NotebookLM updates soon We are expecting Gemini 3.5 Flash and Gemini Omni upgrades, along with a bunch of new features. Which ones do you want the most?
LLMS
-
WWDC 2026: Apple to use Google’s 1.2T-parameter Gemini for Siri
By
–
WWDC 2026 – Apple rents Google's brain to fix Siri. What we can expect: Apple's keynote today is a software reset built around one admission: it couldn't build frontier AI alone. What to expect: – Gemini-powered Siri, a rebuilt assistant on a custom 1.2T-parameter Google model,
-
Hivemind launches continual learning for AI coding agents
By
–
New from Hivemind: continual learning for AI coding agents, available to everyone starting today.
— Chubby♨️ (@kimmonismus) 8 juin 2026
It takes the traces from every agent your team runs (Claude Code, Codex, Cursor, Hermes, Pi) and turns them into reusable skills, then pushes those skills across all of them, all on… https://t.co/MFS1ZpCh4DNew from Hivemind: continual learning for AI coding agents, available to everyone starting today. It takes the traces from every agent your team runs (Claude Code, Codex, Cursor, Hermes, Pi) and turns them into reusable skills, then pushes those skills across all of them, all on
-
LangSmith Engine offers three resolution actions
By
–
For each problem it detects, LangSmith Engine offers three resolution actions. Opens a PR
Writes a targeted code or prompt change + opens against the repository. You can review and merge. Creates a custom inline evaluator
Proposes an evaluator -
AI inference is a latency problem, not throughput
By
–
Why does this matter for AI inference specifically? Training = throughput problem. Inference = latency problem. When a user talks to an AI assistant, tokens have to return fast. Latency, memory access, bandwidth, and interconnect all matter, not just raw compute. In large AI
-
Coding agent retry loop: 10,000 calls huge bill
By
–
A coding agent gets stuck in a retry loop during the night. By morning, it has made 10,000 LLM calls. You now have a four-digit bill. Observability tells you what happened, but stopping these problems before they
-
From Prompt Optimization to Agent Loop with Memory
By
–
Peter put it well. The next stage is not "writing better prompts," but designing a loop that can continuously drive an agent. Let the agent observe, remember, retrieve, act, and improve with each run. But the loop needs memory. Without persistent, retrievable, and auditable memory, the so-called agent loop will eventually fall back to "cramming more context into the prompt."
-
A free AI model trained for $7,800 beats a model 400x larger
By
–
Breaking: A free AI model trained for $7,800 has just outperformed a model 400 times larger in mathematics competitions. It is compact enough to run on a laptop. Weibo's AI lab published the results and put all the
-
Low 20T token budget and bug halt DSV4 pre-training
By
–
Could it be linked to the (very) low 20T token pre-training budget? DSV4 was trained on 33T tokens. More room for knowledge in the case of HLE. Looks like they had to stop it early due to a bug they never root-caused. (Not a great demo for NVFP4 pre-training to be honest.)
-
Nemotron 3 Ultra unable to recover HLE and code performance via OPD
By
–
Nemotron 3 Ultra can't recover perf on HLE, code, etc. via OPD The teacher was trained on DeepSeek-V4-Pro traces (DSV4 Max achieves 37.7% on HLE!). Looks like the MOPD warmup failed to properly init the student? No good trajectory → No improvement via OPD