AI News Radar

These are my thoughts about the InfiMaker K1 5-axis desktop CNC mill after attending the launch event.

Here's a roundup of this week's must-read articles. We'll look into the latest developments on humanoid robots, simplifying embedded system design, photonic IC development, and more. Also, check News Archives - Embedded and Technical Articles Archives - Embedded for the complete list of news and articles from our website. NEWS OpenLight and Tower Accelerate Photonic IC Development The expanded [.…

Aalto University’s $155 open source 6-DoF stylus makes high-precision spatial tracking accessible to makers and labs everywhere.

In 1998, Swatch tried replacing time zones with "Internet Time" — and one maker just built a clock to keep the failed idea alive.

Researchers harness resonance to build rocket- and helicopter-like flying drones, and small boats, driven solely by sound.

Altera Corp. has expanded memory support across its Agilex FPGA portfolio with Quartus Prime Pro Edition Software version 26.1.1. The release adds DDR5, LPDDR5, and LPDDR5X-compatible memory capabilities to additional Agilex devices, giving developers greater flexibility when optimizing system bandwidth, power consumption, PCB space, and component availability. For Agilex 7 M-Series devices, the …

Researchers created a ribbon-like flat optical fiber that detects pressure with up to 1,000x the sensitivity of round fibers.

Tiny magnets, thousands of times smaller than a grain of sand, could one day do our computation for us.

Photoacoustic sensor delivers a claimed ±30 parts per million plus three percent accuracy over a range of 400–5,000 PPM of CO₂.

Eric Xing, President and Professor at Mohamed bin Zayed University of Artificial Intelligence and Carnegie Mellon University presents “AGI Meets the Real World—Toward Reasoning, Planning and Acting Beyond Book Intelligence” at the May 2026 Embedded Vision Summit. LLMs like OpenAI’s GPTs have fascinated the public with their astounding capabilities on… “AGI Meets the Real World—Toward […] The post…

This blog post was originally published at Image Engineering’s website. It is reprinted here with the permission of Image Engineering. As the automotive industry continues its path toward full automation, one area of focus has become the in-cabin monitoring systems, often referred to as driver and occupant monitoring systems (DMS/OMS). These systems use cameras and sensors […] The post In-Cabin I…

Perforated AI has released an enhanced version of ResNet-18 designed to approach the accuracy of larger vision models without imposing a comparable increase in model size or computational cost. The pretrained model is now available for evaluation and transfer learning on Hugging Face. The company’s technology adds artificial dendrites—additional computational structures inspired by biological neu…

Assessing tire defects is another use case in the market for neuromorphic AI chip solutions LAGUNA HILLS, CA, UNITED STATES, August 13, 2026 /EINPresswire.com/ — A global leader in ultra-low power, fully digital, event-based neuromorphic AI, today announced a partnership with Atlanta-based Orama.AOI, which has retrained and optimized BrainChip’s Akida models using its industrial inspection […] Th…

On September 10, 2026 at 9:00 am PT (12:00 ET), SPIE will present the webinar “SPIE AR, VR, MR monthly Fireside Chat with FlexEnable.” Here’s the description, from the event registration page: Each month, hosts Bernard Kress from Google, Christophe Peroz from Google Zurich, and Grace Lee from Mojo Vision interview key leaders to explore […] The post Upcoming Webinar on AR, VR, and MR from SPIE ap…

Turn your old Wii Remote into a smart home controller with OpenMote, an open source, ESP32-powered drop-in replacement board.

The emulation of the numerous types of movement produced by the human body in humanoid robots is arguably among the most complex challenges at the forefront of modern engineering. Designing robots to tackle everyday tasks such as assembly on a production line or cooking in a kitchen demands the coordination of multiple sub-tasks. The automation [...] The post Humanoid Robots Need a Real-Time, Dis…

Build your own multi-room audio system for under $50 with an ESP32, open source tech, and a custom 3D-printed case.

A 2D plotter delivers 3D holograms, etching a programmatic interference pattern into the surface of old CD jewel cases.

