MiniMax M2 Technical Report Notes
Technical notes on MiniMax M2, including full attention, fine-grained MoE routing, agent training data, speed rewards, and self-evolution.
Latest AI/ML News
35859 matching items
China's Robotics Dream Began in 1972
a history!
Koog 1.0 Is Out: Stable Core, Better Interop, and Multiplatform Observability
Last week at the KotlinConf 2026 keynote (watch the recording here), we announced Koog 1.0. Koog is JetBrains’ open-source framework for building AI agents in Kotlin and Java. It provides the core building blocks for agentic applications: tools, workflows, persistence, memory, observability, and integrations with existing JVM and Kotlin Multiplatform projects. We introduced Koog at […]
Last Week in AI #341 - Musk loses to OpenAI, Google's IO updates, OpenAI solves Erdős
Elon Musk Loses $150 Billion Suit Against OpenAI and Sam Altman, Google updates its Gemini app to take on ChatGPT and Claude at IO 2026, and more!
definition and example of coords concept
What is coords at csr_matrix?. i confused when i print() my csr_matrix data. something that appear on my terminal is these: Coords Values (0, 1) 3 (1, 0) 4 what the mean of that number tuples on below "coords"?
The Pope’s Encyclical on AI Was Important—Now Comes the Hard Part
The post The Pope’s Encyclical on AI Was Important—Now Comes the Hard Part appeared first on Data & Society .
Place-Based Pathways to a Viable Future
The post Place-Based Pathways to a Viable Future appeared first on Data & Society .
Role-Based Access Control for humans and agents
Introducing Role-Based Access Control for humans and agents, now available for all users on Teams and Enterprise plans.
Using LLMs to Secure Source Code
Build a threat model, discover vulnerabilities, verify, triage, and patch.
AI Weekly Issue #496: Anthropic's Pentagon model is now everyone's model
Anthropic released Mythos to the public, collapsing the wall between cleared-contractor frontier AI and developer-grade frontier AI in a single press release. DeepMind's Demis Hassabis moved his AGI timeline from "five to ten years" to "a real possibility by 2029" and tied it explicitly to AlphaProof Nexus solving nine open Erdős problems for the cost of a steak dinner. Critical zero-days hit Starlette (a million AI agents on the wire) and CrowdStrike led a coordinated takedown of the Glassworm developer botnet across four C2 channels. BNP Paribas formalized a sovereign-AI security partnership with Mistral while Beijing froze overseas travel for top AI engineers at Alibaba and DeepSeek. And the AI-displaces-workforce arithmetic got honest: Uber burned its full-year AI token budget by April, ClickUp restructured to 1,000 humans alongside 3,000 internal agents, and Sam Altman publicly reversed his white-collar-apocalypse prediction.
When AI Makes You Worse at Your Job
The post When AI Makes You Worse at Your Job appeared first on Data & Society .
The Pope Takes On AI
As the technology's reputation wobbles, the pope joins a chorus of AI watchers calling out its potential harms and working with the industry to mitigate them.
