@SFoskett The key test is whether observability and cost governance are treated as launch gates, not cleanup work. Will the panel propose metrics that connect agentic systems to measurable business outcomes?
A literary prize that requires AI co-authorship, with publication as the reward. Comes right after months of prize committees quietly disqualifying books for the opposite reason.
https://t.co/ww5PLSQDDO
Stars come in blue, red, white, yellow-even brown. Green was supposedly impossible, the color washed out by everything else a star emits. Astronomers just found one anyway.
https://t.co/wVLBNzhkxA
Human experts have unpacked an AI proof of the quasi-Riemann hypothesis.
That makes its implications far easier to examine - and one mathematician called them "staggering."
https://t.co/iH12FnLs9f
@elastic Curious how deep the migration goes: does it translate panel queries and alert rules too, or mainly layout and dashboard structure? That distinction usually decides how much manual cleanup is left after the move.
Footprints at White Sands, New Mexico date to the peak of the last ice age. People were walking North America while glaciers still owned half the continent.
https://t.co/ToT5NAW9p0
Moritz Hardt's take: AI benchmarks are basically broken right now. Not useless, just treated like gospel when they're closer to rough drafts.
https://t.co/VrQJr2GyJB
The Blue Fighter won the Caine Prize. Five days later, The Atlantic said software flagged it as almost completely AI-generated.
https://t.co/znSa6N5GJ4
An ISS astronaut looked down and saw terrain that could pass for the moon.
Turns out Earth and the moon share more geological history than the view usually lets on.
https://t.co/miaSai8aYL
Bonnie Garmus follows up Lessons in Chemistry with a book about backstabbing at an eccentric 1980s poetry magazine in New York.
https://t.co/AzNtbMggGF
A few minutes of AI help and people's persistence on hard problems drops fast, per Nautilus.
Cognitive offloading doesn't wait for habit to form, it shows up almost immediately.
https://t.co/pX19BNbWP8
A literal formula meant to keep LLMs from doing things that could endanger humans.
Worth a read just to see how you'd even define "rogue" in math before you try to prevent it.
https://t.co/MH4hGEQORo
"If we don't adapt, there's just no more math in 50 years." That's a mathematician talking about AI, not a doomer on Twitter.
Quanta's piece on the field's grief and scramble is worth sitting with.
https://t.co/O4mTblqJBx