DeepMind skips memory safeguards, Astra costs 2.5×, IBM leads Apache forecasters
Every URL the pipeline pulled into ranking for this issue — primary sources plus the supporting and contradicting findings each Researcher returned. Inline citations in the issue point back here.
Sources
Recreating a 70-year love story frame by frame blog.google
An elderly couple sitting in a movie theater. Overlayed are “Teulluride Film Festival” and “Love, Rendered”
.blend URL Viewer simonwillison.net
Tool: .blend URL Viewer I’m continuing to have a lot of fun with GPT-6 Astra and Blender (see my TIL ). As a big fan of the Imperial Fabergé Easter eggs , I’ve always thought it would be fun to make some new ones that celebrate popular culture. Yesterday I decided to try out the new ChatGPT Images 2.5 by running this prompt : Generate a photo of a faberge egg that’s themed after the TV show Pluribus - research first It gave me this - honestly not bad for a first attempt! Then, just to see what…
IBM releases SOTA Granite Time Series PatchTST-FM-r2 model with commercial-friendly license huggingface.co
Google’s AI genome system evaluates every possible one-base change arstechnica.com
AlphaGenome runs through every possible one-base substitution in the human genome to flag the rare changes that disrupt splicing or gene regulation. Most single-base edits are silent, so exhaustive scoring helps researchers isolate the few variants that actually matter in non-coding DNA.
Update to Google’s AI weather model improves forecast accuracy arstechnica.com
The updated forecasting system now ingests more raw satellite data instead of pre-processed inputs, mirroring how traditional numerical models improve when fed a broader observation set. The wider input stream lifts forecast accuracy across the model’s predictions.
References
Kotaku on Aronofsky’s Primordial Soup kotaku.com
Darren Aronofsky’s AI Movie Is As Bad As It Sounds: Slop About the Forming of the Country That Is About to Crumble
Newsweek — Aronofsky AI backlash newsweek.com
By using AI to generate ‘everything seen on screen,’ the production avoids hiring thousands of concept artists, animators, and cinematographers, a move seen as a betrayal in a post-SAG/WGA strike environment.
LA Times — Bill Gates at Telluride latimes.com
Festival director Julie Huntsinger described AI as a ‘threat’ and a ‘humanity dilemma’… Gates voiced a mix of ‘wonder and catastrophe,’ comparing the lack of AI control to a science-fiction threat.
MIT Media Lab (Pataranutaporn/Loftus) — AI-amplified false memories media.mit.edu
Generative chatbots induced over three times more immediate false memories than control conditions… approximately 36.4% of participant responses were successfully misled.
Domestic Data Streamers — Citizens’ Office of Synthetic Memories domesticstreamers.com
Images are generated with a ‘dream-like’ or surreal aesthetic, and faces are often obscured to prevent them from being mistaken for objective historical evidence.
Medium — Anatomy of Veo 3 medium.com
Veo utilizes a 3D latent diffusion model… audio signals (spectrograms) are fused directly into the latent space during the generation process… Security is managed through SynthID watermarking.
AI Weekly on Google TimesFM-3 aiweekly.co
Google TimesFM-3 tops forecast benchmarks but weights are restricted under a non-commercial license (v1.0) that prohibits production deployment.
Amazon Science blog — Chronos-2 amazon.science
Chronos-2 introduces a ‘group attention’ mechanism enabling in-context learning across related series and covariates in a single forward pass, achieving the highest win rate among pretrained models on GIFT-Eval.
Grid Dynamics TSFM comparison griddynamics.com
When weather covariates were added, the iTransformer improved accuracy three times more efficiently than PatchTST due to its inherent ability to mix information across variables.
CauKer paper (Paris Descartes preprint) helios2.mi.parisdescartes.fr
Models pretrained exclusively on CauKer synthetic data can match or exceed zero-shot performance of those trained on real-world corpora, with clear scaling laws from 10K to 10M samples.
OpenReview review of CauKer openreview.net
The individual components (GP and SCM) are well-established… evaluation is largely confined to zero-shot classification; further research is needed on downstream fine-tuning.
IBM Research blog — Time series AI for enterprise research.ibm.com
Through the Confluent integration, Flink SQL functions like AI_FORECAST and AI_DETECT_ANOMALIES call Granite models directly on Kafka streams, with Flink’s state managing the historical context window — no external database required.
Mindstudio — GPT-6 Astra benchmarks analysis mindstudio.ai
Artificial Analysis scores Astra at 61 on its Intelligence Index, effectively a tie with GPT-5.6 Sol and trailing Claude Fable 5.1 at 66; Astra is roughly 2.5× more expensive per token than Sol.
Emergent.sh — GPT-6 Astra benchmarks emergent.sh
OpenAI reports Astra hitting 95.9% on BenchCAD 3D-reconstruction and 99.9% on ARC-AGI-3, positioning it as the first model marketed specifically for long-horizon agentic 3D/CAD work.
SJSU ScholarWorks — 3DCodeBench / CodeGen-3D paper scholarworks.sjsu.edu
General-purpose LLMs fail on 21–92% of Blender-script prompts; even when scripts execute, outputs frequently exhibit ‘disconnected geometry’ such as floating chair backrests, and specialised BlenderLLMs outperform frontier models on procedural CAD tasks.
Simon Willison — ‘Code research projects with async coding agents’ (Substack) simonw.substack.com
Skill files are a bigger deal than MCP: they use progressive disclosure so instructions load only when a task matches, versus MCP injecting full tool schemas into every prompt — one study measured ~1,365 tokens per CLI-skill run vs ~44,026 tokens for MCP-based agents.
GitHub issue #201, ahujasid/blender-mcp github.com
Practitioners driving Blender through LLM agents report repeated failure modes — unintuitive stops, focus bugs on macOS, and over-engineered geometry — that require manual intervention despite the ‘autonomous’ framing.
GitHub issue #18294, jsdelivr/jsdelivr github.com
jsDelivr enforces a ~50 MB per-file cap (sometimes as low as 20 MB) on GitHub-sourced assets and returns 403 above that, while raw.githubusercontent.com rate-limits unauthenticated clients around 5,000 req/hour — both bite the moment a browser .blend viewer sees real traffic.
