Opus 4.7 with Apple Health is wildly fun. So far it’s guessed my social activity from noise levels. It noticed an asymmetry in my walk during a certain week (had knee pain). It’s asking why I stopped listening to music a certain month. Way more data in there than I realized.
DATA
-
AI Bottlenecks: Models, Data, or Infrastructure Challenge
By
–
From the @aiDotEngineer London conference floor, we asked a simple question… what’s holding your AI back right now: models or data or infra?
— Snorkel AI (@SnorkelAI) 16 avril 2026
Here’s what @GoogleDeepMind's Cassidy Hardin had to say. pic.twitter.com/KL4IdQSBBHFrom the @aiDotEngineer London conference floor, we asked a simple question… what’s holding your AI back right now: models or data or infra? Here’s what @GoogleDeepMind
's Cassidy Hardin had to say. -
New MedVersa Paper on AI Generalist for Medical Images Published
By
–
I think the adaptive thinking requirement in Claude Opus 4.7 is bad in the ways that all AI effort routers are bad, but magnified by the fact that there is no manual override like in ChatGPT. It regularly decides that non-math/code stuff is "low effort" & produces worse results.
-
What is Data Science: Overview and Applications
By
–
What is #DataScience
by @PythonPr #BigData #DataScientist -
Startup Shuts Down: New Marketplace For Old Company Data
By
–
a company called @simple_closure that helps startups wind down is launching "asset hub", which is a marketplace to sell all the old slack/email data apparently they are working on hardening the PII removal tech for this
-
Predictive Bearing Maintenance Using Leading Indicators
By
–
Dashboard showing machine running hot is lagging indicator. Prescriptive bearing replacement during scheduled Tuesday window is velocity indicator.
-
Three-layer AI filtering transforms industrial alert management
By
–
Three-layer filtering turns 100,000 daily alerts into 35-40 actions. Anomaly detection identifies physics-based deviations. LLM-enabled AI refines to production-critical actions. Expert validation ensures recommendations are executable.
— Lucian Fogoros (@fogoros) 16 avril 2026
Sponsored by Infinit Uptime. #Infinite_iiot pic.twitter.com/zrZS9G1UCeThree-layer filtering turns 100,000 daily alerts into 35-40 actions. Anomaly detection identifies physics-based deviations. LLM-enabled AI refines to production-critical actions. Expert validation ensures recommendations are executable.
Sponsored by Infinit Uptime. #Infinite_iiot -
MERRIN: Multimodal Evidence Retrieval Benchmark for Web Environments
By
–
MERRIN A Benchmark for Multimodal Evidence Retrieval and Reasoning in Noisy Web Environments paper: https://
huggingface.co/papers/2604.13
418
… -
MRCR Evaluation Method Replaced by GraphWalks Approach
By
–
MRCR is a bad eval, we use GraphWalks now. More info here:
-
AI Transforming Healthcare Through Comprehensive Patient Data
By
–
Beautifully written piece by @FAbnousi about how AI for health might look like in the future
The current data in health is limited because it only captures episodic clinical snapshots of what happens to our bodies
The revelation is that there is so much latent knowledge in