fp.
newest
Affiliates Flock to 'Soulless' Scam Gambling Machine
https://krebsonsecurity.com/2025/08/affiliates-flock-to-soulless-scam-gambling-machine/
1
•
todsacerdoti
•
1m ago
•
0 comments
'Isn't Designed to Solve Privacy Concerns,' Grafana CTO on Bring Your Own Cloud
https://www.theregister.com/2025/08/28/grafanas_tom_wilkie_interview/
1
•
rntn
•
2m ago
•
0 comments
Show HN: A high-level search agent
https://www.gensee.ai/tooling.html
2
•
bobby_zhu
•
3m ago
•
0 comments
Uncertain<T>
https://nshipster.com/uncertainty/
1
•
samtheprogram
•
3m ago
•
0 comments
The Toad Report #1
https://willmcgugan.github.io/toad-report-1/
1
•
ingve
•
6m ago
•
0 comments
Dating like it's 1999. No algorithm. No swipes. No signup
https://1999date.com/
2
•
midzer
•
6m ago
•
1 comments
How TRM Labs Scaled Security with Self-Improving AI Vulnerability Agents
https://www.trmlabs.com/resources/blog/scaling-security-in-the-age-of-ai-how-trm-labs-built-self-...
2
•
Frituurpanda
•
6m ago
•
0 comments
GPT-Realtime in the API [video]
https://www.youtube.com/watch?v=nfBbmtMJhX0
1
•
swyx
•
6m ago
•
0 comments
The personhood trap: How AI fakes human personality
https://arstechnica.com/information-technology/2025/08/the-personhood-trap-how-ai-fakes-human-per...
3
•
rbanffy
•
7m ago
•
0 comments
Why stdout is faster than stderr? (2024)
https://blog.orhun.dev/stdout-vs-stderr/
1
•
vinhnx
•
7m ago
•
0 comments
I Ran GPT-5 Against My MacBook. The Laptop Won
https://twitter.com/aviflombaum/status/1961107076731568261
1
•
aviflombaum
•
7m ago
•
1 comments
How to improve your mental health and scroll social media at the same time
https://chadnauseam.substack.com/p/the-simple-way-to-improve-your-mental
1
•
paulpauper
•
8m ago
•
0 comments
In Search of AI Psychosis
https://www.astralcodexten.com/p/in-search-of-ai-psychosis
2
•
paulpauper
•
8m ago
•
0 comments
Building human-centered technology in the AI age
https://www.useinvent.com/blog/building-human-centered-technology-in-the-ai-age
1
•
agnamihira
•
10m ago
•
0 comments
Think Different about WordPress
http://scripting.com/2025/08/28/140604.html?title=thinkDifferentAboutWordpress
1
•
ingve
•
11m ago
•
0 comments
We cut API latency 5x after implementing OpenTelemetry
https://toolhouse.ai/blog/how-we-cut-api-latency-5x-with-opentelemetry-(after-fighting-with-herok...
1
•
orliesaurus
•
12m ago
•
0 comments
CompactifAI Inference API
1
•
Compactifai
•
14m ago
•
0 comments
Microsoft announces new in-house AI models: MAI-Voice-1 and MAI-1-preview
https://microsoft.ai/news/two-new-in-house-models/
5
•
meetpateltech
•
15m ago
•
2 comments
Use of Face Authentication Systems Associated with Japan "My Number Card"
https://www.nec.com/en/global/techrep/journal/g18/n02/180207.html
1
•
rawgabbit
•
16m ago
•
0 comments
Measuring LLM citations with server logs might not work as assumed
https://agentberlin.ai/blog/how-llms-crawl-the-web-and-cache-content
1
•
boatbuilder
•
17m ago
•
1 comments
Ask HN: Windows 11 Update Fail – Linux Distro Suggestions?
1
•
giardini
•
19m ago
•
0 comments
With recent Falcon 9 milestones, SpaceX vindicates its "dumb" approach to reuse
https://arstechnica.com/space/2025/08/with-recent-falcon-9-milestones-spacex-vindicates-its-dumb-...
1
•
LorenDB
•
19m ago
•
0 comments
Show HN: No-Cloud Needed – A License for Devices That Work Without the Cloud
https://j89.net/nocloudneeded/
1
•
Jocund
•
20m ago
•
0 comments
Back to the Future and the Year 2038 Problem: Keeping Embedded Systems on Track
https://www.windriver.com/blog/Back-to-the-Future-and-the-Year-2038-Problem-Keeping-Embedded-Syst...
1
•
mooreds
•
21m ago
•
0 comments
Tic-tac-toe meets Lean 4
https://ochagavia.nl/blog/tic-tac-toe-meets-lean-4/
3
•
todsacerdoti
•
23m ago
•
0 comments
GPT-realtime and Realtime API updates
https://openai.com/index/introducing-gpt-realtime/
7
•
meetpateltech
•
24m ago
•
0 comments
Real-world dataset creation, SFT fine-tuning, and GRPO alignment pipeline
https://github.com/jacobwarren/social-media-ai-engineering-etl
1
•
jwarren92
•
25m ago
•
1 comments
Why tf aren't there voice RPGs yet when LLMs are dirt cheap now?
3
•
jhongodto
•
27m ago
•
1 comments
Markdown Is a Disaster: Why and What to Do Instead
https://karl-voit.at/2025/08/17/Markdown-disaster/
1
•
speckx
•
27m ago
•
2 comments
Trading Bot on Bybit Exchange: Greedy Recursive Arbitrage
2
•
vladimir_spirin
•
27m ago
•
1 comments
Load More
Open in hackernews
ArtSci MagiCalc – Computer Ads from the Past
https://computeradsfromthepast.substack.com/p/artsci-magicalc
1
•
rbanffy
•
2h ago