Shareable cards — Sunday, July 19, 2026

Sunday, July 19, 2026 — top 10 database + AI stories, one card each. Save the image, copy its caption, post.

Phone: tap Share image, pick Save Image in the sheet (LinkedIn's app won't take images from Safari directly), then tap the button again — it opens LinkedIn: start a post, add the photo on top, paste the caption. Desktop: Open → save the image, attach it in the composer, paste the caption.

Hand-drawn sketchnote — AirLLM v3.0 enables 70B+ LLM inference on 4GB GPUs without quantization

Open image

AirLLM v3.0 enables 70B+ LLM inference on 4GB GPUs without quantization

The day's edition as one sketchnote — full stories + sources:
https://offnetchat.com/news/2026-07-19.html?s=26149269af1e40c4#item-26149269af1e40c4

#TechNews #AI

Pinecone Nexus Engine Compiles Business Context for AI Agents

Open image

Pinecone Nexus Engine Compiles Business Context for AI Agents

▸ Nexus structures enterprise data into a reusable layer for AI agent queries.
▸ Ingesting context once reduces token costs and improves accuracy.

Summary + source: https://offnetchat.com/news/2026-07-19/s/2672320c831f908a.html

#AgenticAI #TechNews

Kimi K3 and the 393-point HN signal for open-weight models

Open image

Kimi K3 and the 393-point HN signal for open-weight models

▸ Kimi K3 is generating substantial debate and interest on Hacker News.
▸ The 393-point score signals high technical relevance to the audience.

Summary + source: https://offnetchat.com/news/2026-07-19/s/f297f3c4a259523e.html

#LLM #TechNews

GPT-5.6 closes 30-year gap in convex optimization via prompt

Open image

GPT-5.6 closes 30-year gap in convex optimization via prompt

▸ GPT-5.6 successfully addressed a long-standing convex optimization problem.
▸ The solution was derived directly from prompt interpretation without code generation.

Summary + source: https://offnetchat.com/news/2026-07-19/s/1c98b9c89e38b32b.html

#LLM #TechNews

Fable 5 vs GPT-5.6 Sol: Does /goal prompt help with NP-Hard problems?

Open image

Fable 5 vs GPT-5.6 Sol: Does /goal prompt help with NP-Hard problems?

▸ Tests Fable 5 against GPT-5.6 Sol on NP-Hard problems.
▸ Evaluates if the /goal prompt directive improves outcomes.

Summary + source: https://offnetchat.com/news/2026-07-19/s/737581f31a198380.html

#LLM #TechNews

AirLLM v3.0 enables 70B+ LLM inference on 4GB GPUs without quantization

Open image

AirLLM v3.0 enables 70B+ LLM inference on 4GB GPUs without quantization

▸ AirLLM eliminates the need for quantization to run 70B+ models on consumer-grade 4GB GPUs.
▸ Massive models like DeepSeek-V3 (671B) are now viable on ~12GB VRAM hardware.

Summary + source: https://offnetchat.com/news/2026-07-19/s/26149269af1e40c4.html

#LLM #TechNews

PostgreSQL Meta Commands Boost DBA Efficiency in psql

Open image

PostgreSQL Meta Commands Boost DBA Efficiency in psql

▸ Meta commands operate at the psql client level, distinct from server-side SQL syntax.
▸ They do not require a terminating semicolon, differentiating them from standard queries.

Summary + source: https://offnetchat.com/news/2026-07-19/s/764730b70796e93b.html

#Databases #TechNews

PostgreSQL external_pid_file GUC enables custom PID pathing

Open image

PostgreSQL external_pid_file GUC enables custom PID pathing

▸ Set external_pid_file to point to the path expected by your cluster manager.
▸ Resolves PID file location mismatches without modifying PostgreSQL source or startup scripts.

Summary + source: https://offnetchat.com/news/2026-07-19/s/e5b3d78cc4c82ff5.html

#Databases #TechNews

Moonshine Micro: Speech Recognition and TTS under 500KB

Open image

Moonshine Micro: Speech Recognition and TTS under 500KB

▸ Enables full audio pipeline (STT/TTS) on resource-constrained edge devices
▸ Model size under 500KB eliminates need for large local inference servers

Summary + source: https://offnetchat.com/news/2026-07-19/s/744e466904127176.html

#AI #TechNews

OpenAI introduces AI scorecard measuring ROI via useful work and cost per task

Open image

OpenAI introduces AI scorecard measuring ROI via useful work and cost per task

▸ Shift focus from raw model output to useful work delivered by AI systems.
▸ Track cost per successful task to optimize operational spending.

Summary + source: https://offnetchat.com/news/2026-07-19/s/3ebda52fc80ebe33.html

#LLM #TechNews

Uber's Zone-Failure-Resilient OpenSearch via Odin Isolation

Open image

Uber's Zone-Failure-Resilient OpenSearch via Odin Isolation

▸ Uber uses Odin container orchestration to enforce isolation groups for OpenSearch nodes.
▸ Native shard allocation is combined with custom logic to handle zone outages.

Summary + source: https://offnetchat.com/news/2026-07-19/s/1ee13618510420aa.html

#Databases #TechNews