Built as a set of tools atop a Linux base, the Flipper OS stack promises to do away with project-hopping leading to a "trash system."

OpenLight and Tower Semiconductor Ltd. have expanded the design ecosystem for Tower's PH18DA indium phosphide (InP)-on-silicon photonics platform. OpenLight's photonic process design kit (PDK) is now available within Cadence Design Systems' electronic design automation (EDA) tools, enabling developers to design photonic integrated circuits (PICs) using an established IC design environment. The PD…

200MHz RISC-V part includes the optional vector extensions for improved performance in ML and DSP tasks, and up to 8MB of external PSRAM.

Bluetooth's latest ranging feature comes to the mainstream as Google joins Apple in offering first-party location-tracking item finder tags.

James Imber, Director of Research at Imagination Technologies presents “Self-Compression for Edge Inference” at the May 2026 Embedded Vision Summit. Self-compression is a quantization-aware training technique to reduce neural network size and optimize performance for edge inference. By learning optimal bit depths for weights and activations during training, self-compression achieves… “Self-Compre…

This blog post was originally published at ModelCat’s website. It is reprinted here with the permission of ModelCat. Artificial intelligence has advanced rapidly over the past decade, fueled in large part by the availability of pre-trained models and shared research. One of the most visible outcomes of this progress is the rise of the “model zoo”—collections […] The post Why Model Zoos Fall Short…

A maker reverse engineered an automated cotton candy machine to break free from factory restrictions.

Texas Instruments (TI) has claimed the industry's first commercially-available controller area network (CAN) extended data-field length (XL) transceiver for industrial systems including humanoid robots, industrial robots, and HMI systems that require more data at faster speeds. The TCAN6062 CAN XL transceiver supports payloads up to 2,048 bytes per frame and data rates as high as [...] The post T…

The first official ESPHome starter kit launches today for $40, letting you build custom smart home devices without soldering or coding.

Microchip Technology Inc. has introduced Revision 2.0 PolarFire FPGA Ethernet Sensor Bridge for Ethernet-based AI inferencing at the edge. The smaller, production-ready board for standardized sensor integration, leveraging Nvidia Holoscan Sensor Bridge (HSB) technology, supports twice the number of cameras and offers a lower price point than the first generation. Thanks to the power efficiency, s…

Amit Gupta, Associate Director–Solution Architecture and Head of Robotics CoE at eInfochips (an Arrow company) presents “Introduction to Simultaneous Localization and Mapping: From Block Diagrams to Real Systems” at the May 2026 Embedded Vision Summit. Simultaneous localization and mapping (SLAM) is the core capability that allows robots and autonomous systems… “Introduction to Simultaneous Local…

This blog post was originally published at Synopsys’s website. It is reprinted here with the permission of Synopsys. AI is quickly emerging from its digital confines as something new: physical AI. This evolving incarnation pushes beyond the realm of information — code, text, images, video — and enables machines to sense, decide, and act in the […] The post Physical AI: Bridging Silicon, Software,…

Nordic Semiconductor is simplifying the development of IoT devices that require connectivity across both terrestrial cellular and satellite networks. LooUQ's MTC2-N9151 embedded modem has received Skylo non-terrestrial network (NTN) certification, building on the certification of Nordic's nRF9151 cellular IoT module. This enables developers to use a pre-certified hardware foundation for products …

h3>Smaller, multi-camera platform enables scalable Ethernet architectures for NVIDIA Edge AI systems while lowering power, cost and integration complexity CHANDLER, Ariz., August 11, 2026 — The shift toward compact, high‑performance edge AI systems is redefining sensor connectivity, pushing developers to deliver power‑ and space‑efficient architectures that remain secure, scalable and future‑read…

ADI details how an open, modular software ecosystem, using its CodeFusion Studio, can reduce development effort in embedded system design. The post Simplifying Embedded System Design with a Modern Software Ecosystem appeared first on Embedded .

