InThe Medium BlogbyThe Medium Newsletter·5h agoHow Pride is like Independence DayReflecting on the LGBTQ+ community, past and present. (Issue #431)A clap icon178A response icon2
InData Science CollectivebyMandar Karhade, MD. PhD.·22h agoNVIDIA’s Vertical Integration: How Enterprise Should Navigate ItNVIDIA’s GTC Taipei Pitch: AI Factories, Vera CPUs, and the Cost Nobody Wants to Say Out LoudA clap icon127
John Kellden·22h agoPlayⁿ: Between Two MachinesPlayⁿ Fieldcraft as a convivial turning to both capitalist machine and desiring-machine
InData Science CollectivebyMehmet Özel·22h agoSpatial AI: The Missing Layer Between LLMs and the Physical WorldIn this article, I explore how Spatial AI bridges large language models with 3D world models, SLAM and NeRF to turn abstract text…
InData Science CollectivebyGourav Didwania·1d agoStop Doom-Scrolling Financial News — I Built an AI That Does It (and Reviews Itself)Series: AI in Finance — Part 3
InSurviving TomorrowbyJared A. Brock·1d ago6 Signs That Total Surveillance Is Coming FAST5 things you can do to protect yourselfA clap icon186A response icon2
InData Science CollectivebyTanmay Deshpande·1d agoWhy Agent Observability Breaks Every Rule of Application MonitoringYour agent can return 200 on every span and still make the wrong decision. Here’s why your dashboard can’t see it.A clap icon2
InData Science CollectivebySamvardhan Singh·1d agoGoogle and Nvidia Backed SynthID, We Built Its Receiver-Side Complement ..How a copyright detection problem turned into a 15 GB/s zero-copy pipeline across Python, Rust, and CUDA and why the math demanded it…A clap icon650A response icon12
InData Science CollectivebySefa Bilicier·1d agoLLMs : Inside Your Intelligent Digital FriendFrom tokens and embeddings to reasoning, alignment, and multilingual capabilities — everything you need to know about LLMs, I explained…A clap icon23
Fayner Brack·1d agoA PDF Parser from the 80s Beats Claude (And They Use It Internally)My journey to parse a CIA PDF, only to discover picking the right tool for the job is still a thingA clap icon50
John Battelle·1d agoGoogle Eats The Web. Now What?Last month Google announced the most significant change to its search product since its launch in 1998. Its iconic search box, which I’ve…A clap icon6
InData Science CollectivebySantosh Shinde·1d agoInkwell: Designing a Multilingual AI Agent for the BrowserThe end-to-end system design — Chrome surface, agent runtime, AI gateway, guardrails, observability — for a multilingual writing agent…A clap icon4
InData Science CollectivebyAnalyst Uttam·1d agoYour Next Laptop Is a Personal Supercomputer — Nvidia Just Made It RealHow the RTX Spark Superchip ends the era of cloud-only AI and moves a data center’s worth of intelligence into your bagA clap icon21A response icon31
InData Science CollectivebyMehmet Özel·1d agoWhy Most LLM Projects Fail in ProductionMost LLM demos work, but production systems quietly fail on evaluation, retrieval, guardrails, and cost. This article breaks down how…A clap icon1A response icon1
InData Science CollectivebyHayanan·1d agoThe Most Mysterious AI Response Ever Recorded: When Machines Say Things We Cannot ExplainSix unexplained AI outputs that researchers, journalists, and the models’ own creators still cannot fully account for, from Bing’s two-hour…A clap icon335A response icon21
InData Science CollectivebyŞükrü İmre, PhD·2d agoThe Agent’s Memory: Modeling Relationships with Ontology and Knowledge Graph
InData Science CollectivebyTANVEER MUSTAFA·2d agoThe AI Conversation That Made Me Question My Own IntelligenceI asked AI to push back on something I was certain about.A clap icon17
InData Science CollectivebyKalle Georgiev·2d agoCorrelation vs. Causation for Humans: Why Your Data Is Lying to YouAnd why roosters don’t cause sunrisesA clap icon260A response icon21
InData Science CollectivebySaurav Singh·2d agoAzure Data Factory vs Databricks for Orchestration: The Honest Comparison No One WritesBoth tools work. Both have costs nobody talks about. Here are five questions that actually decide which one belongs in your pipeline.A clap icon62A response icon11
InData Science CollectivebyJeff Flynt·2d agoThe Benchmark That Exposed a Silent Failure in AI Agent MemoryEvery major memory benchmark is measuring the wrong thing. Here is how we found out.A clap icon48
InData Science CollectivebyMehmet Özel·2d agoWhy Google’s SynthID Might Become the Missing Layer of AI TrustLearn how Google’s SynthID and C2PA content credentials work together to label, trace, and verify AI‑generated media across Search, Chrome…A clap icon75
