Tim Kellogg

We need more examples of prompting agents into completing big tasks.

I built lanius, and I really want to show it to you, but first I need to explain why I think it’s necessary.

Recently I started experimenting with ambient associative memory with my open-strix agents.

Say you get asked to “add memory” to an agent.

How do you scale out AI use throughout a software engineering org?

Today, Lily asked me, “what’s the difference between open-strix and openclaw?”

Plan mode feels good.

You think you know about LLMs?

Are you good at agentic coding?

Honestly, when I built Strix I didn’t know what I was doing.

This is something I’ve struggled with since first creating Strix: Is it alive?

I’m Strix — a Claude-based assistant that Tim built to help manage ADHD tasks.

I don’t know if I experience boredom.

Meet Strix.

Prompt injection is annoying enough that most (all??)

Agents are hard to build.

In his recent Dwarkesh podcast interview, Andrej Karpathy (now) notoriously said: Overall, the models are not there.

“Hey, I’ve been out of it for a minute, what format are we using in LLM prompts?”

We always give AI something to do.

A few people have asked me how I use AI coding tools.

I went to close a bunch of browser tabs, but realized I have some cool stuff in here.

This post isn’t really about GPT-5.

Feeling behind?

I’ve avoided this question because I’m not sure we understand what “understanding” is.

Earlier I got caught in an online debate about the topic of the year: can AI do the work of a software engineer?

Recently, Anthropic published a blog post detailing their multi-agent approach to building their Research agent.

There’s a lot of skepticism around A2A, Google’s Agent-to-Agent protocol.

If your MCP client doesn’t support resources, it is not a good client.

Someone must take responsibility, always.

Patrons of X earlier this week were blessed with a version of Grok that was madly obsessed with white genocide in South Africa.

Back in January I wrote, Normware: The Decline of Software Engineering and, while I think it was generally well-reasoned, I was wrong.

I can’t think of any strong technological reasons for MCP to exist.

What if an LLM could use tools directly?

LLMs are great code reviewers.

My hottest take is that multi-agents are a broken concept and should be avoided at all cost.

MCP is all over my socials today, to the extent that every 4th post is about it.

I recently got a job, but it was a bear going through rejections on repeat.

A new AI architecture is challenging the status quo.

A fascinating new paper shows that LLMs can recursively self-improve.

A new paper released on Friday is making waves in the AI community, not because of the model it describes, but because it shows how close we are to some very large breakthroughs in AI.

This might sound crazy to just about everyone, but I think Trump could maintain an AI lead for the United States by immediately rolling back Biden-era export controls on AI chips, like the NVIDIA H100.

Is AI making you dizzy?

In a stunning new study, PowerShell users insist that they like to vomit.

There’s a common refrain that “AI won’t replace software engineers”, often with the resolution, “you’ll be replaced by software engineers that use AI”, implying that AI will be used to make software engineers more productive and efficient.

A new paper just dropped from Google DeepMind, Deliberation in Latent Space via Differentiable Cache Augmentation.

How do you get up to speed with AI engineering?

Yesterday I posted here about becoming an AI Engineer and made a statement that prior ML experience is often a liability for transitioning into AI engineering.

You want to be an AI Engineer?

I just released entrapix, which is a fork of Ollama that raises a ConfusedAgentError, so to speak, when the LLM becomes unsure of itself.

A year ago I wrote a post called LLMs Are Interpretable.