Apple Photos app corrupts images
tenderlovemaking.com
Optimizing ClickHouse for Intel's 280 core processors
clickhouse.com
Gluon: a GPU programming language based on the same compiler stack as Triton
github.com
DeepMind and OpenAI win gold at ICPC
codeforces.com
Tinycolor supply chain attack post-mortem
sigh.dev
YouTube addresses lower view counts which seem to be caused by ad blockers
9to5google.com
Drought in Iraq reveals tombs created 2,300 years ago
smithsonianmag.com
Launch HN: RunRL (YC X25) – Reinforcement learning as a service
runrl.com
Noise cancelling a fan
chillphysicsenjoyer.substack.com
The Asus gaming laptop ACPI firmware bug
github.com
Ton Roosendaal to step down as Blender chairman and CEO
cgchannel.com
U.S. investors, Trump close in on TikTok deal with China
wsj.com
Event Horizon Labs (YC W24) Is Hiring
ycombinator.com
Famous cognitive psychology experiments that failed to replicate
buttondown.com
Anthropic irks White House with limits on models’ use
semafor.com
DeepSeek writes less secure code for groups China disfavors?
washingtonpost.com
Infinite Mac: Resource Fork Roundtripping
blog.persistent.info
Ask HN: What's a good 3D Printer for sub $1000?
Depression reduces capacity to learn to actively avoid aversive events
eneuro.org
Just for fun: animating a mosaic of 90s GIFs
alexplescan.com
Tau² benchmark: How a prompt rewrite boosted GPT-5-mini by 22%
quesma.com
Alibaba's new AI chip: Key specifications comparable to H20
news.futunn.com
UUIDv47: Store UUIDv7 in DB, emit UUIDv4 outside (SipHash-masked timestamp)
github.com
Hm, I can't see any part that talks about the models Aider used? I've observed "surface level without deeper analysis" issues mostly with the non-CoT models, but I can usually overcome this through use of a CoT model and iterating on prompting.
I wonder what was the exact query and model used - it would be nice to have even just a snippet.