Tom Michiels, System Architect at Synopsys presents “Understanding Transformers: From LLMs to Context-Aware Multimodal Models” at the May 2026 Embedded Vision Summit. Transformers have become the foundation of modern AI, reshaping how products are built and how businesses operate. In this talk Michiels explains why transformers replaced earlier models, what… “Understanding Transformers: From LLMs…

Whisper runs end-to-end on Chimera. Encoder and decoder compiled as native GPNPU kernels, INT4 weights, FP16 attention, top-1 token match against the float32 reference. Scales to four cores with a flag, no recompile. This blog post was originally published at Quadric’s website. It is reprinted here with the permission of Quadric. I ran a […] The post Your NPU Learned to Listen appeared first on E…

Silicon Motion Technology Corp. has introduced its MonTitan SSD reference design kit (RDK), incorporating the company's next-generation PerformaShape technology. Presented at FMS2026, the platform is designed for agentic AI infrastructure, where enterprise SSDs are increasingly required to operate as a persistent memory tier for applications such as key-value (KV) cache offload. The solution targ…

Open commercial licensing, benchmark‑leading reasoning and inspectable decisions bring autonomous vehicles, including robotaxis, closer to production and widescale deployment. This news blog was originally published at NVIDIA’ website. It is reprinted here with the permission of NVIDIA. For robotaxis and other autonomous vehicles (AVs), the hardest problems aren’t the everyday scenarios. They’re …

Amit Kumar, Director of Product Management and Marketing at Cadence and Amit Sulakhe, Director in the Vision Group at Cadence Pune present “Exploring Radar SLAM: Advancing Localization and Mapping for Automotive, Robotics and Beyond” at the May 2026 Embedded Vision Summit. Reliable localization and mapping are foundational for autonomous vehicles… “Exploring Radar SLAM: Advancing Localization […]…

This blog post was originally published at Synaptics’ website. It is reprinted here with the permission of Synaptics. Developers and system architects today face a growing demand to enable large language model variants on device. They are facing pressure to support transformer-capable models on constrained devices to ensure data privacy, eliminate cloud API charges, and provide […] The post The E…

Here's a roundup of this week's must-read articles. We'll look into the latest developments in high-bandwidth memory, Cadence's agentic AI platform for PCB Design, and a Ku-Band RFoF transmitter. Also, check News Archives - Embedded and Technical Articles Archives - Embedded for the complete list of news and articles from our website. NEWS Partnership Delivers AI-Powered Design to Engineers [...]…

Data center and fiber network solutions provider DC BLOX has obtained a new green Senior Secured Credit Facilities loan for $850 million, up from the original $265 million. The financing will support the company's development of hyperscale data centers throughout the southeastern U.S. The additional capital will also support existing and future projects, including those already pre-leased. “Our b…

Edgecore Networks unveiled its Praxis edge AI services platform for AI service providers at this year’s MWC. Praxis is built for managed service providers, Internet service providers and AI-driven SaaS and VaaS products. The platform supports its five hardware configurations through Synaptics Incorporated and Qualcomm Incorporated from 1 to 70 TOPS. "AI is moving to the edge, and the companies th…

Edge AI infrastructure solutions provider Armada announced a 500+ job deal with Johnson Controls to construct modular data centres at a new plant in Arizona. Armada raised $230M in an oversubscribed Series B round, valuing the company at $2B. The funds will expedite deployment of the U.S. AI stack while meeting growing demand from customers across various industries. "The AI race will not be won …

Structure Research , an independent digital infrastructure research and consulting firm published its 2026 State of Environmental Impact Report . The report tracks developments from 2020-2025 with metrics such as carbon emissions, energy consumption, renewable energy use, and water consumption from reported environmental data from 38 data center providers and nine hyperscale cloud platforms. In 2…

Blackstone and Google are forming a joint venture to create a new TPU cloud company. Google Cloud said it will provide data center capacity and operations, networking, and access to the Google Cloud Tensor Processing Units ( TPUs ). Alternative asset manager Blackstone plans to kickstart the new business with a $5 billion investment that will add 500 MW of capacity in 2027. “This joint venture wi…

