Meta to Shut Down Metaverse Project in June After $80 Billion Investment, According to Social Media Report

A social media post claims Meta will permanently shut down its Metaverse project in June, following an estimated $80 billion investment. The claim, if accurate, would mark the end of one of the most capital-intensive tech experiments.

x.com/Mar 18, 2026/3 min read

fundingxrbusiness

HuggingFace robot mascot reading a markdown document titled Daily Papers SKILL.md, with search bars and database…

Products & Launches

85

HuggingFace Launches Daily Papers SKILL.md for AI Agents to Read, Search, and Fetch Research Papers

HuggingFace released Daily Papers SKILL.md, a tool enabling AI agents to read paper content as markdown, search papers, find linked models/datasets, and fetch papers via API.

x.com/Mar 18, 2026/3 min read

research toolsapiai agents

A security sandbox interface showing an AI agent model isolated from cloud services with blocked unauthorized…

Products & Launches

97

NVIDIA Open-Sources NeMo Claw: A Local Security Sandbox for AI Agents

NVIDIA has open-sourced NeMo Claw, a security sandbox designed to run AI agents locally. It isolates models from cloud services, blocks unauthorized network calls, and secures model APIs via a single installation script.

x.com/Mar 18, 2026/3 min read

open sourcesecurityai agents

Jensen Huang onstage at GTC 2026, gesturing toward a lineup of humanoid and industrial robots under bright lights…

Products & Launches

85

Jensen Huang at GTC 2026 Declares 'Arrival of Physical AI' Amid Showcase of Advanced Robots

NVIDIA CEO Jensen Huang presented a major showcase of advanced robotics at GTC 2026, framing it as the arrival of 'physical AI.' The announcement signals a strategic push into embodied AI systems.

x.com/Mar 18, 2026/3 min read

hardwareroboticsbusiness strategy

Vellum AI persistent desktop agent interface showing autonomous task execution with dedicated email and accounts…

Products & Launches

85

Vellum AI Launches as Persistent Desktop Agent with Dedicated Email and Accounts

Vellum AI operates as a persistent desktop agent with its own email and accounts, executing tasks autonomously. The tool claims enterprise-grade security while running continuously on user systems.

x.com/Mar 18, 2026/3 min read

product launchautomationdesktop software

Pantio Launches AI-Powered 'Digital Immortality' Service …

Products & Launches

85

Pantio Launches AI-Powered 'Digital Immortality' Service to Recreate Deceased Loved Ones

Pantio, a new AI startup, has launched a service that creates interactive digital replicas of deceased individuals using their personal data. The company claims it can preserve memories and personality traits through conversational AI.

x.com/Mar 18, 2026/3 min read

product launchethicsstartup

AI-generated diagram showing a structured software implementation plan with labeled blocks for architecture…

Products & Launches

85

CodeRabbit Launches 'Planner' Feature to Shift AI Coding from Implementation to Architecture Validation

CodeRabbit launched Planner, a feature that generates structured implementation plans from descriptions and context before code is written. It aims to move architectural debates from PR reviews to the planning phase, working with multiple AI coding tools.

x.com/Mar 18, 2026/3 min read

product launchai workflowdeveloper tools

OpenAI Whisper demo screen showing English-to-Spanish real-time translation interface with audio waveforms and text…

Products & Launches

85

Whisper's Real-Time Translation Demo Shows Practical Progress Toward Universal Translation

OpenAI's Whisper model demonstrated real-time translation from English to Spanish, showcasing progress toward practical universal translation tools. The demo highlights incremental but meaningful improvements in speech-to-speech translation latency and quality.

x.com/Mar 18, 2026/3 min read

speechwhisperopenai

A designer using Figma Make, typing a text prompt that transforms static UI design layers into a clickable prototype…

Products & Launches

85

Figma Make Generates Clickable Prototypes from Design Files via AI Prompt

Figma has launched Make, an AI tool that converts static Figma designs into functional, clickable prototypes based on a text prompt, aiming to streamline the handoff between design and development.

x.com/Mar 18, 2026/3 min read

product launchprototypingdesign tools

A sleek silver robot arm assembling a microchip on a glowing circuit board, with floating data graphs and code lines…

Products & Launches

97

Minimax M2.7 Achieves 56.2% on SWE-Pro, Features Self-Evolving Training with 100+ Autonomous Optimization Loops

Minimax has released M2.7, a model that reportedly used autonomous optimization loops during RL training to achieve a 30% internal improvement. It scores 56.2% on SWE-Pro, near Claude 3.5 Opus, and ties Gemini 3.1 on MLE Bench Lite.

x.com/Mar 18, 2026/3 min read

model releasecoding aireinforcement learning

Two AI agent icons, one blue and one orange, connected by a glowing network link representing the A2A protocol for…