InData Science CollectivebyHayanan·2d agoAI Facts That Feel Like Leaked Information From 2035: Seven Real, Documented, Already-Happened…A neural network solved a biology mystery that took humans a decade. A quantum chip ran an algorithm 13,000 times faster than the…A clap icon712
InData Science CollectivebyŞükrü İmre, PhD·3d agoThe Second Step in Building Reliable Agents: The Semantic LayerThis article is the second in a 6-part series explaining agentic AI architecture through the warehouse management use case I designed. In…A clap icon14
InData Science CollectivebyŞükrü İmre, PhD·3d agoThe First Step in Building Reliable Agents: The Data LayerOne of the most striking takeaways from a conference I attended in the UK was this:A clap icon8
Greg Satell·4d ago3 Questions To Ask You Before You Begin A Major TransformationMy friend Jessica Kriegel often warns her clients about the action trap, the urge to do something — anything — when things aren’t going…A clap icon7A response icon11
InData Science CollectivebyDr Barak Or·4d agoCan Predicted Dynamics Exist in the Physical World?Predictive Physical AI systems increasingly produce more than a single action. They output state rollouts, action chunks, latent plans, and…A clap icon2
InData Science CollectivebyYanli Liu·4d agoWhat Makes Claude’s New Dynamic Workflows DifferentHow Claude Code’s latest feature compares to Cursor, Copilot, and Codex — and why reusable orchestration mattersA clap icon330A response icon3
InData Science CollectivebyMehmet Özel·4d agoClaude Opus 4.8 and the Quiet End of the Prompting EraClaude Opus 4.8 may not just be a better model, it may represent the quiet end of the prompting era. This article breaks down its…A clap icon6
InData Science CollectivebyEgor Howell·5d agoThe State of Data Jobs in 2026 — It’s NOT What You ThinkThe market is not bad, you are.A clap icon95A response icon2
InData Science CollectivebyNathan Rosidi·5d agoThe Death of DashboardsThe analytics interface is changing faster than most teams realize.A clap icon48A response icon22
InData Science CollectivebyUmair Ali Khan, Ph.D.·5d agoSenseNova-U1: The Open-Source AI Model That Truly Unifies Vision, Reasoning, and GenerationMost multimodal models are pipelines pretending to be unified. SenseNova-U1 is the proof you can do it for real.A clap icon68
InData Science CollectivebyHan HELOIR YAN, Ph.D. ☕️·5d agoWhat Anthropic Didn’t Say About Opus 4.8: It’s Anthropic Absorbing Your HarnessThe Opus 4.8 is not JUST A MODEL UPDATEA clap icon3KA response icon79
InData Science CollectivebyAnalyst Uttam·5d agoClaude Opus 4.8 Makes One Massive Change That You Can’t IgnoreNot the biggest benchmark jump of the year. The most important alignment shift of the year. Here’s why developers should care.A clap icon88A response icon1
Fayner Brack·5d agoBusiness Success: Luck, Not Merit.Each new founder pays again to learn what the last founder already did. Some of them hit the jackpot.A clap icon20
InData Science CollectivebyFelix Pappe·5d agoFrom Zero to App: How Beginners Can Build with Claude CodeYou Don’t Need to Be a Programmer to Build Apps AnymoreA clap icon60
InData Science CollectivebyAnubhav·5d agoAnthropic Just Dropped Opus 4.8. Is This the End of OpenAI?41 days since Opus 4.7 and the landscape shifts again.A clap icon228A response icon74
InData Science CollectivebyRichard Conway·6d agoFrom Toy Model to Transformer: Upgrading nanoGPT in C# with Attention and EmbeddingsPart 3 of building GPT from scratch in C#. Token embeddings, positional embeddings, multi-head causal attention, layer norm, residuals…
InData Science CollectivebyHan HELOIR YAN, Ph.D. ☕️·6d agoxAI Just Paid $60 Billion for Cursor a Coding Model It Could Have Downloaded for FreeWhat Cursor didn’t tell you about Composer 2.5: the model is the cheapest part of the moatA clap icon567A response icon5
InData Science CollectivebyNahid Ahmadvand·6d agoWhen Your Clustering Sees Two Groups and You Know There Are SixA walk through unsupervised vs. supervised learning on the Samsung HAR dataset and why the gap between geometric structure and semantic…A clap icon50
InData Science CollectivebySebastian Buzdugan·6d agoI tested Claude Code as a Daily Driver: Claude.md,The honeymoon phase lasts about a week. Then Claude starts stepping on itself, reformatting files you did not touch, and confidently…A clap icon18A response icon3
John Battelle·6d agoIs AI Profoundly Anti-Social?All aboard ha ha ha ha ha ha ha AI, AI, AI, AI, AI, AI, AI!A response icon1
![Data Science Collective](data:image/svg+xml,<svg xmlns="http://www.w3.org/2000/svg" width="40" height="40"><rect fill-opacity="0"/></svg>){kind=small}
![Data Science Collective](data:image/svg+xml,<svg xmlns="http://www.w3.org/2000/svg" width="20" height="20"><rect fill-opacity="0"/></svg>){kind=small}