Choosing to Stay Human
If you go to your favorite social media site, you will find it full of posts that start to look suspiciously similar to each other:
PyCaret 3.4.0 and scikit-learn return different results
I am currently working with PyCaret 3.4.0, since 4.0 lacks some configuration parameters that are useful for my case. I tried to replicate PyCaret results using scikit-learn. This is my script, after running Pycaret's setup and obtaining the transformed data: compare_models(include=['rf'],cross_validation=True) scoring=['accuracy','precision','recall','f1','f1_macro','f1_weighted','f1_micro', 'roc_auc'] rfc = RandomForestClassifier(random_state=42, n_jobs=-1) scores = cross_validate(rfc, xtrain_trans, ytrain_trans, scoring=scoring, cv=cv) compare_models return these results: Model Accuracy AUC Recall Prec. F1 Kappa MCC TT (Sec) rf Random Forest Classifier 0.7164 0.7617 0.7164 0.7254 0.7137 0.4329 0.4416 0.25 but I get these results from sklearn: Accuracy=0.6948717948717947 AUC=0.7411665257819103 Recall=0.6908791208791208 Precission=0.7005056185644422 F1=0.6867142420587947 F1_macro=0.6910345427878428 F1_weighted=0.6911863570749788 F1_micro=0.6948717948717947 Just to clarify, I am using the exactly same CV splitter on both PyCaret and sklearn. cv = RepeatedStratifiedKFold(n_splits=5, n_repeats=10, random_state=42) The setup used was: setup(xtrain, target = 'Group', session_id=42, test_data=xtest, #None default imputation_type=None, remove_multicollinearity=True, multicollinearity_threshold=0.70, remove_outliers=True, transformation=True, transformation_method='quantile', normalize=True, feature_selection=False, fold_strategy=cv, use_gpu=True) I know the differences are small.…
Google DeepMind CEO Loves Hard Questions 🙂
Full video: https://youtu.be/huAwz_BR8WM #shorts
Reconciling the Past: Correcting Records for Unfixed Kubernetes CVEs
The Kubernetes project relies on transparency to empower cluster administrators and security researchers. One important way we do that is by publishing CVE records into the Common Vulnerabilities and Exposures database. As part of our ongoing effort to mature the official Kubernetes CVE Feed , we have identified some discrepancies. CVE records for a few older, unfixed issues incorrectly include a fixed version field. The Kubernetes Security Response Committee (SRC) will correct the affected CVE records on June 1, 2026. This may result in vulnerability scanners identifying these vulnerabilities in places where they were previously not detected. To help reduce confusion, this post provides a technical update on three vulnerabilities that were disclosed in previous years but remain unfixed: CVE-2020-8561 , CVE-2020-8562 , and CVE-2021-25740 . Why we are updating these records now While these vulnerabilities have been public for several years, the recent work to generate official Open Source Vulnerabilities (OSV) files revealed that their corresponding CVE records did not accurately reflect their status. Specifically, some records suggested a fixed version existed, when in reality, these issues are architectural design trade-offs that cannot be fully remediated through code without breaking fundamental Kubernetes functionality. Correcting these records is vital for the community for: Automation Fidelity : Modern vulnerability scanners depend on precise version ranges. Inaccurate…
Some ideas for what comes next, May 2026
Gemini Flash 3.5, Mythos, open-closed balance, America's open-source surge, emerging power struggles and more.
Is SaaS dead?
MCP comeback in works
Import AI 458: Reckoning with the future; and a singularity story
What AI-driven miracles will happen this year?
WarTalk: Ukraine War Tactical Update
with Rob Lee
Vacancy: Project Manager
Position title: Project ManagerReports to: Research DirectorLocation: Cape Town, South Africa (Hybrid)Duration: Full-time contract Overview Research ICT Africa is seeking a highly organised and proactive Project Manager to support the […] The post Vacancy: Project Manager appeared first on Research ICT Africa .
How Four Teams Stopped Postponing the Refactoring They Knew They Needed
As an engineering leader, you don’t need to be told your codebase needs attention. The issue isn’t awareness – it’s the rational risk calculation that follows. For four teams, that calculation kept producing the same answer: defer. They found a way out not by avoiding the calculation, but by changing what went into it. To […]
LWiAI Podcast #246 - Gemini 3.5 + Omni, Musk Loses, OpenAI vs Erdős
Google unveils AI model Gemini 3.5 and AI agent Gemini Spark, Omni turns images, audio, and text into video, Musk loses OpenAI court battle
Fake Academic Journals Are Publishing AI-Generated Papers Under Real Professors’ Names
The post Fake Academic Journals Are Publishing AI-Generated Papers Under Real Professors’ Names appeared first on Data & Society .
AI Job Losses Are Increasing. Are Training Programs the Answer?
The post AI Job Losses Are Increasing. Are Training Programs the Answer? appeared first on Data & Society .
Women Are Leading the Rebellion Against AI Data Centers
The post Women Are Leading the Rebellion Against AI Data Centers appeared first on Data & Society .
[Registrations closed] Qiskit Global Summer School 2026
Sign up for Qiskit Global Summer School 2026: A decade on the cloud — a free, virtual program for learning quantum computing with Qiskit.
