love the bm25-first call over vector dbs. most teams jump to vectors before measuring anything

I like this direction.

Agents having direct access to credentials always felt a bit scary.

This seems cleaner, even if it just moves the trust somewhere else.

Yup! I think the terminologies we're going to be seeing more and more of are "credential exfiltration" and conversely "credential brokering" as a solution to that.

the image generation genuinely got me excited, outputs are impressive. but editing what it creates still feels underbaked.

Wild that privacy became a feature and not the default. Building in this space too and the no uploads needed angle is surprisingly hard to communicate to users who've been trained to expect everything to live in the cloud.

I think the majority of users want their data in the cloud. Edit on desktop, send link to work to client, client looks at it on their phone. No need to export. Change something, client sees it immediately.

Even if it's not business. Sister shares video she edited of party. Mom points something out, sister makes edit.

I'm not saying there's no use to local only solutions. I'm only saying that I suspect for most users, the benefits of in the cloud are something they want.

Yes and yet perfectly logical when the most successful business models are about selling private data.

Man, I feel this. I'm literally grinding to get the first 30 active members for my own project right now and the chicken-and-egg phase is brutal.

Honestly, just be the courier. Pick one route, find the senders manually, and drive the packages yourself.

the cal newport quote in the post is doing a lot of work. the strain required to craft a clear memo or report is the mental equivalent of a gym workout.

fine, but the gym analogy breaks down somewhere. in a gym, the person who actually lifts heavier gets noticed. in software, the person with the right bio and the right network gets noticed, regardless of whether they've ever lifted anything real.

you can spend three years learning compilers properly and have a handful of readers. someone else ships a wrapper on a saturday and lands a pmarca quote tweet by monday.

coding the old way is good for you. i'm not convinced it's what gets you noticed. the strain was never really what got rewarded in the first place.

I agree with you sentiment although I disagree with the sentence

> coding the old way is good for you. i'm not convinced it's what gets you noticed.

You won’t go completely unnoticed if you’re good at your job but you can only be noticed from your deliverables, which is a slow process. You can buff up your presence by talking about it a lot, yes, but you won’t get 0 attention for hard work.

the real unlock here isn't the design output. it's that non-designers can now have a visual conversation instead of writing a 3 page spec that nobody reads.

half the design problems I've seen weren't bad taste. they were bad communication between the person who knew what they wanted and the person who could build it.

love this. one person, 119GB, two drives, rsync. no kubernetes, no distributed cluster, no nonsense. just works.

this is the kind of setup that lets you actually go to bed without checking your phone every 20 minutes.

not a bad strategy. one repo that shows you can duct tape commodity hardware together and let an agent handle the workflow beats 50 leetcode solutions on your profile.

the interesting part is the cost ratio. commercial flying probe setups run 5 figures with proprietary software. this is a weekend build that gets you most of the way there for under 500 bucks.

the "stable long-term companion" framing from jwr's comment is the part that sticks with me. every company I've worked with that chased the new shiny ended up spending more time on migrations than on the actual product.

stability is boring to talk about but it's the thing that actually lets small teams survive long enough to win.