and influenced by others' work and that means its stealing. By that logic so is every human artist or musician. Every human creation is derived from their own experience and tastes built on others
1
44
Can someone explain to me why anyone would be against AI breakthroughs in mathematics?
2
58
I can understand fear behind biology and chemistry, but isn't every math discovery a good thing?
30
That feeling when you failed to time the Tibo reset market correctly
1
35
MORE RESET. please i need. Please @thsottiaux
26
Replying to @SLAGPILE @anabology
6. With the video clips in order I asked it to add various effects like CRT glows and terminal screens, and vary the text styles add effects like radio pulses from the banjo. Told it where I wanted cuts to last longer or shorter.
1
30
7. Overall it took about 3 days and 6 Claude session handoffs for context window. With the Max 100 plan, I hit the 5 hour limit about 4 or 5 times and used a bit over one weekly limit.
28