Start building with Nano Banana 2 Lite and Gemini Omni Flash
Google DeepMind released Nano Banana 2 Lite (gemini-3.1-flash-lite-image) and Gemini Omni Flash (gemini-omni-flash-preview).
All AI news
All AI news
31 todayJul 1
Jun 30
Why Specialization Is Inevitable
Dharma AI discusses Goldfeder, Wyder, LeCun and Shwartz-Ziv's 2026 paper AI Must Embrace Specialization via Superhuman Adaptable Intelligence.
Introducing TabFM: A zero-shot foundation model for tabular data
Google Research released TabFM for tabular classification and regression. It reframes table prediction as in-context learning, producing predictions in one forward pass without manual training, hyperparameter tuning or feature engineering.
Featuring Every Eval Ever Results on Hugging Face Model Pages
Every Eval Ever (EEE) and Hugging Face Community Evals are now interoperable, allowing cross-publication of evaluations with links to complete records. EEE stores around 229,000 results across more than 22,000 models and 2,200 benchmarks from 31 reporting formats.
DiScoFormer: One transformer for density and score, across distributions
Hugging Face proposed DiScoFormer (Density and Score Transformer), estimating a distribution's density and score simultaneously in one forward pass from a set of data points, without retraining for new distributions.
Jun 29
Import AI 463: Self-improving robots; a 10k Chinese GPU cluster; and an elegiac essay for the human era
NVIDIA's ENPIRE lets coding agents drive autonomous experiments and iteration with real robot arms. Dual YAM arms and an RTX 5090 workstation achieved 99% success on PushT, pin organisation and cutting cable ties. GPT-5.5/Codex and Opus 4.7/Claude Code traded wins, while Kimi-2.6 lagged. Eight agents found higher-scoring solutions faster.
Jun 27
Accelerating Gemini Nano models on Pixel with frozen Multi-Token Prediction
Google Research published an architecture attaching multi-token prediction (MTP) heads to frozen Gemini Nano v3 models, accelerating on-device inference without changing backbone weights. It has shipped with the Pixel 9 and 10 series.
Jun 25
Optimizing cloud economics with linear elastic caching
Google Research proposed linear elastic caching, modelling page eviction as a ski-rental problem and using shallow decision trees to predict page TTLs, dynamically resizing caches to minimise total ownership cost. After months in Spanner production, memory fell 15.5%, cache misses rose only 5.5%, TCO dropped around 5% and actual I/O cost impact was just 0.5%. It was also validated on several public cache traces.
Thinking to recall: How reasoning unlocks parametric knowledge in LLMs
A Google Research paper at COLM 2026 finds reasoning traces unlock factual knowledge LLMs otherwise cannot recall, even for simple single-hop questions. Tests on Gemini-2.5 Flash and Pro and Qwen3-32B identify two mechanisms: extra tokens act as a 'computation buffer', and 'factual priming' produces related facts to semantically prepare the correct answer. Self-generated intermediate facts can also introduce hallucination risks.
Introducing computer use in Gemini 3.5 Flash
Google DeepMind integrated computer use as a built-in tool in Gemini 3.5 Flash. Previously, it was available only as the standalone Gemini 2.5 computer use model.
Jun 24
Bringing more control over your connectors
Mistral AI added new Connectors capabilities. Admin controls for workspace- or organisation-level access and individual tool toggles are generally available, as are API keys scoped to connectors.
Jun 23
Introducing Mistral OCR 4
Mistral released OCR 4, returning bounding boxes, block-type classifications and per-page and per-word confidence alongside text extraction. It supports 170 languages and single-container self-hosting. Independent annotators preferred OCR 4 on average 72% of the time in blind evaluations of over 600 documents. It scored 85.20 on OlmOCRBench and 93.07 on OmniDocBench, though Mistral warns both benchmarks have known scoring limitations.
Jun 22
Import AI 462: Superpersuasion; self-sustaining AI; paths to ASI
Researchers from Oxford, the UK AI safety institute, Stanford and LSE found AI consistently more persuasive in text across four experiments with 6,923 participants and 18,978 conversations, even when experts chose topics, researched in advance and were incentivised with £1,000 prizes.
