Mike Brummett

code mode: let the code do the talking (aka, after w/i/m/p) wherein I ponder the implications of every user having a little coding buddy, and every "app" being directly programmable on demand. sunilpai.dev/posts/after-wi… lmk what you think.

Criminally underrated ♥️

@AaronCQL A couple more you should give a shot: - [Variant UI](variant.com) - [Uncommon](unc.mn)

On a quest to try every AI design tool out there. So far I've tested Pencil, Paper, and Figma MCP. What else should I try? Mainly looking for fast iteration, design quality, and ease of use as a design noob (ie. Good built-in prompts and templates). Will review each one.

If you remember AutoGPT & BabyAGI back 2023 The ideas were fansinating, but model simply not ready However, something big shifted in Dec 2025. LLM didn’t just get better - it crossed a threshold. The key unlock is harness engineering + Models: • Environments for memory • Verification loops • Generic atomic tooling This opens up the biggest opportunity in 2026: We’re moving from copilot agents (assistive tools) → to long-running autonomous agents that is always-on. Here are my learnings

It is hard to communicate how much programming has changed due to AI in the last 2 months: not gradually and over time in the "progress as usual" way, but specifically this last December. There are a number of asterisks but imo coding agents basically didn’t work before December and basically work since - the models have significantly higher quality, long-term coherence and tenacity and they can power through large and long tasks, well past enough that it is extremely disruptive to the default programming workflow. Just to give an example, over the weekend I was building a local video analysis dashboard for the cameras of my home so I wrote: “Here is the local IP and username/password of my DGX Spark. Log in, set up ssh keys, set up vLLM, download and bench Qwen3-VL, set up a server endpoint to inference videos, a basic web ui dashboard, test everything, set it up with systemd, record memory notes for yourself and write up a markdown report for me”. The agent went off for ~30 minutes, ran into multiple issues, researched solutions online, resolved them one by one, wrote the code, tested it, debugged it, set up the services, and came back with the report and it was just done. I didn’t touch anything. All of this could easily have been a weekend project just 3 months ago but today it’s something you kick off and forget about for 30 minutes. As a result, programming is becoming unrecognizable. You’re not typing computer code into an editor like the way things were since computers were invented, that era is over. You're spinning up AI agents, giving them tasks *in English* and managing and reviewing their work in parallel. The biggest prize is in figuring out how you can keep ascending the layers of abstraction to set up long-running orchestrator Claws with all of the right tools, memory and instructions that productively manage multiple parallel Code instances for you. The leverage achievable via top tier "agentic engineering" feels very high right now. It’s not perfect, it needs high-level direction, judgement, taste, oversight, iteration and hints and ideas. It works a lot better in some scenarios than others (e.g. especially for tasks that are well-specified and where you can verify/test functionality). The key is to build intuition to decompose the task just right to hand off the parts that work and help out around the edges. But imo, this is nowhere near "business as usual" time in software.

Introducing SpacetimeDB 2.0. Web development at the speed of light. Come learn what fast really means.

Code Mode is all you need, very excited about this direction for MCP blog.cloudflare.com/code-mode-mcp/

@nikgeneburn Real talk @nikgeneburn: - will you increase loot? - marked room loot is very disappointing - boss ammo/gear is low tier

The amount of crap I get for putting out a hobby project for free is quite something. People treat this like a multi-million dollar business. Security researchers demanding a bounty. Heck, I can barely buy a Mac Mini from the Sponsors. It's supposed to inspire people. And I'm glad it does. And yes, most non-techies should not install this. It's not finished, I know about the sharp edges. Heck, it's not even 3 months old. And despite rumors otherwise, I sometimes sleep.

OpenAI🤝Cerebras openai.com/index/cerebras…

@anything lfg!

Introducing Anything Max: Vibe Coding that's leaps above Lovable and Bolt We've raised money at a $100M valuation and built what we believe is the future of vibe coding. We asked 100 vibe coders to build their apps side by side on Lovable, Bolt, and Anything Max and they rated Anything Max the winner across all 3 categories - accuracy, design, and 'overall'. Here's why: • Full-stack control: Max can test backend hooks, branch database states, and debug issues, because Anything owns the full infrastructure. • Max can load up your app in its own browser and click on all buttons like a human tester to find all edge case bugs, then trace the bug across the stack - could be a frontend, backend, or a database issue (only we can do this, read #1) and autonomously fix it with 97% accuracy. Lovable and Bolt build prototypes, but Max users are building production-ready apps and already charging money for them. Blake built a gut biome app to $10K run rate Anthony built a referral tool to $20k in revenue Yuri built a suite of apps doing $40K Build your app with Max: createanything.com/max -------------------------------------------- We're hosting a $100K Hackathon to help people grow their app to $10K MRR. - We'll teach you everything we know about growing to 1M users. - You'll have 30 days to build a real product in public and get paying customers for it. If you do it well, you can start the New Year with a functioning business. Retweet and comment “LFG”, and we’ll send you a $100 discount code and the link to participate

@tarkov you broke your AI again. Latest update made AI PMCs and Scav bosses unbeatable 1-tap aimbots again. Please roll it back to the 1.0 launch AI!

Doomscrolling is an attempt to binge-watch the current season of our shared reality. You're tapping into the instinct that if you just keep reading, you'll find out how the story goes! That's not how reality works, though, and no amount of scrolling will speed it up.

@nikgeneburn awesome, thanks!

adding more servers in to cut matching times

@nikgeneburn 2 of my friends took today off work to play- we've been trying to wipe our PvE accounts since 7am, and it is 3pm now

oh bro, yes it is. the load increased overall: website - 10x auth center - 4x launcher - 4x yes, its on new server infrastructure prepared for release, but still - overloaded. yes, this is not an excuse - we are working to fix that. 6 years old flashbacks

@nikgeneburn Absolutely, yes- myself and whole PvE friend group is going to wipe.

A question for PvE players. A reminder - you will need to wipe your PvE profile to have an option to complete the game in PvE mode. But this self wipe is not a mandatory - you can continue to progress on current PvE profile, but you will not be able to complete the game and receive ending and so on. So the question for PvE players - will you wipe you acc on release to start all over and Escape from Tarkov?

I've been using @tan_stack Start for a new project and it's super good. The server functions completely replace the need for TRPC/GraphQL/REST, the middleware is composable and fully typed, and having TSRouter's nice typing and stateful search params is icing on the cake. A+!

@nikgeneburn Not rewarding: the thrill of spotting the boss, engaging them and actually winning the firefight… all to be disappointed when you go to loot them. That is not rewarding.

on release, boss chances will be around 75% and their loot/gear will be nerfed for the sake of early game. Then their typical loot will be slowly returned back. Ok?

🤠