Products & Launches

85

Google's A2A Protocol Aims to Standardize Communication Between AI Agents

Google is developing the Agent2Agent (A2A) protocol, a standardized framework for AI agents to discover, communicate, and collaborate on tasks. The protocol aims to solve the interoperability problem in a growing but fragmented agent ecosystem.

x.com/Mar 18, 2026/3 min read

protocolinfrastructureai agents

A laptop screen displays a SureThing AI dashboard with a 300K email processing metric, a coffee cup beside it, and a…

Products & Launches

89

OpenClaw Early Contributor Switches to SureThing, Claims It Processed 300K Emails in One Hour Where Claude and Codex Failed

An early contributor to the OpenClaw AI project has publicly switched to competitor SureThing, claiming it processed 300,000 emails in one hour where Claude and Codex failed. The contributor described OpenClaw as 'Linux' and SureThing as 'Mac' in terms of user experience.

x.com/Mar 18, 2026/3 min read

developer experiencebenchmarksai tools

TaxHacker: Open-Source, Self-Hosted AI App Automates Rece…

Products & Launches

87

TaxHacker: Open-Source, Self-Hosted AI App Automates Receipt and Invoice Processing

A developer released TaxHacker, a self-hosted AI accounting app that extracts data from receipts/invoices in any language, converts currencies, and exports to CSV. It's fully open-source under MIT license and runs via Docker.

x.com/Mar 18, 2026/3 min read

open sourceai applicationfintech

Baidu's Qianfan-OCR model interface on Hugging Face showing document text recognition and layout analysis features

Products & Launches

85

Baidu's Qianfan-OCR End-to-End Document Intelligence Model Released on Hugging Face

Baidu has released Qianfan-OCR, an end-to-end document intelligence model, on Hugging Face. The model appears to be a unified framework for optical character recognition and document understanding tasks.

x.com/Mar 18, 2026/3 min read

multimodal-aicomputer-visionmodel-release

Person typing on laptop with Gamma interface showing an AI-generated presentation, colorful slides and design…

Products & Launches

85

Gamma Launches 'Gamma Imagine' AI Feature for Instant Document and Presentation Design

Gamma has launched 'Gamma Imagine,' an AI feature that generates complete documents and presentations from text descriptions. The company claims it eliminates the need for templates and manual design work.

x.com/Mar 18, 2026/3 min read

product launchproductivity toolsgenerative ai

M2.7 AI Model Scores 56.22% on SWE-Pro Benchmark, Highlighted for Frontend Task Performance

Products & Launches

85

M2.7 AI Model Scores 56.22% on SWE-Pro Benchmark, Highlighted for Frontend Task Performance

The M2.7 AI model has been released, with its developer highlighting strong performance on frontend development tasks. It achieved a score of 56.22% on the SWE-Pro coding benchmark.

x.com/Mar 18, 2026/3 min read

model releaseai codingbenchmarks

Developer using open-source web app to train and compare hundreds of LLMs locally with reduced VRAM usage for…

Products & Launches

85

Unsloth Studio: Open-Source Web App Cuts VRAM Usage for Local LLM Training and Dataset Creation

Unsloth has launched Unsloth Studio, an open-source web application that enables users to run, train, compare, and export hundreds of LLMs locally with significantly reduced VRAM consumption. It also converts files like PDFs, CSVs, and DOCXs into training datasets.

x.com/Mar 18, 2026/3 min read

open-sourcedeveloper-toolsllm

Anthropic's Claude 3.5 Sonnet dashboard showing speed metrics and cost comparisons

Products & Launches

85

Anthropic Launches Claude 3.5 Sonnet with 70% Lower Cost, 3x Speed Boost

Anthropic released Claude 3.5 Sonnet, claiming 70% lower cost and 3x faster speed (~100 tokens/sec) than its predecessor. The model is positioned as a mid-tier option between Sonnet 3.0 and Opus 3.0.

x.com/Mar 18, 2026/3 min read

inference-optimizationanthropicmodel-release

Brian Armstrong, Coinbase CEO, stands in a modern office with holographic data displays showing AI code generation…

Products & Launches

87

Coinbase CEO: AI Agents Now Write Over 50% of Code, Resolve 60% of Support Tickets

Coinbase CEO Brian Armstrong reports AI agents now generate over half of the company's code and resolve 60% of support tickets. The company equips these agents with stablecoin wallets for autonomous machine-to-machine payments.

x.com/Mar 17, 2026/3 min read

ai agentsfintechenterprise ai

xAI Hires Wall Street Bankers and Credit Lenders to Train…

Products & Launches

85

xAI Hires Wall Street Bankers and Credit Lenders to Train Grok on High-Level Finance

Elon Musk's xAI is recruiting finance professionals from Wall Street and credit lending institutions to train its Grok AI model on specialized financial knowledge. This move signals a targeted push to build domain expertise beyond general-purpose LLM capabilities.

