In the 2010s, if you had technical gaps (say, you couldn’t write CSS), you had to either rely on a skilled colleague or just hope that the answer to your exact problem was out there on the internet. Today, everyone can write sort-of-okay CSS by delegating the task to an LLM....
Before large language models, people who worried about AI safety often talked about the “boxing problem”. It goes like this. Suppose some genius figures out artificial intelligence in a late-night coding session on their laptop. Because they’re a genius, they’re smart enough...
Here’s the big idea in Keith Johnstone’s book Impro:
Children are naturally creative, but are violently formed into repressed adults by Western culture and education
The process of becoming more creative and expressive is largely a process of unlearning these habits of repr...
The anonymous blogger Gwern recently completed a thirteen thousand word post called Human-like Neural Nets by Catapulting, in which he offers a theory about why LLMs don’t possess truly flexible human-like intelligence, and how we might train LLMs that do. Theories like this...
Software engineers are often told to “start playing politics”, but most engineers have no idea what that means.
Their reference point for “playing politics” comes from fiction like Game of Thrones. Are they supposed to raise an army and depose the CEO, or poison each other a...
As a software engineer, how well do you have to understand your own codebase?
My guess is that people who work on small codebases with low-turnover teams (say, Redis or games like The Witness) would say “obviously you have to understand it completely, otherwise you can’t do ...
Every post I publish represents at least two things I’ve learned: the thing that prompted me to write the post, and the thing I learned in the course of writing it. If I don’t learn anything new while I’m writing, it’s not interesting enough to publish.
Typically I learn way...
The European Union AI Act is Europe’s attempt to comprehensively regulate AI usage. A big part of that is the requirement that AI-generated content be identifiable: either tagged with a watermark or with what the Act calls “digitally signed metadata”1. Since all this becomes...
The European Union AI Act will begin to be enforceable in August 2026, one month from now1. One of the biggest new requirements is Article 50, which requires all AI outputs to be “detectable as artificially generated”. In other words, if LLM providers want to do business in ...
Stating the obvious is surprisingly useful. Most of your knowledge lives below the threshold of conscious awareness, so it’s possible for a piece of writing to remind you of what you already know. It’s common to know you don’t like something without being quite sure why, and...
Many people claim that AI inference is unprofitable to serve, and thus must be subsidized by an ocean of dumb money from investors who believe that some future AI model will come to dominate the world economy. When that dumb money goes away, so will AI products. According to...
People who think current AI use is unsustainable often rely on the claim that inference GPUs only last “three years at the most” under load1. The idea here is that once the AI bubble money drains away, current infrastructure will rapidly become obsolete, and there won’t be e...
Many engineers should be doing less work. I don’t necessarily mean producing less code or fewer changes, but literally working fewer hours in the day. When they do work, they should be working at a slower pace. I like to aim to be running at 80% utilization by default: unles...
The relationship engineers have with product management is more dysfunctional than with any other part of the company. There’s no shared culture or language like there is with other engineers, and the rules of “who gets to tell who what to do” aren’t as clear-cut as they are...
Programmers were better back in the day, weren’t they? Back when we had real programmers. Not just people who got paid to write code, but people who lived it, who were obsessed with their craft, and whose code was a lively expression of themselves. Hackers were hackers in th...
Where are all the AI-generated projects? This is a common question from AI skeptics: if LLMs are so good at writing code, where is the tsunami of new AI-generated apps, services and games?
I personally don’t find this to be much of a paradox. Writing code is only one of the ...
There are only two ways to use LLMs in a computer program: as part of a pipeline, or as an agent. In other words, either you express the control flow of the program in code, or you give a LLM tools and allow it to manage the control flow itself1.
Here’s how you might structu...
In April last year, Kelsey Piper discovered that OpenAI’s o3 model was surprisingly good at figuring out where a photo was taken from. Like human “geoguessr” pros, o3 could sometimes take a nondescript photo of a beach and tell you exactly where it is. Here’s the example Kel...
It’s common and correct to say that “all code is technical debt”. Adding code is a necessary evil for developing new features: you almost always have to do it, but each line of code adds to the complexity and maintenance burden of the system. All future changes to the system...
The engineer who says no all the time is a real archetype among senior and staff engineers. Their role is to slow things down, to block the development of features that add complexity, and to ensure that as little code gets written as possible (since code is a liability).
We...
A bit over a year ago I wrote How I use LLMs as a staff engineer. Here’s a brief summary of what I used AI for last year:
Smart autocomplete with Copilot
Short tactical changes in areas I don’t know well (always reviewed by a SME)
Writing lots of use-once-and-throwaway rese...
Ever since Golden Gate Claude I’ve been fascinated with “steering”: the idea that you can guide LLM outputs by directly manipulating the activations of the model mid-flight.
DeepSeek V4 Flash
I was inspired to write this post by antirez’s recent project DwarfStar 4, which is...
This year Elon Musk has started banging the drum about building AI datacenters in space. As the only person who owns a successful space company and a (moderately) successful AI company, this is a sensible way to boost his profile and net worth. Is it a sensible way to build ...
Thinking Machines just released Interaction Models. This is their first real AI model release1 after a year of work and two billion dollars of capital. What is an “interaction model”? First, it’s not a frontier model. Thinking Machines is not yet competing with OpenAI, Anthr...
In Many anti-AI arguments are conservative arguments I argued that left-wing anti-AI sentiment1 is partly a backlash to two unrelated events around the rise of ChatGPT: the crypto mania of 2022 and the pro-Donald-Trump push many big tech CEOs made in 2024. If the timing had ...
Like other kinds of puzzle-solving, software engineering ability is strongly heavy-tailed. The strongest engineers produce way more useful output than the average, and the weakest engineers often are actively net-negative: instead of moving projects along, they create proble...
Incidents are boring. Most of what you actually do during an incident is wait: for some other team to investigate, or for a deploy to finish, or for the result of some change to become apparent, or for someone else who’s been paged to come online. It’s stressful, but there’s...
Dwarkesh Patel1 recently posted an award for the best answers to four key questions about AI. It’s partly a challenge and partly a job interview, since some of the winners will get offered a role as a “research collaborator”. I don’t want the job, but I do want to write down...
The most influential piece of writing about staff engineers in the last decade has to be Will Larson’s Staff engineer archetypes. He argues that the “staff engineer” title covers at least four very different roles: the team lead, the architect, the solver, and the right hand...
I don’t think there’s compelling evidence that using AI makes you less intelligent overall1. However, it seems pretty obvious that using AI to perform a task means you don’t learn as much about performing that task. Some software engineers think this is a decisive argument a...
