0. GPT 6.1 Sol: Near-Astra intelligence for a fifth of the price (openai.com)
1066 points · 952 comments · by crorella
OpenAI introduced GPT-6.1 Sol, a lower-cost model that approaches GPT-6 Astra’s performance in coding, computer use, professional workflows, and research, with API pricing of $2 per million input tokens and $10 per million output tokens. [src]
1. Everybody’s home. No one’s coming over (derekthompson.org)
833 points · 738 comments · by barry-cotter
Americans’ monthly home hosting rate has plunged from 42% in 1975 to 12% in 2026, reflecting broader declines in face-to-face socializing driven by busier dual-earner households, intensive parenting, shrinking friendship networks, and increasingly convenient solitary entertainment. [src]
2. America.gov (america.gov)
778 points · 740 comments · by plesiv
America.gov is inaccessible because the site returned a 403 error and requires JavaScript and cookies to continue. [src]
3. Dots: Always-on agents (openai.com)
766 points · 646 comments · by alvis
OpenAI introduced “dots,” always-on AI agents powered by GPT-6 Astra that can connect to more than 4,000 apps, work autonomously across channels, learn user preferences, and support organizations through specialized, permission-controlled agents. [src]
4. Ask HN: What are you reading?
467 points · 883 comments · by dan-bailey
The author is reading the entertaining LitRPG series *Dungeon Crawler Carl* and finishing Cory Doctorow’s *Enshittification*, while seeking a more intellectually challenging next read. [src]
5. Livenerf: Has Opus 5.5 been nerfed yet? (github.com)
922 points · 392 comments · by bryan0
The livenerf GitHub project is tracking whether Claude Opus 5.5’s performance changes after launch, but with only six of 30 days collected, it has not yet produced evidence of “nerfing.” [src]
6. DraftKings is using AI to behaviorally target chronic gamblers (eff.org)
570 points · 432 comments · by paimapi
The EFF says DraftKings uses AI trained on customers’ betting records to identify likely losing gamblers and target them with promotions, arguing the practice exploits vulnerable users and underscores the need to ban online behavioral advertising. [src]
7. How Delhi cut electricity loss from 50 to 5 percent (spectrum.ieee.org)
592 points · 336 comments · by rbanffy
Delhi cut electricity losses from over 50 percent in 2002 to 5–6 percent in 2026 by privatizing distribution, modernizing infrastructure, deploying digital monitoring and meters, combating theft, improving billing, and engaging communities, raising grid reliability above 99.9 percent. [src]
8. macOS Golden Gate Is a Buggy Mess (squareorbits.com)
491 points · 360 comments · by SquareOrbits
A user’s review of macOS 27 Golden Gate finds numerous visual inconsistencies, forgotten legacy assets, broken features, crashes, and system hangs, arguing that Apple’s stability-focused release remains notably unpolished despite its promise to refine Tahoe. [src]
9. 500k facial scans at UK stations yield no arrests, 1 false positive (theguardian.com)
518 points · 304 comments · by ilamont
A six-month British Transport Police facial-recognition trial scanned over 500,000 faces at London stations, costing £320,786 and producing one false alert but no arrests directly attributable to the technology. [src]
Brought to you by ALCAZAR. Protect what matters.