Webinar: Rise of the Neocloud and Future of AI Infrastructure Thursday May 28, 2026 | 10am PT / 1pm ET Join Structure Research for an update, analysis and discussion about the emerging Neocloud infrastructure services market. We will trace the neocloud sector's short history and draw parallels with first generation cloud and the rise of hyperscale. This session features proprietary marketshare da…

Red Hat announced expanded capabilities for sovereign and private clouds, empowering organizations with greater data and technology control. New features include simplified compliance, production-ready landing zones, and rapid deployment of AI and cloud services. Building on-premises telemetry for data sovereignty and localized software delivery for regional resilience, Red Hat are rolling out a …

Iceotope , a provider of liquid cooling solutions, raised $26 million in a Series B funding round. Proceeds will be utilized to advance product development, broadening of the patent portfolio, and accelerate partnerships. The round, one of the largest ever for cooling technologies, was co-led by Two Seas Capital and Barclays Climate Ventures to reinforce interest in climate-smart solutions. "Secu…

IREN , a vertically integrated AI cloud provider has signed a five-year AI infrastructure cloud services contract with NVIDIA . The contract is valued at approximately $3.4 billion and IREN will provide NVIDIA with managed GPU cloud services for AI and research workloads, including orchestration and cluster management software in collaboration with Mirantis . The services will utilize air-cooled …

Sharon AI , an Australian neocloud company, announced a five-year cloud computing infrastructure agreement. The deal is worth approximately US$950 million with a global technology company that has significant Asia-Pacific presence. Sharon AI will achieve cloud solution deployment in Australia at NEXTDC data centers, returning revenue starting late 2026. “We are thrilled to continue to expand our …

Implement an end-to-end fine-tuning pipeline for tool-calling language models. This tutorial covers parsing trajectories, structured tool-call extraction, Qwen-compatible ChatML rendering, and efficient LoRA adaptation using PyTorch. The post Fine-Tuning Tool-Calling LLMs: A Complete Guide Using XYZ-Aquila-SFT and Qwen3 appeared first on MarkTechPost .

arXiv:2608.12325v1 Announce Type: new Abstract: Autonomous reasoning is among the most scientifically and economically motivating topics in AI today. Historically the purview of symbolic AI, recent advances have mainly emerged from deep probabilistic generative models. Despite immense interest and rapid progress, the generative AI community has not clearly converged on operational definitions for…

arXiv:2608.12345v1 Announce Type: new Abstract: Language models are increasingly deployed as co-scientists, yet their ability to uphold research integrity under institutional pressure remains unmeasured. We introduce IntegrityBench, a benchmark evaluating misconduct classification, ethical action reasoning and artifact-grounded decision making across 36 paired tasks under a 5-level implicit-expli…

arXiv:2608.12346v1 Announce Type: new Abstract: This position paper argues that modern AI alignment methods - originally designed to prevent harmful output - are dual-use technologies that may easily be misused by malicious actors for censorship and manipulation. By mapping current alignment techniques to the possibility and actual cases of misuse, we show that the quest for a "perfectly aligned"…

arXiv:2608.12368v1 Announce Type: new Abstract: Agreement with human judgments is a common proxy for evaluating the alignment of large language models (LLMs). Yet agreement in final labels does not show that human annotators and models rely on the same moral grounds. Two agents may reach the same judgment while appealing to different principles, contextual assumptions, or interpretations of the s…

arXiv:2608.12371v1 Announce Type: new Abstract: Stream-processing systems increasingly operate across heterogeneous mobile edge--cloud infrastructures, where workload volatility, resource contention, and stringent quality-of-service (QoS) requirements complicate decentralized scheduling. This paper proposes \emph{MAS-DecStream}, whose main contribution is \emph{LLM-MR-CNP}: an extension of the cl…

