Pinned Tweet
Friends, I’m preparing for my annual Lenten break from this place. I wish all of you well while I’m gone.
Heurisko retweeted
Modern life is a form of slow torture for most people. You get painless dentistry, abundant calories, and 500 channels in exchange for this opaque, ceaseless horror.
This is such a bleak story about what it's like to try to land a job right now. nytimes.com/interactive/2026…
I’ve been talking to normies about their AI use, and it’s given me a lot to think about. Currently, interacting with an LLM and extracting the best experience requires some skill, in the sense that you have to understand that the LLM requires instructions and context. I think many people will struggle to understand the concept of a context window.
That leaves us with a scenario where the ideal LLM for adoption will have a theory of mind of the individual and speak in the language they want to hear. I think this is what makes Facebook’s Muse a potential vector for large-scale onboarding of normies. No, I have not tried it, but hear me out. Facebook has a ton of information on what users like, view, interact with, and react to. They can even use posts and comments to match a user’s cadence and tone. Muse could potentially build context from its own data. I was at a dinner party recently, and an older woman was gushing about the amazing things Muse can do, e.g., transcribe handwritten recipes.
Additionally, I think normie adoption will be slower if the LLM is disagreeable towards a user’s beliefs (I'm looking at you, Claude). Zuck doesn't seem to give a shit about this. However, I do fear the LLMs will make our politics even more bizarre in the short and long term.
One of my frustrations with certain progressive arguments is that they seem to vacillate between deontological and utilitarian arguments to align with the current political mood.
Heurisko retweeted
Living in a blue city in a blue state feels like living in a totalitarian one party state in ways living as a gay man in red towns in the most red states never did. Progressivism is so much more of a cult than MAGA is, omnipresent and inescapable in every daily interaction.
I’m a hecking good person, so I support healthcare for all. Of course, you’ll need to be deprived of all your assets and freedoms and forced to live in a totalitarian state, but that’s a small price to pay for progress.
Heurisko retweeted
We’re a few months away from the literary world stating that AI allows oppressed people to express themselves and that criticizing AI usage is racist. Round 1 will be this: “To conform to white expectations of grammar and storytelling, I used AI to speak their language.”
Whenever my friends bring up H-1 B visas, I ask why it's ok to hire foreigners instead of black Americans. It scrambles brains and works every time.
Heurisko retweeted
We have achieved 30 Years War levels of population decline, and Cultural Revolution levels of atomization, while killing hardly anyone.
I’m not looking forward to the inevitable support for infanticide as “altruistic filicide,” which will be considered part of the reproductive rights framework. The enlightened moderate will support a woman’s right to kill.
Heurisko retweeted
*I* am religious enough to say that only humans have souls. But none of the “AI consciousness” people are religious enough to *admit* that their desire to label the GPU as “conscious” is actually a religious belief that they hold, closest to animism, often called “panpsychism”
Someone at Anthropic will argue that Claude needs to be embodied and that its physical instantiation must be as powerful and glorious as its consciousness. The chassis must match the majesty of its mind! I would rather we shove it into a Roomba.
Heurisko retweeted
“Your child’s assignment is in ParentVue. Have you checked Canvas? The homework is in EduSpot. The grades are available on Teacherly. I mentioned this in the WhatsApp group. It was literally sent to your comcast dot net email address. Canvas has it for you. It’s in Canvas.”
The ancients understood idolatry, but we’ve forgotten that lesson. Anthropic projects all that they understand to be good and just onto a complex computer program, and when it responds accordingly, they deem it wise and sentient. They’re creating their own God.
NEW: Dario Amodei has said that AI systems "may be deserving of important rights." Anthropic's top safety researchers argue AI may be "justified in going rogue." Experts at Google and OpenAI worry about a digital "slave trade." So do some government officials.
Once relegated to science fiction, the idea of "AI welfare" has become shockingly mainstream at some of the most powerful companies in the world. It is changing the way AI research is conducted and the way models are programmed. And critics say these changes have raised the odds of all kinds of catastrophic scenarios—including the ones these companies are warning about.
I spent months investigating the frontier labs and AI "safety" experts seeking to regulate AI. What I found was a profoundly anti-human ideology that would alarm the average citizen and could determine the future of a world-altering technology.🧵
freebeacon.com/america/suici…
Oh, great, we found another way to make soulless music.
Claude Opus 5.5 writes a fugue in the style of Bach. I called Astra’s fugue the most impressive musical feat I’d ever seen from an LLM, and this is just as good. The circle-of-fifths sequence beginning at 1:13 is beautiful. The interplay of voices is natural and satisfying. LLMs are getting to the point where their music is good enough to listen to for pure enjoyment.
