Opus 5.5 in medium, that is the magic level.
Weekend hack: HackerTube. I was annoyed with YouTube's recommendations so I had Claude filter down 43M comments over 20 years of HN to 6,454 videos that keep being recommended. *Significantly* better viewing experience, ~2 hours with Opus 5.5:
hackertube.org/
- Excellent recs without any logins (state is kept in local browser session)
- Thumbnail is a representative frame from middle of the video instead of the super duper A/B test hyper-optimized user uploads
- Underlying data from the comment analysis as one zip file, point your Claude at it and ask for your own version
- HackerTube Longs ✊
A bit more on the process:
I use Claude Code (desktop app for Mac), Opus 5.5 on Medium effort as my default. I only talk to Claude and have it do everything, the only "manual work" I helped with in tandem was in the browser (reminiscent of MenuGen): 1) I had to create some kind of API key from Google Cloud UI to query YouTube for video metadata, 2) I booted up a VPS box for serving and 3) I bought the domain name. Once the box was up I gave Claude the IP and told it update, secure, harden, deploy and optimize the code, and it did the whole thing.
Many awesome projects that would have normally taken weeks/months even can now be few-shotted for breakfast. What else are we not thinking of.
It would be great to have an .agentignore file. That could be then enforced by Harness or future OSes.
In the morning in Germany, people are burning wood in their houses, to an extent that it smells like bay area wildfire days.
TIL you do not need to find an ATM to get cash. You can get it from the supermarket cashier. What an upgrade.
I thought biocompute is when the compute runs on biological systems. This seems more like video model with a good PR story.
today, we introduce a little mischief to the world.
for four years, we’ve been brewing real brain cells on electrodes to understand how they compute. now, we’ve turned those discoveries into a video model anyone can use.
from 0.5¢ a second, mischief 1 is at least 5x faster than its base and the cheapest image-to-video model on @fal and @OpenRouter. soon to be on @awscloud marketplace.
it’s not a delusion anymore. biocompute is real, commercial and this is just the beginning.
happy halloween.
Jules was my first hosted AI product, that would just prototype away in a GitHub. And launch was only 2 years ago, crazy. theverge.com/tech/1022729/go…
Die Welt entwickelt sich manchmal rückwärts.
🚨 Ab dem 1. November 2026 wird für Online-Bestellungen aus Ländern außerhalb der EU, die direkt an Privatpersonen in der EU geliefert werden, eine Bearbeitungsgebühr in Höhe von 2 € je Warenposition erhoben.
➡️ Weitere Informationen: presseportal.de/blaulicht/pm…
It has started, some of the PsiQuantum technology, used for other commercial endeavours.

ALT Today, PsiQuantum is unveiling its Optical Circuit Switch initiative with an 8×8 OCS demonstration vehicle built on the low-loss silicon photonics platform—the same platform we’ve spent nearly a decade developing for utility-scale quantum computing. In lab testing, the demo achieved industry-leading results with typical insertion loss of 1.2 dB, 40 dB cross talk, and reconfiguration in under 3 ms. We’re sharing samples with select prospective customers and advancing the platform toward commercial products. https://lnkd.in/dT_aTw82
Now is the time for the hospitality industry to get customers back from booking.com. A booking API that my agent can search through. Low fees, industry controlled. @HOTREC_EUROPE
It is funny that now a days, we do know the question, and we do know the answer, but we don't have the knowledge to move from question to answer. #42
Ready to listen.
as o turns 25
and while still alive
let's play
tell us where
on damienrice.com
photos by robbie fry
video edit by bruna braga
Reflecting a bit on scientific progress today, with AI being a standard tool in science. What crazy Nobel prizes will we see in 2036 and 2040s.
This is such an important message. Not to downplay what the labs have achieved, this is amazing, and progress beyond my dreams even five years ago. The physics nobel prize this year reflects this well, it takes a lot of time and commitment to realize an idea IRL. Exciting future.
I don't know enough about the topic. Is there a big remaining knowledge gap how to train on 100000 GPUds over how to train on 4000 GPUs? Because in-House know-how building is all that counts, at this point of the game.














