Romin

I want to make /init more useful- what do you think it should do to help setup Claude Code in a repo?

🚨 BREAKING: Tencent has killed the “next-token” paradigm. Tencent and Tsinghua has released CALM (Continuous Autoregressive Language Models), and it completely disrupts the next-token paradigm. LLMs currently waste massive amounts of compute predicting discrete, single tokens through a huge vocabulary softmax layer. It’s slow and scales poorly. CALM bypasses the vocabulary entirely. It uses a high-fidelity autoencoder to compress chunks of text into a single continuous vector with 99.9% reconstruction accuracy. The model now predicts the “next vector” in a continuous space. The numbers are actually insane: - Each generative step now carries 4× the semantic bandwidth. - Training compute is reduced by 44%. - The softmax bottleneck is completely removed. We’re literally watching language models evolve from typing discrete symbols to streaming continuous thoughts. This changes the entire trajectory of AI.

We are so unbelievably early with agents right now. The majority of companies aren’t even using coding agents at scale, let alone for the rest of knowledge work. We’re still mostly in the chatbot era of work for most of AI right now. Diffusion of tech takes time, even in the most breakneck of markets, because there are major workflows that need to be reinvented, any regulated or large business has huge governance processes for deploying new tech or agents, data needs to get into well-organized environments, and there’s technical literacy that needs to be established. All things that get solved, but takes time nonetheless. A point of comparison for technology diffusion: in 2010, a time by which every person in silicon valley knew that cloud was the future, AWS revenue was $500 million, Azure had only launched that year, and GCP was called Google App Engine. By 2025, these 3 platforms generated around $225 billion in revenue. And that’s only about 60% of the cloud market. So from the moment the tech industry saw the future of cloud to today, the market is nearly 1,000 times bigger. And it’s still growing at an insane rate. The same will happen for agents. Coding agents are like the early days of cloud computing when developers got on board for initial use cases. Then came the bigger workloads. This gives you a sense for how early we actually are in this transformation.

STRIPE NOW LETS YOU PICK YOUR AI MODELS, SET YOUR MARKUP, AND BILL CUSTOMERS FOR LLM TOKENS AUTOMATICALLY. BUILDING AN AI WRAPPER BUSINESS JUST GOT A LOT EASIER.

I'm squarely in the camp of believing that there is an insatiable demand for software, and until that stops, software engineers are going to be in overwhelmingly short supply in the long term. AI is forcing companies to recalibrate their resources (we're all seeing that now with layoffs and freezes), but once AI is more-or-less saturated among most companies, hiring more engineers will, once again, be the only way to build more product than the competition. And yes, being better than the competition is all that matters. Not being better than the competition 2 years ago. The pre-AI world is gone and not coming back. Readjust your expectations. Companies that think they'll be able to last with tiny headcounts are going to get crushed by the companies who choose to grow and force AI adoption across large teams. Yes, there's a brief period where adoption hasn't fully hit yet, but once it has, the low water mark for productivity will adjust and you still need more people to build more product, unless you have some secret AI sauce that no one else has access to (indie solopreneur building saas products, that's not you!).

prediction cone/safe triangle — this is something we take so much for granted in modern day native UIs. but it's not the same for most web-based dropdown menus. it took me a while to implement this here. Amazon, macOS, Windows all implement some version of this.

Most people take melatonin right before bed. That might be one of the worst times to take it. Melatonin isn't a sedative, it's a chronobiotic. It doesn't knock you out. It signals to your circadian clock that it's dark. And the effect it has on your sleep timing depends entirely on WHEN you take it. There's a concept called the Phase Response Curve. It shows that melatonin taken ~3 hours before your usual bedtime produces the largest phase advance, meaning it shifts your internal clock earlier, so you fall asleep sooner and wake up easier. But take it AT bedtime? You can actually push your clock in the wrong direction. That's a phase delay. You end up falling asleep later over time, not earlier. The sweet spot sits just before dim light melatonin onset, the point when your brain would naturally start releasing melatonin. For most people, that's roughly 3 hours before habitual bedtime. Timing > dose. Lewy et al. (1998), Burgess et al. (2010), and Challet et al., J Pineal Res (2024).