Most data science jobs are actually problem definition jobs. Before you write a line of code, ask: — What decision is this driving? — What does “good” look like? — What data shouldn’t we use? #DataScience #MachineLearning #AI #RStats #Python

Spyglass MTG Unveils AI Navigator Amid Rising Demand for Enterprise Governance -- radicaldatascience.wordpress.com/2026/01/22/s... #AI @SpyglassMTG #Governance

As I was working on the 2nd edition of my ML book, I kept thinking: The real job of a data scientist isn’t to build models. It’s to help people make better decisions under uncertainty. That hasn’t changed—and probably never will. #DataScience #MachineLearning #AI #RStats

What’s one data tool or concept you rediscovered in 2025 that surprised you with its usefulness? For me: data.table in R. Lightning-fast joins. Elegant logic. #RStats #DataScience #MachineLearning

The Demise of Tech Journalism -- medium.com/@ddgutierrez...

SO much is happening, don't miss any of it: AI News Briefs BULLETIN BOARD for January 2026 -- radicaldatascience.wordpress.com/2026/01/15/a... #AI #GenAI #LLM

NEW UPDATES! AI Industry Predictions for 2026 -- radicaldatascience.wordpress.com/2026/01/14/a... #AI #GenAI #LLM #predictions @imply.bsky.social

UPDATED! Keeping a Pulse on DeepSeek -- radicaldatascience.wordpress.com/2025/06/16/k... #AI #GenAI #LLM #deepseek

The longer I work in data science, the more I value: — Sharp definitions — Clear goals — Modest, testable models Hype fades. Clarity endures. #DataScience #MachineLearning #AI #RStats

BREAKING TODAY! LambdaTest Rebrands to TestMu AI, the Agentic AI platform for Quality Engineering -- radicaldatascience.wordpress.com/2026/01/12/l... #AI #AgenticAI #GenAI #LLM

I always tell my students: The best model isn’t the most accurate. It’s the one you can explain clearly to a decision-maker who wasn’t in the room when it was built. #DataScience #MachineLearning #AI #RStats

DDN Powers Integrated Compute, Data, and Offload at Scale for NVIDIA Rubin Platform -- radicaldatascience.wordpress.com/2026/01/08/d... #AI #hardware #CES26

BREAKING from CES: NVIDIA DGX Spark and DGX Station Power the Latest Open-Source and Frontier Models From the Desktop -- blogs.nvidia.com/blog/dgx-spa... #AI #hardware #GenAI #LLM @NVIDIAAI @nvidia #CES2026

R remains my go-to for exploratory analysis. — Compact syntax — Powerful visualization — Mature statistical packages No language supports statistical reasoning like #RStats. #DataScience #MachineLearning #AI

BREAKING TODAY! AI Industry Predictions for 2026 -- radicaldatascience.wordpress.com/2026/01/05/a... #AI #GenAI #LLM #predictions @imply.bsky.social

In 2026, the differentiator won’t be the LLM—it’ll be: — who owns the context — who integrates memory — who designs with guardrails Next-gen AI is less about size, more about fit. #AI #GenAI #LLM

Keeping a Pulse on DeepSeek! mHC: Manifold-Constrained Hyper-Connections -- radicaldatascience.wordpress.com/2025/06/16/k... #AI #GenAI #LLM #DeepSeek

2025 has been quite a year in AI! Here is a complete list of all the updates. Exciting to say the least -- radicaldatascience.wordpress.com/ai-news-brie... #AI #GenAI #LLM

What a month of December! Don't miss all the updates: AI News Briefs BULLETIN BOARD for December 2025 -- radicaldatascience.wordpress.com/2025/12/22/a... #AI #GenAI #LLM

2025 saw huge advances in AI. But most companies still struggle with: — dirty data — vague goals — misaligned metrics The core challenges aren’t new. They’re just louder. #AI #GenAI #LLM

NEW book! Regression Modeling Strategies - hbiostat.org/rmsc/ #DataScience #Statistics #MachineLearning #AI #Rstats

While working on the 2nd edition of my book, I became even more convinced: Teaching statistics is really about teaching judgment under uncertainty. Formulas are easy. Thinking in distributions is hard. #DataScience #MachineLearning #AI #RStats #Statistics

NEW! Percona Launches Percona Packages to Enable AI-Ready, High-Performance Database Infrastructure -- radicaldatascience.wordpress.com/2025/12/22/p... @percona.bsky.social #database #AI #SQL #PostgreSQL

AI-native startups are no longer focused on training new models. They’re focused on: — context-aware agents — secure memory layers — human-in-the-loop UX The edge isn’t in scale. It’s in orchestration. #AI #LLM #GenAI

BREAKING TODAY! Palo Alto Networks and Google Cloud Forge Landmark Agreement to Help Customers Securely Accelerate Cloud and AI Initiatives -- radicaldatascience.wordpress.com/2025/12/19/p... @paloaltontwks.bsky.social #AI #cloud

BREAKING TODAY! Enlightra emerges from stealth with $15M to build the future of energy-efficient laser links for AI and data centers -- radicaldatascience.wordpress.com/2025/12/18/e... #AI #datacenter