arXiv:2608.12372v1 Announce Type: new Abstract: AI systems are increasingly employed as decision aids, decision delegates, or autonomous decision-makers. This position paper argues that in many settings, particularly high-stakes decision-making, we need accurate cognitively-aligned AI systems that reason similarly to their users, and faithfully communicate their reasoning. We review evidence that…

arXiv:2608.12373v1 Announce Type: new Abstract: Large language models are increasingly used in strategic and advisory contexts, yet their safety alignment is typically evaluated in English only. We test nine models from six providers and ask whether the language of a prompt can change a model's decision in a high-stakes scenario. We use single-turn game-theoretic vignettes in which a model advise…

arXiv:2608.12385v1 Announce Type: new Abstract: As large language models serve more requests, cumulative inference cost is becoming increasingly important relative to one-time training cost. The two inference phases stress hardware differently: prompt prefill is parallel and typically compute-bound, whereas autoregressive decode is sequential and often memory-bandwidth-bound. Conventional width o…

arXiv:2608.12389v1 Announce Type: new Abstract: Cross-domain zero- or few-shot personalization aims to generate user-preferred responses in unseen conversational domains from only a handful of target-domain interactions. Existing adaptation methods struggle to calibrate update magnitude under sparse evidence and thus overfit, whereas history-transfer methods often entangle user preferences with s…

arXiv:2608.12395v1 Announce Type: new Abstract: We describe Research Assistant, an internal LLM-based system developed at AstraZeneca to help scientists and clinicians explore biomedical questions across a broad range of data sources. The system provides a chat-style interface that brings together evidence from scientific literature, knowledge graphs, chemistry, clinical trials, safety resources,…

arXiv:2608.12426v1 Announce Type: new Abstract: Large language models are increasingly deployed in settings that require simultaneous adherence to multiple explicit constraints - reasoning structure, safety boundaries, output schemas. Individual constraints are handled proficiently, but the compositional regime, where many must hold jointly, remains poorly characterized: how rapidly does performa…

arXiv:2608.12428v1 Announce Type: new Abstract: Memory is a core component of AI agents, enabling them to accumulate experience, maintain personalization, and adapt over long-term interactions. However, existing memory systems often remain fixed after development, limiting their ability to adapt their memory models, organization strategies, and procedural knowledge through continued use. We prese…

arXiv:2608.12476v1 Announce Type: new Abstract: Long-term agent memory is usually treated as select--store--retrieve, but retrieval does not decide whether contradictory, superseded, retracted, deleted, or stale records may support an outgoing claim. We introduce Governed Persistent Memory (GPM), an auditable bitemporal state-transition model with source-bound admission, derived lifecycle state, …

arXiv:2608.12522v1 Announce Type: new Abstract: LLM-based program evolution systems such as FunSearch and AlphaEvolve have shown strong ability to discover novel algorithms, but typically optimize each task in isolation, discarding search experience after completion. We introduce $\varepsilon$-MemEvo, a framework for cross-task knowledge transfer in LLM program evolution. $\varepsilon$-MemEvo sto…

arXiv:2608.12555v1 Announce Type: new Abstract: Predictive explanation methods attribute a model output; they do not, by themselves, attribute an intervention effect on the real-world outcome. We introduce the Causal Attribution Score (CAS), a compact score architecture for causal explanation. CAS starts from an identified interventional coalition game, allocates the joint intervention contrast w…

Turning off this setting won't affect invisible benchmarks used to identify an AI generated file.

Meta released Glimmer this week, an open-weight AI model anyone can download and run on their own hardware — a contrast to Muse Spark, the company’s more powerful model that stays locked behind its own APIs. The release landed alongside a letter from Mark Zuckerberg arguing AI should be “for everyone” rather than controlled by a handful of labs, but as Equity’s […]

The idea that GPUs are poorly suited for agentic workflows may be a misconception, according to French startup Kog.

Natural gas prices could triple in some parts of the U.S., which could saddle hyperscalers with massive bills to power their AI data centers.