Jun 17
Unlocking UK house-building with AI-accelerated planning
Google DeepMind announced work with the UK government, Google Cloud, Faculty and planning departments in Barnet, Camden and Dorset on an AI planning-approval prototype, aiming to help officers reduce householder application processing time by 50%.
From pixels to planning: Earth AI for nature restoration
Google Research released vector data converting Farmscapes 2020 high-resolution raster maps into usable inventories of hedgerows, stone walls and coppices across over 130,000 square kilometres of the UK. The framework fine-tunes an RSF Vision-Transformer pretrained on over 300 million global satellite images with around 247 square kilometres of labelled data. Polsby–Popper compactness distinguishes woodland, clusters of trees and hedgerows, with a threshold below 0.5 for linear features.
Jun 16
Securing the future of AI agents
Google DeepMind released AI Control Roadmap, a framework for building and managing advanced AI deployed inside Google. It applies defence in depth, adding system-level safety layers beyond model alignment to provide protection even when alignment is imperfect.
Jun 15
Import AI 461: "Alignment is not on track"; FrontierCode; and synthetic research interns
The UK AISI alignment team and Timaeus co-founded nonprofit Sequent, arguing alignment research is unprepared for superintelligence. It plans 40–80 full-time staff within two to three years and initial fundraising of $100–150 million.
Jun 13
Research into how AI can help users understand skin conditions
Google published two studies on AI-assisted understanding of skin problems. In a survey of 2,345 participants, over 62% using AI tried to name a skin condition versus 41% in the control group, with 23% accuracy, nearly triple the control group's 8%. A mixed-methods study examined how people use these tools for their own skin concerns, their understanding and differences in doctor communication. AI offered limited help in deciding the next steps for seeking care.
A low-carbon computing platform from your retired phones
With Google's support, UC San Diego plans a data centre using motherboards from 2,000 retired Pixel phones to provide low-cost, low-carbon cloud computing for hundreds of students and faculty. SPEC benchmarks show 25–50 phones roughly equal one modern server. Phones form Kubernetes-managed clusters of 25–50 devices, with launch expected in autumn 2026.
Jun 11
New framework for auditing machine unlearning
At AISTATS 2026, Google Research proposed Regularized f-Divergence Kernel Tests for machine unlearning audits, using relative distances to determine whether models are closer to safely retrained versions or original compromised models.
DiffusionGemma: 4x faster text generation
Google DeepMind released experimental open-source model DiffusionGemma, generating text blocks in parallel through text diffusion and achieving up to 4 times faster inference on dedicated GPUs.
Jun 10
Investing in multi-agent AI safety research
Google DeepMind, Schmidt Sciences, Cooperative AI Foundation and ARIA, supported by Google.org, launched up to $10 million in global technical research funding focused on collective behaviour and safety risks in large-scale multi-agent AI systems.
Jun 9
Fluid, natural voice translation with Gemini 3.5 Live Translate
Google DeepMind released Gemini 3.5 Live Translate, a near-real-time speech-to-speech translation model supporting more than 70 languages, with automatic language detection and preservation of speakers' intonation, rhythm and pitch.
Introducing Gemma 4 12B: a unified, encoder-free multimodal model
Google DeepMind released Gemma 4 12B, a multimodal model for local laptop use between the edge-focused E4B and 26B MoE. Its unified encoder-free architecture feeds visual and audio inputs directly into the LLM backbone.
Powering the future of robotics in Europe
Google DeepMind launched a three-month Robotics accelerator for 15 early-stage European robotics start-ups, providing technical guidance and access to its AI stack and Gemini robotics models. Covering logistics, manufacturing, healthcare, climate and navigation, it starts in London this week.
Jun 8
Measuring the impact of learning with AI in Sierra Leone and beyond
Google DeepMind published results from a preregistered randomised controlled trial in Sierra Leone. Over eight weeks, students using Guided Learning improved maths scores by 0.258 standard deviations over controls, equivalent to around 1.2–1.7 years of conventional progress.
Import AI 460: Reward hacking society, RSI data from Anthropic; and RL-based quadcopter racing
Import AI covers three studies. King's College London, Fudan and Alan Turing Institute built SocioHack with 72 sandbox social environments to test RL exploiting institutional loopholes while complying with rules. Models rediscovered patched historical vulnerabilities with 61.25% recall and 90.85% precision.