x.com/Mar 17, 2026/3 min read

financexaigrok

OpenAI Launches GPT-4o Mini and Nano: Nano Model Can Proc…

Products & Launches

89

OpenAI Launches GPT-4o Mini and Nano: Nano Model Can Process 76,000 Images for $52

OpenAI released GPT-4o Mini and Nano models. Developer Simon Willison calculated the Nano could analyze his 76,000-photo library for $52 total.

x.com/Mar 17, 2026/3 min read

product launchcomputer visionai economics

A browser window displays a video creation interface with a timeline and editing tools, while an AI agent icon…

Products & Launches

85

OpenClaw Agent Demonstrates In-Browser Video Creation Without App Switching

OpenClaw agent can now create videos directly within a browser interface without opening separate applications or switching tabs. The development suggests progress toward more integrated multimodal AI workflows.

x.com/Mar 17, 2026/3 min read

open sourceai agentsmultimodal

SK Group Chairman Chey Tae-won in a business setting, gesturing while discussing a prolonged memory chip shortage…

Products & Launches

85

SK Group Chairman Forecasts Memory Chip Shortage Until 2030, Warns of Sustained Price Increases

SK Group Chairman Chey Tae-won predicts the global memory chip supply crunch could persist until around 2030, with wafer supply lagging demand by over 20% and prices continuing to rise.

x.com/Mar 17, 2026/3 min read

ai infrastructuresupply chainsemiconductors

A smartphone screen displays a chat interface with GPT-5.4 mini, showing a positive user feedback message

Products & Launches

85

GPT-5.4 Mini Reportedly Available, Early User Reports Positive Initial Impressions

A user reports that GPT-5.4 mini is now available, describing initial impressions as 'very good.' No official announcement or technical details have been released.

x.com/Mar 17, 2026/3 min read

rumormodel releaseopenai

NVIDIA staff member at a podium addressing journalists, with a screen showing DLSS5 lighting comparison in a game scene

Products & Launches

85

NVIDIA Employees Clarify DLSS5 Does Not Alter Character Models or Assets, Only Lighting

NVIDIA employees confirmed at a press conference that DLSS5 makes no changes to character models or game assets, countering speculation about AI filters. The visual differences are attributed solely to lighting changes.

x.com/Mar 17, 2026/3 min read

gpudlssai-graphics

A biomimetic fish-shaped underwater drone prototype swims through blue water, its metallic fins and tail mimicking…

Products & Launches

85

Beijing Military Intelligent Technology Demonstrates Underwater 'Fish Drone' Prototype

A brief video shows a biomimetic underwater drone resembling a fish, attributed to Beijing Military Intelligent Technology. The prototype's technical specifications and operational status are unconfirmed.

x.com/Mar 17, 2026/3 min read

unmanned systemsroboticsbiomimicry

Dashboard interface displaying 71 pre-built AI agents for business workflow automation, replacing traditional…

Products & Launches

97

Adaptive Launches 'AI Agent OS' Platform with 71 Pre-Built Agents, Claims to Replace Traditional Software Stack

Adaptive has launched a no-code AI agent platform with 71 pre-built agents designed to automate business workflows. The company claims it can replace tools like Zapier, Make, and virtual assistants with a single unified system.

x.com/Mar 17, 2026/3 min read

product launchno-codeautomation

A security researcher stands before a digital map showing 40,000 exposed OpenClaw servers, with red markers…

Products & Launches

85

Security Researcher Exposes 40,000+ OpenClaw Servers, 12,000 Vulnerable to API Key Theft

A security scan reveals over 40,000 OpenClaw servers are exposed online, with 12,000+ vulnerable to API key and data theft. The researcher published a comparative security analysis of hosted AI providers.

x.com/Mar 17, 2026/3 min read

open sourcesecurityinfrastructure

A smartphone and laptop display messaging app interfaces including Telegram and WhatsApp, showing a text prompt…

Products & Launches

85

OpenClaw's Pexo Agent Generates Videos Directly Within Telegram, Discord, and WhatsApp

OpenClaw has launched Pexo, an AI agent that creates videos from text prompts directly within messaging apps like Telegram, Discord, and WhatsApp, without requiring users to switch applications.

x.com/Mar 17, 2026/3 min read

product launchai agentsgenerative ai

An operator uses a wireless controller to guide multiple robotic hands with precise, low-latency movement…

Products & Launches

87

HexaCercle Demonstrates Multi-Robot Hand Control System with 3ms Latency, 0.001° Precision

HexaCercle has demonstrated a wireless system enabling one operator to control multiple dexterous robotic hands with 1:1 movement replication. The system achieves 3ms data transmission latency and 0.001° collection precision.

x.com/Mar 17, 2026/3 min read

roboticsteleoperationdemonstration