Meta released Glimmer this week, an open-weight AI model anyone can download and run on their own hardware — a contrast to Muse Spark, the company’s more powerful model that stays locked behind its own APIs. The release landed alongside a letter from Mark Zuckerberg arguing AI should be “for everyone” rather than controlled by a handful of labs, but as Equity’s […]

Z.ai released GLM-5.3 on August 14, 2026. The model reuses the 743B GLM-5.2 base unchanged. Every reported gain comes from scaled post-training: more long-horizon task environments, more environment types, longer training. Terminal-Bench 3.0 moves from 4.6 to 28.3, and DeepSWE v1.1 from 46.2 to 66.9. Cybersecurity moved further than Z.ai says it planned, with CyberGym at 84.5% and ExploitBench mo…

Cactus Compute released Needle 2, an open 45M-parameter model for tool calling, device use, and structured extraction. The full model is a single 14MB binary that runs a session in about 28MB of RAM. It leads both Seal-Tools splits while targeting hardware with no GPU and no NPU. The post Meet Needle 2: An Open 45M-Parameter Tool-Calling Model That Ships as a 14MB Binary and Runs a Full Session i…

This tutorial provides a complete workflow for building a compact, reasoning-focused language model. By streaming the SupraLabs reasoning corpus from Hugging Face, we apply quality filters and curate data for Supervised Fine-Tuning (SFT). Using SmolLM2-135M-Instruct and LoRA, we demonstrate an end-to-end pipeline—from dataset analysis and heuristic cleaning to efficient training and inference—ena…

Built as a post-training variation on Z.ai's open source model GLM-5.2, Writer says the new system should provide deployment-ready capabilities at a much lower price.

AI is expensive, Ali Ghodsi tells TechCrunch. With so many investors wanting into his latest round, he said yes to more than planned.

OpenAI is launching a preview of a sped up version of its latest, most powerful model, in an effort to court enterprise users.

IBM plans to train and certify tens of thousands of consultants on OpenAI's technologies as part of this deal.

Anthropic researchers found AI agents can clash, collude, and coordinate in unexpected ways, raising new questions about whether today’s safety tests capture the risks of multi-agent systems.

Google has released Gemini 3.7 Flash, a refinement of Gemini 3.6 Flash with algorithmic improvements to its reasoning core. It handles text, images, audio, and video across a 1M-token context window with 64K-token output, and supports customizable thinking configurations. Coding results move notably: 43.6% on FrontierCode 1.1 Main versus 34.4%, 65.3% on DeepSWE v1.1, and 1588 Elo on WebDev Arena.…

OpenAI has replaced chief revenue officer Denise Dresser after just nine months on the job, tapping Wiz president and chief operating officer Dali Rajic to take on frontier lab's top sales job.

The video shows Sheets canvas in action.

Liquid AI released LFM2.5-VL-3B, a 3.1B-parameter vision-language model built for on-device deployment. It averages 80.7 on ScreenSpot-v2 and lifts RefCOCO grounding from 57.1 to 87.9. Function calling is new to the VL line, with ToolSandbox moving from 26.4 to 59.5. The model fits in roughly 3 GB and decodes 228 tokens/s on an Apple M5 Max. The post Liquid AI Releases LFM2.5-VL-3B: A 3B Vision-L…

Microsoft is simplifying Copilot by combining its consumer and business apps, and dropping AI-generated podcasts, Group Chats, Deep Research, and its Mico character.

Nvidia has a plan to make sure its GPUs won't lose value. It wants to convince a new crop of financiers to keep lending for AI buildouts.

The tech giant has considered a nine-figure budget for the payments, according to the WSJ.