Students often assume that models create insight. I remind them: models compress what you already understand. If you don’t understand the domain, the model won’t rescue you. Assumptions come first. #AI #DataScience #MachineLearning #RStats

BREAKING TODAY! Molmo 2: State-of-the-art video understanding, pointing, and tracking -- radicaldatascience.wordpress.com/2025/12/16/m... #AI #GenAI #video @ai2.bsky.social

What was the first dataset that changed how you saw the world? Not the cleanest. Not the biggest. The most revealing. Mine was a public health dataset—messy, human, urgent. # DataScience #MachineLearning #RStats

DECEMBER has been a very busy month! AI News Briefs BULLETIN BOARD for December 2025 -- radicaldatascience.wordpress.com/2025/12/11/a... #AI #DataScience #MachineLearning #LLM #GenAI

R users: revisit the power of data.table. For large data wrangling jobs, it beats dplyr on speed and syntax economy. It’s a reminder that performance doesn’t always require Python. #DataScience #MachineLearning #AI #RStats

NEW! Couchbase AI Services Put Enterprises in Control of Agentic AI -- radicaldatascience.wordpress.com/2025/12/11/c... @couchbase.bsky.social #AI #database #AgenticAI

The most disruptive AI systems aren’t the biggest—they’re the most grounded. LLMs are impressive. But the future is in: — retrieval-augmented generation — domain-specific tuning — workflow integration #AI #GenAI #LLM

BREAKING TODAY! Axonis Emerges From Stealth With Federated AI Architecture That Brings AI to the Data -- radicaldatascience.wordpress.com/2025/12/10/a... #AI #FederatedAI

BREAKING TODAY! Linux Foundation Announces the Formation of the Agentic AI Foundation (AAIF), Anchored by New Project Contributions Including MCP, goose and AGENTS.md -- radicaldatascience.wordpress.com/2025/12/09/l... #AI #AgenticAI @linuxfoundation.org

In most real-world data projects, you don’t need 97% accuracy. You need: — clarity of assumptions — repeatability — stakeholder buy-in A flawed but explainable model often beats a perfect black box. #DataScience #MachineLearning #AI #RStats

NTT Scientists Contribute Fifteen Research Papers to NeurIPS 2025 -- radicaldatascience.wordpress.com/2025/12/08/n... #AI #LLM #GenAI #NeurIPS @neuripsconf.bsky.social @nttresearch.bsky.social

NEW! Lumia, Agentic AI Security platform, Secures $18M Seed Round, Appoints Former NSA Director to Head the Advisory Board -- radicaldatascience.wordpress.com/2025/12/05/l... #AI #security #governance

BREAKING TODAY! imper.ai Launches With $28 Million in Funding to Pioneer Real-Time Defense Against AI-Driven Impersonation and Social Engineering Cyber Attacks -- radicaldatascience.wordpress.com/2025/12/04/i... #AI #deepfake

While working on the 2nd edition of my book, I’ve realized: The hardest part of data science isn’t the code. It’s helping people understand what matters—Which metric, which trade-off, which insight. Tools change. Judgment persists. #DataScience #MachineLearning #AI #RStats

BREAKING TODAY! Approximate Domain Unlearning: Enabling Safer and More Controllable Vision-Language Models -- radicaldatascience.wordpress.com/2025/12/02/a... #AI #VLM #vision @neuripsconf.bsky.social #NeurIPS #neurips25

And that's a WRAP! A whole month of accelerated compute! AI News Briefs BULLETIN BOARD for November 2025 -- radicaldatascience.wordpress.com/2025/11/21/a... #AI #LLM #GenAI

The gap between “demo” and “deployment” in GenAI is enormous. It’s not the model—it’s: — latency — cost — governance — security LLMs are easy to call. Hard to trust at scale. #AI #LLM #GenAI

BREAKING TODAY! Minitap Raises $4.1M to Make Mobile Development 10x Faster with AI -- radicaldatascience.wordpress.com/2025/12/01/m... #AI #mobile

What's one dataset that taught you more than any textbook ever could? For me: — messy healthcare billing data — multiple date formats — unclear labels It was a masterclass in defining the problem. #DataScience #MachineLearning #AI #RStats

I teach my students: Don’t confuse fitting a model with solving a problem. A good model tells you something useful. A great model tells you something actionable. The rest is just math. #DataScience #MachineLearning #AI #RStats

I’ll say it again: #RStats is still the best language for statistical modeling, especially in applied settings. Why? — Native support for inference — Elegant formulas — Visual diagnostics baked into the workflow #DataScience #MachineLearning

We’re now beyond GenAI as a novelty. The frontier: — Alignment with real tasks — Memory + context integration — Traceability The next wave isn’t smarter LLMs—it’s smarter interfaces. #AI #GenAI #LLM

Classiq and BQP Demonstrate Hybrid Quantum-Classical Breakthrough for Scalable Digital Twin and CFD Simulation with NVIDIA -- radicaldatascience.wordpress.com/2025/11/21/c... #quantum

You don’t need deep learning for every problem. You need: — Clean data — Clear framing — Honest evaluation Simplicity scales better than complexity that no one trusts. #DataScience #MachineLearning #AI #RStats