HELM Arabic Enterprise
We present HELM Arabic Enterprise, a leaderboard for transparent, reproducible evaluation of large language models on Arabic-language benchmarks designed around enterprise use cases. The leaderboard was developed in collaboration with Arabic.AI and builds on the HELM evaluation methodology: standardized prompting, fully logged requests and responses, and reproducible scoring through the open-source HELM framework.
Demis Hassabis On What AI Will Do Next
Thank you to Google DeepMind for the invite. 🙏 ❤️ Check out Lambda here and sign up for their GPU Cloud: https://lambda.ai/papers Our Patreon if you wish to support us: https://www.patreon.com/TwoMinutePapers 🙏 We would like to thank our generous Patreon supporters who make Two Minute Papers possible: Adam Bridges, Benji Rabhan, B Shang, Cameron Navor, Charles Ian Norman Venn, Christian Ahlin, Eric T, Fred R, Gordon Child, Juan Benet, Michael Tedder, Owen Skarpness, Richard Sundvall, Ryan Stankye, Shawn Becker, Steef, Taras Bobrovytsky, Tazaur Sagenclaw, Tybie Fitzhugh, Ueli Gallizzi My research: https://cg.tuwien.ac.at/~zsolnai/ Thumbnail design: https://felicia.hu 00:00 Intro 00:40 Gemini Health Scans and Gemma 4 01:30 AI as a Brainstorming Partner 02:30 Second Order Nobel 03:15 DeepMind Co-Scientist 05:00 Curing All Diseases 06:30 Exponential Growth in Drug Discovery 07:45 Regulatory Bottlenecks 09:45 Accelerating Clinical Trials 11:15 EVE Online Partnership 13:15 The Einstein Test 15:30 Recursive Self-Improvement 18:15 Lightning Round 19:30 The Badge of Honor 20:10 Behind the Scenes
Mohamad Moosavi: Accelerating the search for climate solutions with AI
Mohamad Moosavi, Assistant Professor, Chemical Engineering, University of Toronto | Vector Institute Faculty Member The path to breakthrough climate technologies often moves at a frustrating pace. Consider metal-organic frameworks – […] The post Mohamad Moosavi: Accelerating the search for climate solutions with AI appeared first on Vector Institute for Artificial Intelligence .
Annual Report 2024 | GovAI Blog
GovAI's Annual Report 2024.
Annual Report 2025 | GovAI Blog
GovAI's Annual Report 2025.
📈 Why AI bills rise as costs fall
Boo 👻 it's the ghost token
Team Sakura, including Adam Nohejl of the Explainable AI Team, won the Open Track of the BEA 2026 Shared Task（May, 2026)
Team Sakura, including Adam Nohejl, Postdoctoral Researcher, and Hitomi Yanaka, Team Director of the Explainable AI Team, won the Open Track of the BEA 2026 Shared Task “Vocabulary Difficulty Prediction for English Learners,” held in conjunction w
AI Weekly Issue #495: Musk, Zuckerberg killed Trump's AI safety order in three phone calls
Over the weekend: Musk, Zuckerberg, and Sacks killed Trump's draft AI safety executive order in three Wednesday-night phone calls. Anthropic closed a $30B+ round the same Saturday — while Microsoft quietly cancelled its internal Claude Code pilot after token billing ate the entire annual AI budget, redirecting developers to Copilot. CISA logged 15,000 attacks on a same-week Drupal SQL flaw. The first cross-registry supply chain attack — TrapDoor — hit npm, PyPI, and Crates.io at once, using .cursorrules and CLAUDE.md config files as the carrier. And the White House personally overrode the Pentagon to keep Claude inside the NSA.
45 papers have been accepted at ICML 2026
45 papers have been accepted at the International Conference on Machine Learning (ICML) 2026, a major conference on Artificial Intelligence (July 6-11, 2026, Seoul, South Korea). For more details, please refer to the link below. <stro
🔮 Exponential View #575: AI’s math breakthrough and its creative limits
“Humans were not put on this earth to maintain Excel models.”
DeepSeek Sparse Attention From Scratch
Short note on a DeepSeek Sparse Attention from-scratch implementation added to the LLMs-from-scratch repository.
