trying out @flueai today the goal - create (yet another) PR review agent to help distinguish signal from noise in large PRs with larger PRs its helpful to have an agent pre-screen the changes and highlight the import parts so I can pay extra attention to it will report back

the 2 most important problems in the AI era is infrastructure and security all businesses will face these problems in some form or another in the next 5 years if you solve one of these (or both) you'll be set for success

So basically, today PHP got two critical bugs that let anyone run code on your server without even logging in, which is bad, and then the details leaked before the patch was even out, which is worse, and it works on WordPress and MediaWiki and Magento and Joomla and Laravel, and Censys says like 1.5 million servers are running a version with no fix, so the NSA was like "take your system offline," and the FBI was like "take your system offline," and CISA and the UK and CERT-EU were all like "take your system offline," and then the Prime Minister of Japan, the actual Prime Minister, posted at 5:55 in the morning telling 48.8 million people to take their system offline, so everyone took their system offline, like the Dutch cybersecurity agency took its own website offline, and the Dutch parliament went offline, and the Port of Rotterdam went offline, and Finland and France and Denmark started pulling government sites, and Wikipedia went down, and Archive.org went down, and Asana and GOG and Something Awful went down, and the University of Vienna's webmail went down for some reason, and that's just PHP. Because GitLab also dropped a CVSS 10.0, which is the highest score they give out, for a bug that's been there since 2014, so that's every self-hosted GitLab for twelve years, and then LiteLLM 1.107.0 got hijacked and shipped a credential stealer and a remote access trojan in every single wheel and the Docker images, including latest, and you didn't even have to import it, just installing it was enough, and it's the second time this year, and then Railway paused every deployment on the entire platform because someone got into the environment variables, so now everyone has to rotate every secret they've ever stored there, and TiDB Cloud said someone was walking around inside with stolen credentials for basically a month and an extortion group says they took 10.6 terabytes, and Zeabur got breached for the second time in two months and is just shutting down, like the whole company, and Node.js patched remote code execution and a TLS bypass, and CISA says Plex is being actively exploited, and someone took over 300 Tor exit relays to steal SSH passwords, and SafePay ransomware has hit over 600 critical infrastructure orgs, and India's biggest bank filed a data loss incident, and Japan has been in an official state of emergency in cyberspace since October 9, which honestly explains the 5:55 AM post. Oh, and eMudhra, a certificate authority that's issued over 60 million certificates, whose entire job is proving websites are who they say they are, was supposed to be using private keys that are random numbers over a hundred digits long, and one of them was 4. Just 4. And the others were like 66 and 237. Anyway, it's Wednesday.

or just vibe coding in general man

For years, maybe a decade, I've wanted to build a writing tool that just did a few things I couldn't find anywhere else. A system that mimicked how I think when I write. A system that provided clear affordances, features, and places to do the things I wanted to do. A way to play with my writing as I'm writing it. A way to settle into just the right way to say something. It never made sense to make it a 37signals product because it's not commercially viable, nor would it be worth pulling people off their other work to hack on this. Plus, it would have taken months the old way. So this weekend I just made it myself, with Claude's help. It's called Write_On and the short video walks you through it. Essentially it's "alternative control" at the word, sentence, and paragraph level, plus a way to dim stuff back, and stash stuff near by. Not version control, but alternative control. You'll see what I mean when you watch the video below. So here it is. Write_On.

It’s NOT a hot take that we will be reviewing less code soon But today you shouldn’t follow loud voices who have been full sending some of the worst code we’ve ever seen Stick with Ruby or any language for that matter if your team is executing Use AI to elevate your weaknesses

Rust is a good prompt compilation target for the moment, but so is C++. And soon assembler. Then microcode. Myopic to think we're going to stop the agentic drill bit until it reaches computing bedrock.

