Brian McKelvey

@mehrab_build @levelsio Wow, this is so fucking gross and abusive, flooding the search index with shit content just to game the system. Shame! SEO like this is why we can’t have nice things. 🤮

Unpopular opinion: "AI makes everyone a developer" is true the same way "cameras makes everyone a photographer"

- Drafted a blog post - Used an LLM to meticulously improve the argument over 4 hours. - Wow, feeling great, it’s so convincing! - Fun idea let’s ask it to argue the opposite. - LLM demolishes the entire argument and convinces me that the opposite is in fact true. - lol The LLMs may elicit an opinion when asked but are extremely competent in arguing almost any direction. This is actually super useful as a tool for forming your own opinions, just make sure to ask different directions and be careful with the sycophancy.

@jasonbosco Agree completely! This has also been my experience.

I've had this laundry list of TODOs sitting in my backlog, that I've always wanted to get to, but couldn't convince myself of the value vs effort involved. Thanks to coding agents, that list is now finally shrinking in size. This actually feels... really good! Still no one-shotting here. I still review every line of code and in some cases, I've had to punt on the idea because the changes involved were too invasive... but at least the coding agent helped do the research for me and helped me clarify my thinking. And that was still worth it.

Expectation: the age of the IDE is over Reality: we’re going to need a bigger IDE (imo). It just looks very different because humans now move upwards and program at a higher level - the basic unit of interest is not one file but one agent. It’s still programming.

My autoresearch labs got wiped out in the oauth outage. Have to think through failovers. Intelligence brownouts will be interesting - the planet losing IQ points when frontier AI stutters.

@jasonbosco 😅🫠

“We traced the root cause of the issue that caused the downtime, and turns out it’s all ’s fault. They generated bad quality code that was subtly wrong, and other agents did not catch it either, and it was pushed to production. We’ve updated our prompts to fix the situation.” - Root Cause Analysis in 2028.

@jasonbosco Yikes!

Never imagined I'd see this statement in an AWS status page

@jasonbosco @anumeta10 @grok Post is now unavailable sadly

Why @obsdmd is 100% user-supported and not backed by VC investors: 1. We want to stay small, we don't need to hire lots of people 2. We follow strict principles that we do not want compromise 3. Our users are happy to support us, we don't need VC money Obsidian will not exist forever, no app will. However, the files you create in Obsidian are yours, and can hopefully last for generations. VCware is built with a five year horizon, it is not built to live on for decades. Many startup founders raise VC money because they need the upfront capital to build their product, or they see it as a shortcut to growth. For some products the capital truly is necessary, but too often it's fueled by impatience and the inertia of Silicon Valley. In the short term, VCware tends to subsidize pricing to acquire users. It's easier to grow if your product is cheap or free. But this generally comes at the cost of hoarding user data, and locking in customers. Once you're in you can't get out. To keep raising money, VCware must paint an increasingly enormous vision of their future, which becomes impossible to live up to. This leads to increasingly disparate priorities that gradually make the product worse. What starts off as a useful app become burdened with crap. Eventually all VCware must exit. That means being acquired or going public to pay back investors. It's expected that 9 out 10 startups will fail. That's just part of the math in a VC portfolio. The startups that have big exits pay for the ones that fail. It is now possible for tiny teams to make principled software that millions of people use, unburdened by investors. Principled apps that put people in control of their data, their privacy, their wellbeing. These principles can be irrevocably built into the architecture of the app. Principled people have always been able to make principled software. The difference is that now you need far less money and far fewer employees to reach far more customers. That wave is only just beginning. If you have principles and enough patience, being 100% user-supported is by far the most fun way to build.

@jasonbosco And now that everybody is already conditioned to accept the rough edges, quality continues to decline and the way is paved for the new frontier of AI slop software.

@jasonbosco Funny, I feel like we’ve been seeing this kind of shoddy decline in software for the last 10 years, due to, honestly, “agile” methodology (as commonly practiced), and businesses not giving a shit about quality because they know consumers are used to everything having rough edges.