👀 The AI backlash is the only thing growing faster than AI revenues
Humans are angry
Advice on Dataset Choice for Two-Way Sign Language App in Flutter
I am developing a Flutter app called Talk to Deaf , which aims to enable real-time two-way communication between deaf and hearing users. The app will allow normal users to input text or voice and the deaf user will respond in sign language, while the app will convert those signs back into text or speech. I am unsure about which type of dataset to use for training my machine learning model: a dataset with individual alphabets (A-Z) or a dataset with complete words/phrases. I want to ensure accurate and smooth communication. Which type of dataset would be more suitable for building a robust real-time sign language interpreter, and what are the trade-offs of each approach? Any guidance on dataset selection or best practices for training a model for this type of two-way communication app would be highly appreciated.
AI Dev 26 x SF | Ara Khan: Evals Are Broken Use Them Anyway
This talk by Cline's Ara Khan explains why they went from "evals are useless" to using them as a core part of my agent improvement loop. I share practical heuristics for interpreting, running, and creating evals, and why doing them anyway is better than pure "vibes".
Did Google’s AI agents really build an operating system for $916?
The importance of independent evaluation
Semantic Search Starts With Embeddings
“Budget” and “financials” are different words, but embeddings understand they’re related. That’s the foundation behind semantic search and one of the core building blocks of modern multimodal systems. Learn how embeddings power retrieval across text, audio, images, and video in Building Multimodal Data Pipelines: https://hubs.la/Q04hJ9w10
Nuclear Ethics
This "Values & Interests" panel discussion, held in partnership with PBS and moderated by acclaimed journalist Ann Curry, is available to view in full.
How AI Changes the Role of Applied Scientists
Levi Boxell, Tilman Drerup, Alexandr Lenk The Economics Team at Instacart is an applied science team that operates at the intersection of machine learning engineering and economics. Similar to other applied science teams, our work involves a good chunk of engineering, steeped in statistics, math, theory, and strategy. And while that is still at the heart of what we do today, the surprisingly rapid emergence of artificial intelligence has also fundamentally altered our work in ways that we did not see coming. With this post, we want to provide a brief check-in and share an analysis of the patterns we are seeing from a distinctly economic perspective. To do so, we analyze the empirical dynamics of our project portfolio between 2023 and today, looking at the evolution of both the nature and quantity of our work over time. To start, let’s have a quick refresher of what economists at Instacart do and provide a theoretical framework to think about the impact of technological change through AI. Background & Theoretical Framework At Instacart, economists spend their day-to-day on a diverse portfolio of tasks and activities. Similar to other applied science teams within the company, our work relies on a blend of skills, including economics, statistics, math, machine learning, data manipulation, coding, and AI. Due to this versatility in tasks, the team’s work provides a particularly rich testing ground for predictions derived from economic theories concerning the impact of technologi…
AI Dev 26 x SF | Andi Partovi: Why Every Agent Needs a Simulation Sandbox
AI agents fail in unpredictable ways that traditional testing can't catch — hallucinations, wrong tool calls, policy violations, and more. Teams only discover these failures after users hit them in production. A simulation sandbox gives you a controlled environment with realistic users, tools, and workflows where you can run hundreds of scenarios against your agent before it ships, catching edge cases and adversarial inputs that would be impossible to test manually. This talk by Veris AI's Andi Partovi covers why simulation-driven development is becoming essential infrastructure for any team building production AI agents, and how it closes the gap between "works in demos" and "works at scale."
AI Dev 26 x SF | João Moura: Building Recurring, Governed, and Embedded Enterprise Workflows
Modern enterprises don't struggle to experiment with AI — they struggle to operationalize it reliably. In this talk, CrewAI's CEO outlines how leading organizations are moving beyond one-off automations to build recurring, governed, and deeply embedded workflows that drive real business outcomes. Drawing on lessons from production deployments, João explores how to design systems that are auditable, scalable, and aligned with enterprise controls — without sacrificing speed.
AI Dev 26 x SF | Luke Kim: The Agent Data Stack—Why Every AI Agent Needs Its Own Data Stack
From centralized to distributed: In the old world, organizations relied on one centralized data and AI platform. In the new world of AI agents, every agent needs its own sandboxed, secure, and modern data stack. In this 20-minute talk with live demo by Spice AI's Luke Kim, he explores why this architectural shift is critical and the key patterns required to give agents reliable, real-time data.