No AI, no backspace reflections on my ~4 year journey with @supermeme_ai we made over $200k lifetime revenue. based on where you're looking from, this could be a lot or this could be nothing. But the money is lower down the ROI list for me Context - supermeme is an ai meme generator built by three co-founders (@ramsri_goutham @nwbotha and me) out of a whatsapp group. it's a side project for all 3 of us and we don't spend more than a few hours on it every month In no particular order, here are the things i've personally gained or learnt from it 1. personal brand memes was (and hopefully is) my social media identity. I built a following by sharing corporate tamil memes and creating parody sketches. building supermeme made that even more real, the fact it was revenue generating attracted a lot of podcast and speaking opportunities. it also gave me the confidence to start my own youtube channel in tamil where i talk about building microsaas products 2. product dev exposure for me i'm not a developer. beyond to-do lists apps for online courses, i've never shipped code. supermeme was the first place where i could actually dev see work happening in real time. tech conversations between ramsri and nico used to scare me but i'm glad i got exposed to all that. eventually, with claude code and codex help, i actually started shipping things to a real live product app that was being used by thousands of people 3. building a product (vs being a product manager) i was already a pm by the time we started building supermeme. but as a pm, you hardly feel like you're making decisions that matter - its either forced upon you or been shot down by senior engineers/designers. but with supermeme, i was responsible for many decisions we took. it was also not a one way street, many times i had to work hard to convince ramsri and nico (they don't blindly accept ideas) 4. the 'founder' tag the tech world has hierarchies, even if we deny it. you only know about certain whatsapp groups and meetups if you carry a founder tag on linkedin. i was never too comfortable flaunting the founder tag because i was either freelancing or in a full time job but i saw that people who've done less waving their founder status all over the place, so felt better about how i was using it 5. getting attached to an idea and getting detached i was guilty of being too attached to product ideas. even in freshworks, i would passionately (in hindsight, emotionally) campaign for an idea going all out. i was not a pm or dev, so i would be angry and pms and devs for not seeing my idea. with supermeme, i had a lot more control but it still wasn't easy to get things built. i would still get frustrated when my ideas aren't accepted - but over the years, i started getting emotionally detached from an idea or product. i started being objective and i don't know if that helped the product but it certainly helped my mental health 6. software building myths broken after running supermeme with 3 people, no project management tools, minimal budget - i started seeing software development for what it is - rather, what it's not: a complex meeting driven year long activities. i have a fair sense of how quickly things can be done so i can spot the inefficiencies of a product/dev org from outside even if they can't 7. my unshakeable belief on seo seo drives majority of our revenue. if i was more data driven, i could tell you how many clicks we attract but i know it's a lot. as the sole gtm person, i didn't understand SEO at the beginning but we still attracted the right traffic. it taught me that if you just build the right things and use customer language on your website, you'll attract the right audience. seo starts there, not in keyword research - those come later. later, with more nuanced understanding of seo - i was able to a do a lot more, still doing. our goal is to be the default meme generator of the internet - we can do that just with seo. 8. competitor obsession is real but not very useful we would get competitors, some using the exact same language that we did. some would show up higher than our search results. all of this used to upset me. i would try to take counter action but it was fueled by emotions. over the years, i realized that it doesn't matter. there's enough room for everyone and our lives don't depend on this alone, it's okay. anyone can do what they want, we just have to do our thing 9. there's no one way to build or scale saas supermeme isn't a 'successful' saas by global saas standards. our revenue staggers, degrows. we don't convert at a high percentae. sure, if we had played a few cards differently - we could've either made a lot of money or be wildly more successful but we didn't. it was deliberate. i lucked out by having two co-founders who shared the same vision. we knew that we wanted to keep supermeme as a side project while we go on with our lives. we could've raised funds, we didn't. we could've sold it off, we didn't. we're okay with keeping it and keeping it at this scale. 10. team matters i am grateful for both nico and ramsri, beyond their amazing engineering skills - they're just good people. no drama. high trust. willing to put in the intellectual rigor when needed. this has made the entire experience stress free. i can't say that about other teams i've been part of, i've always walked away more exhausted even if we've built something good 11. an idea doesn't have to take up your whole life traditional startup wisdom is that you have to go 'all in'. supermeme's early days were like that for me. i was properly all in - it was my identity. slowly, it wasn't anymore. that's okay. I got interested in other things, strated a family, a new job, picked up hobbies and supermeme's relevancy in my life reduced. that's okay. i can't imagine doing this as a vc-funded founder. this made me realize that i can still run this without it being an obsession, maybe in the future that obsession will come back and we'll do more. for now, i have the privilege of being a bootstrapped founder who can afford to take time off the project besides this, there are so many other micro learnings and regret, but these are the broad themes. the experience was humbling in many places, we've gotten abused over email, weren't able to afford tools and products that we needed and got harsh feedback. i'm just proud and happy that we're afloat as a business, we pay our taxes. we take a decent sum out of it every month and it feels good that it's coming from something that 3 of us built from scratch. of course, lot more to do and we'll do it. at our own pace, at our own liking, in our own ways. but for now, i just wanna acknowledge how far we've come.

ANTHROPIC BUILDS BIOLOGY LAB TO PUSH AI INTO DRUG DEVELOPMENT Anthropic has quietly established a Bay Area wet lab, moving its life-sciences ambitions beyond computer simulations and into physical biological experiments. The company wants Claude to direct laboratory robots with limited human intervention, potentially accelerating research into rare and traditionally “undruggable” diseases. Anthropic says it will focus on preclinical research rather than clinical trials, avoiding direct competition with pharmaceutical companies.

We're adding support for AGENTS.md to Claude Code. Starting today in version 2.1.277, if there is no CLAUDE.md in a folder, Claude will check for and use AGENTS.md. You can toggle this behavior in /config.

Today we introduce TIN: a powerful and reliable full-text search extension for Postgres. TIN works with complicated WHERE clauses, replication, backups, and maintains correct transaction visibility. It's also mind-blowingly fast.

The p(doom) discussion gets all the headlines, but the p(abundance) scenery is far more like and deserves much more engagement.