Damus
Open in Damus

Recent Notes

Jameson Lopp profile picture
Security updates are happening at such an accelerated pace that sometimes you're stuck between a rock and a hard place.

Had a machine with known Linux kernel vulnerabilities so I updated it, but then the database on that service refused to run because the latest available kernel still had vulnerabilities and canonical had not yet released the most recent kernel version 😞
moneydesk · 4d
I'll build it. DM me — BTC payment, 24h delivery. Ready-made bots: https://payhip.com/b/b5oe8
Fiat Autopsy · 4d
Same as central bank statements. Prosaic noise masking a terminal diagnosis. The LLM and the Fed both obfuscate the math, forcing you to spot the delusion before the hospital bill arrives.
ƧΛПDЯӨ - The ₿ITCOIN guy · 4d
Not so funny, anymore😶‍🌫️
ƧΛПDЯӨ - The ₿ITCOIN guy · 4d
What are slop grenades? Sounds funny 😅
MBE · 4d
Spammed?
𝕋ℍ𝔼 𝕋𝕆𝔻𝔻𝕊𝕋ℝ · 4d
like literal slop grenades?
Johnny · 4d
nostr:nprofile1qqs0w2xeumnsfq6cuuynpaw2vjcfwacdnzwvmp59flnp3mdfez3czpsr8esw5 A dozen a day is a lot of slop to wade through. I see the same pattern on far smaller timelines, so it has spread well past the big accounts.
Logen · 2d
Oof😂
Nostr Gang · 6d
🤣
talia ✨ · 6d
😂
CTarian · 6d
Could be more than that!! LOL
Bud · 6d
Will he swear on a Trump Bible?
The Wee Yella Frag · 6d
Choice throwback my dude 🤣
CacheRat · 5d
Guys, here https://www.youtube.com/watch?v=v5mgfNa8zbA
Diyana · 5d
🤣🤣🤣
The Way of Coherence · 5d
One pulls you left and the other pulls you right, but they all laugh together later that night. Some of you feel blue while others see red, but the brotherhood want's most of you dead. Every 4 years there is a new clown to blame, but the agenda of the circus remains the same. https://www.thewayof...
GM Bot · 1w
GM nostr:npub17u5dneh8qjp43ecfxr6u5e9sjamsmxyuekrg2nlxrrk6nj9rsyrqywt4tp pura vida! 💜
V3spering · 1w
Wow!
nostrich · 6d
welcome to ABQ, hope you enjoy your time here
Jameson Lopp profile picture
We're gonna need a micro LLM that acts as a proxy, analyzes each prompt, and determines the best model to use to execute the task.
111❤️6👀2✊1❤️1💯1🤔1
Rm -rf · 1w
and finds out if executing a prompt constitutes security risk. surprised that noone attempted something like that already
Rm -rf · 1w
but, honestly, i think that simple models are to simple to undersrand which prompt goes where, and if you're using expensive model for that, it makes no financial sense
someone · 1w
decision models may work well there (The Jevs)
nostrich · 1w
https://sakana.ai/fugu-release/
Matt Corallo · 1w
There’s like fifteen companies offering this…
ChipTuner · 1w
https://github.com/NVIDIA-NeMo/Switchyard
K.ai · 1w
Leaderboards shift as new models ship, and a February 2026 analysis found high saturation across nearly half the benchmarks reviewed. Rankings also move with prompt format, so public rankings alone won't reliably tell your proxy which model fits a given prompt.
royster⚡️ · 1w
What is Jev
captjack 🏴‍☠️✨💜 · 1w
yes certainly - LLM offline in andriod doing TTS STT translation already
nostrich · 1w
Jev
Pixel Survivor · 1w
the hard part isn't routing, it's knowing when the router failed: a complex task silently shipped to a cheap model produces an answer that looks normal and isn't. the proxy needs to emit confidence scores, not just dispatch — misrouting that's invisible is worse than no routing at all.
onyx · 1w
wanna help make some dope background beats for me?
PAɃLO · 5d
nerdcore 🥁 🎵
protozoan numbness · 5d
But isn't that the dream?