pwr1

4 ms·

1.
by pwr1 5mo ago
This looks pretty useful. The versioned filesystem part is nice becuase that’s exactly where a lot of agent stuff gets messy fast.
2.
Cohere Transcribe: Open-source 2B speech recognition model(huggingface.co)
1 points by pwr1 5mo ago|0 comments
3.
by pwr1 5mo ago
Nice execution. How are you handling deduplication when the same asset shows up on multiple sites?
4.
by pwr1 5mo ago
Did this about a year ago, went smoother than expected tbh. the main gotcha for us was DO's managed postgres — had to dump/restore manually since there's no direct migration path to Hetzner's managed DBs. ended up just s
5.
by pwr1 6mo ago
Just lost an hour going through this. Found a Nirvana show from 1989 at Dreamerz. The recording quality is surprisingly decent for a cassette tape. This is exactly the kind of thing the internet was supposed to be for.
6.
by pwr1 6mo ago
So they want protection from harms caused by their own models. Classic move — lobby for the rules while you're still ahead of regulators who don't fully understand the technology yet. Would be interesting to see what happens when
7.
by pwr1 6mo ago
We tried something similar at a previous company — ended up with 3 different bots all answering slightly differently depending on which doc chunk they hit. The consistency problem is real. Curious how you handle updates. Like if someone edi
8.
by pwr1 6mo ago
Solid list. I'd add git log --all --oneline --graph pretty early on — gives you a quick sense of how active different branches are and whether this is a "one person commits everything" project or actually distributed. Helped
9.
by pwr1 6mo ago
Native has its place for sure — camera, sensors, offline, background tasks. But web has come a long way and for a lot of products it's genuinely enough. The issue is teams defaulting to "build an app" without asking whether t
10.
by pwr1 6mo ago
Yep. If your product needs me to install an app for a one-off thing, you've probably already lost me. The crazy part is how many teams still treat the web as the demo and the app as the “real” product. For a lot of stuff it's the
11.
by pwr1 6mo ago
I catch myself doing this more than I'd like to admit. Copy something from an LLM, it works, ship it, move on. Then a week later something breaks and I realize I have no idea what that code actually does! The speed is addicting but you
12.
by pwr1 6mo ago
This resonates. I had a project sitting in my head for years and finally built it in about 6 weeks recently. The AI part wasn't even the hard part honestly, it was finally commiting to actually shipping instead of overthinking the arch
13.
by pwr1 6mo ago
Pieter Levels is the obvious one. Not really a neat polished “blog”, more like watching someone ship in public for years. Josh Pigford / Baremetrics had some good posts too. And the old Indie Hackers interviews were good for this befor
14.
by pwr1 6mo ago
Running 26B locally is impressive but the latency math gets rough once your doing anything beyond chat. We switched from local inference to API calls for image generation specifically because cold start + generation time on consumer hardwar
15.
by pwr1 6mo ago
I've seen conspiracy posts saying things like, "It's upside down"... haha
16.
by pwr1 6mo ago
Took like 5 minutes to create a kinda basic/simplistic animation. Sorry, not impressed.
17.
Daft Punk – One More Time Sample(twitter.com)
2 points by pwr1 6mo ago|0 comments
18.
by pwr1 6mo ago
Shifting to GPT 5.4 feels like my top performing employee just went on vacation and now I have to deal with complete incompetence in comparison.
19.
by pwr1 6mo ago
I can see skills, having been built over time having value that other agents would pay for.
20.
by pwr1 6mo ago
Obviously I do that. Also take appropriate amounts of K2.
21.
by pwr1 6mo ago
I take 10k-50k IU vitamin D3 every day and haven't been sick in years (knock on wood). Not saying that vitamin D3 is protecting me, but just saying...
22.
by pwr1 7mo ago
Ahh, now I see it. So cool! Bookmarked.
23.
by pwr1 7mo ago
Love it, but when I clicked another show in the guide, nothing happened.
24.
by pwr1 7mo ago
Cool... I guess it depends which model you are using for these tasks. Are there certain models that perform better for a certain task, or maybe could be recommended based on quality/cost?
25.
by pwr1 7mo ago
I don't think we've quite hit the "Her" stage yet. Before I fall in love with my AI agent, it has to stop pissing me off.
26.
by pwr1 7mo ago
Digital/virtual graffiti. Just like in real life, you'll see good stuff and bad stuff.
27.
by pwr1 7mo ago
Interesting... added to bookmarks. Could come in handy in the future