Say you walk into a newly constructed house. Everything looks pristine and perfect. Almost suspiciously good. But as you live in the house, you start noticing small things. A few floor panels creak when you step on them. One stair is a different height than the rest. One light switch works in the opposite direction of all the others. None of these are catastrophic. Over time, you get used to them. But you also change how you behave. You step more carefully because of that one stair. You flip switches twice to make sure you got it right. You stop trusting what should be automatic. --- Then you see your builder on social media. They’re bragging about how they built this entire house in a few weeks - something that used to take a year - thanks to some fancy new technology. “No plumbers. No electricians. No carpenters. Just me and my machine. I tell it what to do, and it does it. I didn’t even have to review what it did under the hood.” --- This post isn’t about building houses.

@jasonbosco Literally, writing code is the hard won skill that I enjoy. It’s literally the only actually fun part of a software engineering job. It is certainly a bit demoralizing to think about the fun part being delegated away.

Say you got a new job as a software engineer, and you’re told that your only work is to review everyone else’s code, all day long. Would you still enjoy the job?

@jasonbosco I guess now we know why engineers are so often given crappy specs… specs are boring for everyone 😜

So. much. to. read. when you work with LLMs. The more you plan ahead, the longer your plan document becomes and it becomes a chore. Almost... almost... makes me want to just write the code myself to save myself from the drudgery of reading the spec a dozen times.

@jasonbosco @meilisearch 😬

Hey @Meilisearch, hi. You published an article on your blog titled: "Algolia vs Typesense vs Meilisearch: Which search engine powers better developer experiences in 2025?" on Oct 29, 2025. And you're now running ads for it on X. Either it is AI-generated slop, or you're being really intellectually dishonest and intentionally misleading users. I really hope it's the former. > Typesense... while primarily designed for single-node deployment... This is false. And hilarious, considering that this is the predicament you've been stuck in for years now, having designed your entire product around a single node architecture, and have tried to argue with me that a single node is sufficient for High Availability. From your X thread on the same topic: > Typesense is... lighter on AI capabilities Again, hilariously false. You support probably half of the AI features Typesense supports at the depth it does, and you've been playing catch up on this front. And you have zero qualms about writing this. SMH. > Typesense: GPL-3.0 (restrictive for commercial products) Lol. What. Stop spreading FUD around OSS licenses. I hope you're not also recommending people stop using Linux, because Oh no, it's GPL licensed. You know what's restrictive? Gating your features behind a non OSS license called "Meilisearch Enterprise Edition license". The classic rug pull. --- Life Pro Tip for you: if you want to win folks over, be honest. Using slimy tactics like this doesn't inspire confidence in you.

@jasonbosco Omg 100% 😂

State of a code base, after you’ve heard “You’re absolutely right” a few times.

@jasonbosco 😁

Happy “my daily scheduled job ran twice this morning due to DST” to all who celebrate

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+!

@aymen_ux @_heyrico 💯💯💯

Option 1 or option 2? Projects from client 🤔

@aarjithn @tannerlinsley TBH I like the ergonomics of the typescript version better. It’s much more explicitly clear what’s happening.

@tannerlinsley What is wrong with decorators though? Feels pretty nice in this example comparing python with ts implementation of writing mcp Python with decorators #tools" target="_blank" rel="nofollow noopener">github.com/jlowin/fastmcp… TS #quickstart" target="_blank" rel="nofollow noopener">github.com/punkpeye/fastm…

I'll continue to die on the hill of "magic directives are bad API interfaces". They're not type-safe, they're not extensible, no native runtime control, they're just strings. ... I can't believe I'm saying this, but I would rather debate @ directives. Pls rewind history.

@iamdavidhill @opencode Fig. 2, and it’s not even close.

help settle a debate which is easiest to parse for finding a file/folder