Dyna Robotics has released Dyna-2, a world-action model pre-trained on more than one million hours of egocentric human video. The technical report establishes three results: a scaling law on human data to 1M hours, the first transfer of that law to unseen robot data, and evidence that video co-training drives cross-embodiment generalization. The post Dyna Robotics Introduces Dyna-2: A World-Actio…

SpaceXAI released Grok 4.6 on August 12, 2026 — a post-training upgrade over Grok 4.5, not a larger base model. It ties GPT-5.6 Sol Max at 61 on the Artificial Analysis Intelligence Index, ships 500K context and a new xhigh reasoning level, and holds pricing at $2/$6 per million tokens. The coding benchmarks are where it still loses. The post SpaceXAI Releases Grok 4.6: A 500K-Context Frontier Mo…

Is Anthropic's new watermarking system a travesty? Some have taken to social media to complain that it is.

Build a custom LLM post-training pipeline using AllenAI’s Open Instruct framework. This comprehensive guide walks through Supervised Fine-Tuning (SFT), Direct Preference Optimization (DPO), and Reinforcement Learning with Verifiable Rewards (GRPO), optimized to run efficiently on 16GB hardware without needing heavy distributed computing infrastructure. The post AllenAI Open Instruct Tulu 3 Post-T…

NVIDIA's open 30B MoE targets the agent execution layer, with Switchyard routing each step to the cheapest capable model. The post NVIDIA AI Releases Nemotron 3.5 Lightning: A 30B Open MoE with 3B Active Parameters, and NeMo Switchyard Model Router appeared first on MarkTechPost .

AMIE promotional video

“GeoPT” helps AI models understand the basics of physics so they can simulate how objects respond to things like wind and water more efficiently and accurately.

Advisor UI in Google Ads and Google Analytics

By focusing on electrolytes, MIT scientists are making sodium-metal batteries a more practical energy storage option.

July AI recap header

Study finds non-experts deferred to LLM-based diagnostic assistance, even when it was wrong, while clinicians caught AI errors.

An expert in machine learning, statistics, and computation, Rakhlin succeeds Professor Ankur Moitra.

Illustrations of a laptop, an AI spark, messages, code, and a 3-D cube

Director of CSAIL and MIT professor honored for her contributions to robotics, artificial intelligence, and autonomous systems.

MIT students and postdocs discussed science funding and research with policymakers in Washington during the MIT Science Policy Initiative’s annual Congressional Visit Days.

The visionary PhysioNet platform launched 25 years ago, based on a system developed at MIT in the 1970s. It has become one of the most comprehensive biomedical and clinical data repositories in existence.

Managed Agents Gemini 3.6 Flash, Hooks and Triggers

Illustration of a black magnifying glass in a white circle on green grass surrounded by items related to fun activities like tennis and games

An illustrated black magnifying glass with a sparkle in a white circle surrounded by a dinner party tablescape

PhD student Lauren Fortier is building on the experience she gained operating a nuclear plant for the Navy to solve a critical hurdle in the wider adoption of the energy source.

Initial research projects advance national priorities across natural resources, manufacturing, nuclear physics, and more.

Known for his clear and elegant writing style, Bertsekas shaped fields from control and optimization to large-scale computation and artificial intelligence.

Gentle Monster glasses, Warby Parker glasses, a prompt asking for the history behind a pictured building, and a prompt asking to book a table at a pictured restaurant

Assistant Professor Bailey Flanigan has arrived at complex computational methods for helping democracy thrive.

Connected apps rendering

Text "Gemini Omni and Personal Avatars in Google Vids" surrounded by various images

Researchers developed an automated framework that helps AI models generate CAD programs more accurately and efficiently.

Assistant Professor Pat Pataranutaporn describes a new interface that lets everyday users glimpse inside an AI's neural network before their chatbot ever says a word.

Through research and entrepreneurship, Professor Devavrat Shah is helping to design methods that can handle constant decision-making using limited computational resources.

MIT students designed, built, and tested a jet engine with AI copilots, assessing AI’s usefulness in developing high-performance aerospace systems.

Google Images logo surrounded by illustrations of people searching for different images

Managed agents feature bundle launch

June Pixel Drop hero

Google, the New York Jobs CEO Council and Urban Assembly hosted an AI summit for 150 education and industry leaders.