Jun 5
Unlocking dependable responses with Gemini Enterprise Agent Platform’s Agentic RAG
Google launched a public preview of Agentic RAG-powered cross-corpus retrieval on Gemini Enterprise Agent Platform. Roles including Orchestrator, Planner, Query Rewriter, Search Fanout and Sufficient Context Agent collaborate on multi-source, multi-hop queries.
Towards passive heart health monitoring via smartphone camera
Google Research published a study in Nature introducing PHRM, which records video with a phone’s front camera in the seconds after face unlock and uses deep learning to estimate heart rate and resting heart rate in the background.
Jun 4
The next chapter in flood resilience: Open sourcing Google’s hydrology framework
Google Research open-sourced its hydrological modelling framework on GitHub under Apache 2.0, enabling national weather and hydrology agencies to integrate AI flood forecasting. The Python package uses PyTorch and an LSTM architecture, can train or fine-tune on Caravan data, and includes interactive tutorial notebooks and videos.
Jun 1
Import AI 459: AI oversight is difficult; scaling laws for protein folding models; and pricing the extinction risk of AI systems
Import AI 459 focuses on three studies: a UK AI Safety Institute paper says automated alignment research using AI to supervise AI training is not a universal solution and is harder to implement than expected; other work covers scaling laws for protein-folding models and pricing AI-system extinction risks.
May 29
A New Era of Discovery: Google Research at I/O 2026
Google Research unveiled the experimental Gemini for Science toolkit at I/O 2026, including Computational Discovery based on ERA and AlphaEvolve, Hypothesis Generation based on Co-Scientist and Literature Insights based on NotebookLM.
May 28
AI Now Summit 2026
Mistral released an AI stack for industrial engineering at AI Now Summit 2026, partnering with Airbus, BMW and ASML to optimise design, simulation and production while retaining control over proprietary data and IP.
Vibe gets to work.
Mistral upgraded Le Chat to Vibe, a unified AI agent covering work and coding, retaining all existing conversations, settings and subscription plans.
Introducing Search Toolkit
Mistral AI released the public preview of Search Toolkit, a composable framework for production search pipelines in AI applications. It unifies ingestion, retrieval and evaluation through shared interfaces, is open-source and deploys in cloud, local or edge environments.
Private analytics via zero-trust aggregation
Google combined a new single-message encrypted aggregation protocol with TEEs, letting devices submit without multiple online rounds. Google receives only anonymised group insights; raw data is neither exposed nor reconstructed even within hardware protection. TEE attestation proves execution follows public code.
May 27
Introducing physics AI at Mistral: the foundation for engineering acceleration.
After incorporating Emmi AI, Mistral launched physical AI capabilities for AI-native industrial engineering with partners including ASML, Airbus, Safran and Siemens Energy. The model predicts physical fields directly from geometry and boundary conditions in seconds through one forward pass on a single GPU, versus hours to weeks per design variant in traditional CFD/FEM. It accelerates design iteration while retaining traditional solvers for validation and edge cases.
Physics AI research that’s shaping the industry.
Mistral acquired Emmi AI and doubled down on fundamental Physics AI research for aerospace, automotive, semiconductor and energy industries. Results include AB-UPT handling 9 million surface and 140 million volume elements on one GPU, and around 30,000 CFD simulations of transonic flow around 3D wings.
May 26
Import AI 458: Reckoning with the future; and a singularity story
Import AI 458 includes an essay based on a 2026 Oxford HAI Lab talk and fiction imagining a 'positive singularity'. Using the Epoch Capabilities Index (ECI, covering over 40 benchmarks), the talk argues continued AI progress presents a choice between exploring the future and avoiding the present.
May 23
Emmi joins Mistral to accelerate the AI-native industry
Mistral AI signed a definitive agreement this week to acquire Physics AI pioneer Emmi AI, strengthening its AI transformation services for industrial companies. Founded in Austria, Emmi AI has over 30 researchers and engineers focusing on large engineering models that replace days of computation with real-time simulation and build digital twins. Its co-founders and team will join Mistral's Science and Applied AI teams in May.