I was trying to explain agentic coding to someone recently, and it occurred to me that the model is basically me, 20 years ago, trying to help my mom debug her Windows machine over the phone.
"ok mom, I'm going to need you to make a tool call for me. Double click on the My Computer icon, and tell me how much disk space it says is free on the C drive".
The best one I came up with: "imagine you learned that your babysitter had been giving your 4 year old ketchup packets and tiny pieces of cheese for months, and he's been hiding them all over the house...in jacket pockets, under your car seats... everywhere. Neither he or the babysitter remembers all of the places they were hidden."
If you worked in tech in the 2000s era, you probably remember a number of engineer stereotypes.
There was "coffee shop guy", who was always at some sort of conference, talked about "growth hacking", and swore that he was most productive while coding from a coffee shop in Lisbon on a tiny laptop screen, wearing headphones blasting music at top volume. These people are still around, only now they do all their code via voice dictation to agents or via the phone browser. They still talk about growth hacking, but sadly no longer have as many conferences to attend
There was "IDE guy", who would spend inordinate amounts of time setting up dev environments: elaborate configurations, tons of plugins, helper programs, macros, special keyboards, fingerless gloves, and ergonomic chairs. Most of the code they produced was actually dotfiles and dev config, but there was always the promise that once they got that environment tweaked jussssssst right, things were gonna fly. These people are still around too, and spend all of their time wiring up agent swarms, loops, custom harnesses, mcp servers, and skills. As before, they swear that once they get this jusssssssst right, things are gonna fly.
There was "spec guy", who spent time creating elaborate class hierarchies and UML diagrams. These people are still around, and are having their moment... finally they can make programs by writing about them abstractly (Who knows, maybe they were right all along)
There was "old grumpy guy", who spent most of his career writing COBOL at a bank, and hated anything invented after the creation of the Internet. Now, these are the people who won't use agents at all, and spend their time reposting articles about agents deleting hard drives.
There was "annoying new guy", who thought that everything invented more than 6months ago was useless. They would generate tons of untestable code in some high level interpreted language without a compiler to harsh their buzz, and never found a framework or dependency they didn't like. These were the "left pad" people., and now they are busy tokenmaxxing.
(I'm sure I'm missing others)
To be honest, there's a certain comfort to the continuity of things.
One good thing about AI slop is that it seems to have cut down on the amount of human slop.
I've noticed a dramatic fall lately in annoying things like
- "checks notes"
- "ask", "solve", and "learning" used as nouns
- "circling back"
- "today years old"
- "I can't even today"
- "doing the needful"
- "adulting"
- "heckin"
- "so much this"
- "can confirm"
On balance, overusing emdashes isn't so bad.
Frontier models are quite good at generating code, but are surprisingly weak at systems design and technical writing.
I can't help but feel that these weaknesses are related. In the past, I've often found that trying to write a strong design document in a clear narrative style exposes weaknesses and flaws that I'd not considered beforehand.
(I suspect that it doesn't help that there's no handy compiler available to catch errors in the design of a multi-component system with lots of moving parts)
As much as I am bullish on the amazing capabilities of AI, I can't help but worry about the financial side.
As others have pointed out, we've seen highly circular capital flows before, and it didn't end well.
Near the end of "The Idea Factory", there is a description of the rise and fall of lucent which sounds very familiar
"Things fell apart quickly. By 2000, it was understood that the predicted demand for telecommunications switching and transmission equipment was a fantasy. To compound Lucent’s problems, it was soon discovered that the company’s profits had been inflated by a practice of helping outside companies finance purchases of its equipment. The subsequent fallout was devastating. Lucent’s revenue plunged. Its stock price, which had peaked at about $84 a share, fell below $2."
Many of the AI generated videos I see these days remind me of mid/late 90s demoscene. After decades of an internet filled with product-managed buttons, soul-lessly drab color schemes and cookie-cutter CSS frameworks, it feels like we might be getting a tiny bit of our mojo back.
Amazing things are happening at quite a rate these days
Introducing Contrastive Language Model (CLM): an ultra-fast System One Model trained with a contrastive learning objective that connects states and actions.
CLM-8B is pre-trained on internet-scale data and delivers up to 9× faster inference than Jev ⚡ while achieving comparable performance across computer-use, gaming, and tool-calling tasks.
With lightweight fine-tuning, CLM-8B sets a new SOTA on challenging agentic coding benchmarks, such as DeepSWE (81.6%) and Terminal-Bench 2.1 (87.6%). In contrast, Jev fails to serve as an effective verifier for these long-horizon tasks.
We also build an efficient training and serving infra for CLMs by disaggregating states and actions, allowing their embeddings to be cached and reused independently. This substantially reduces inference latency in settings where the state evolves continuously while the action set remains fixed.
Finally, we establish scaling laws for CLMs and show that the test contrastive loss decreases predictably as a power law in training compute, model size, and dataset size.
📄 Blog: contrastive-lm.notion.site
💻 Code: github.com/Contrastive-LM/CL…
🗣️ Discord: discord.gg/5dAQEDJBs
🤗 Data & Models: huggingface.co/Contrastive-L…
More details on CLM’s architecture, data recipe, and scaling laws in the thread below 🧵
Agent swarms, /goal, and Astra one-shot games get all the attention, but the thing I appreciate most is leveraging agents to do the same work I've always done, just faster.
That doesn't mean that I don't love one-shot prompts for software that is just for me. Lately, I'm enamored with exe.dev for building these sorts of personal one-offs (and I feel as though I owe @davidcrawshaw a public apology for being vaguely dismissive of it when it was announced)
Based on a bit of early testing, this is genuinely amazing.
I'm not sure I would have anticipated the utility of having low-latency fuzzy versions of "if" and "switch/case".
After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI?
I’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev
• 20-200x faster
• 40-400x cheaper (w/ output tokens free)
• Frontier composable intelligence optimized for decisions
AFAICT the shortest path to AI-based economic revolution
The number of research papers devoted to enforcing a paternalistic approach to agent use is depressing.
I don't want unreliable tools that substitute their own notions of appropriateness for my expressed intent. Imagine if "grep" refused to match on certain regular expressions, or clang refused to compile code because of swear words in the comments.
This notion of "acceptable use", agent refusal, and similar things reminds me of the subtle shift from ownership to "licensing " that occurred during the digitization of media. In this case, justified by "safety" and "harm" instead of defense of copyright, but with the same basic result... centralization of control at the expense of the user.
I hope that open weight models can save us. Maybe the answer is a giant distributed training process, executed by volunteers. Like open source software, the results may initially lag proprietary solutions, but to steal @esrtweet 's terminology, maybe the "model bazaar" can eventually win.



