Zlio Everything in AI and robotics, on one page.
Updated 00:31 UTC · Next update 04:00 UTC
Zlio Daily, the newsletter, is coming soon. The day's essential AI and robotics news in your inbox. Leave your email and we'll tell you when it launches.
One email when we launch. No spam, unsubscribe anytime.

Top stories

Archive, page 7

Wednesday 14 October 2026

Pocket Gamer.biz · 17h #

Playgama launches tools for AI agents to publish games

Playgama released tools that let AI agents publish and improve games on its platform, aiming to let autonomous agents handle distribution and optimization tasks.

PlaygamaClosedAPI
OpenAI · 9h #

Third Moment Research runs 10,000 agents for 36+ hours

AI lab Third Moment Research uses GPT-6 Astra, GPT-6.1 Sol, Codex and Daybreak to run 10,000 agents that operate for days and improve with use, OpenAI says.

OpenAIClosedOfficial
Science Robotics · 17h #

Fast Feature Field predicts events for robot perception

Published in Science Robotics, Fast Feature Field (F3) is a general predictive representation of events aimed at improving robot perception.

Science Robotics · 17h #

Ultra Mobility Vehicle robot drives, balances and jumps

Science Robotics describes the system design of the Ultra Mobility Vehicle, a bicycle robot capable of driving, balancing and jumping.

Science Robotics · 17h #

MonoRace uses monocular AI for champion-level drone racing

Science Robotics presents MonoRace, a robust monocular AI approach that reaches champion-level performance in autonomous drone racing.

OpenAI · 1h #

Block builds AI-powered security tooling with OpenAI

Block's security team uses OpenAI technology to investigate threats faster and propose patches in minutes, per an OpenAI customer story.

OpenAIClosedOfficial
@Dheemanthredy · 17h #

Maya-2 open-weight Indic voice model claims top performance

An X post announces Maya-2, billed as the world's best Indic voice AI model with fully open weights, claiming it outperforms ElevenLabs v3, Cartesia Sonic 3.5 and Bulbul v3 across Indian languages; free to builders soon.

♥ 12 on X
Maya-2OpenOpen weights
Konsulteer · 17h #

CYANIS AI launches Sovereign Inference Cloud

CYANIS AI launched Sovereign Inference Cloud, a service for running AI inference inside a customer's own jurisdiction and infrastructure. Pricing and availability were not disclosed.

CYANIS AI
Idéal Investisseur · 17h #

Quadient launches AI agent for customer account management

Quadient launched an AI agent that automates customer account management and email handling for businesses.

QuadientClosedAPI
CoinMarketCap · 17h #

CoinMarketCap argues agentic DeFi scaling hinges on execution

A CoinMarketCap piece says agentic DeFi is stalled and that execution, not just agent capability, is how it scales. It is an opinion/analysis post with no new product or event.

Press Release Hub · 12h #

NEURA Robotics launches Neuron AI foundation model for robots

German robotics unicorn NEURA Robotics introduced Neuron AI, a multimodal foundation model that gives robots perception and memory for physical AI and humanoid applications.

NEURA RoboticsClosed
DagangNews · 18h #

ReadyServer bundles AI model access into every VPS plan without token billing

ReadyServer now includes AI model access with every VPS plan, billing without per-token charges, a flat-rate option for developers running hosted AI workloads.

ReadyServerPaid
www.tuni.fi · 18h #

Finland's Robostudio teaches social robotics as civic skill

Tampere University's Robostudio brings social robotics to the public, teaching Finns how robots interact with people as a new everyday civic skill.

Tampere University
Bloomberg · 10h #

Mittal and tech executives call for stronger AI regulation

Sunil Mittal warned AI is outpacing regulation, while Microsoft and OpenAI executives said warnings alone are not enough and urged new AI safety laws; billionaires amplified the calls amid an Anthropic worm-risk warning.

조선일보 · 18h #

PhoneBot humanoid robot built from outdated smartphones

A 'PhoneBot' humanoid robot is built from repurposed outdated smartphones, using the phones' cameras and sensors for perception in a low-cost robotics approach.

TechCrunch · 5h #

Deezer to let listeners filter out AI-generated music

Deezer said a toggle letting listeners hide fully AI-made or AI-generated music from search and playlists will roll out by the end of the year.

Deezer
TechRepublic · 18h #

Databricks to acquire Neon for $1B

Databricks has agreed to acquire Neon, a serverless Postgres database startup, in a deal reported at around $1 billion, in a move to boost AI-driven database workloads.

Databricks
- Disrupt Africa · 18h #

South Africa's KhayaHR builds AI-first workforce platform

South Africa's KhayaHR is building an AI-first workforce platform targeting the frontline economy, per Disrupt Africa.

KhayaHR
Stock Titan · 9h #

Nvidia and Danaher partner to integrate AI into scientific instruments

Nvidia and Danaher are collaborating to link AI with lab experiments, starting with select instruments, to advance autonomous science.

Nvidia
Caixin Global · 18h #

Chinese startup StepFun launches smartphone to challenge giants

Chinese AI startup StepFun launched its own smartphone, aiming to compete with major handset makers by putting its AI models on dedicated hardware.

StepFunOn sale
Inc42 · 18h #

Spill Games raises $10.5M for AI-led mobile game studio

Spill Games raised $10.5 million to scale its AI-led mobile gaming studio, per Inc42.

Spill Games
CoinMarketCap · 1h #

Coinbase marks one year of x402 agent payment stack

Coinbase touted its agent payment stack on the business unit's first anniversary as x402 micropayments on Base passed 138 million, with reports saying AI agents moved over 200 million payments and that half may be fake.

CoinbaseAPICode
Stripe · 18h #

Stripe powers payments for Gradium's global voice AI

Gradium, a global voice AI service, runs its payments on Stripe. Case study describes how the company bills customers for voice AI usage.

StripeOfficial
NVIDIA · 18h #

Nvidia details dynamic sparse grids in NanoVDB

Nvidia research describes dynamic GPU sparse grids and voxel-sequential processing in NanoVDB, techniques for efficiently representing and updating sparse 3D volumes on GPUs.

Nvidia
Reddit · 18h #

Alibaba: Qwen-Image 2.1 free up to 300K images monthly

A Reddit user shares Alibaba's emailed reply on Qwen-Image 2.1 commercial licensing: despite the research-only Qwen Research License, commercial use is free up to 300K images per month, with paid tiers above.

AlibabaOpenOpen weights
NVIDIA · 10h #

UNCW researchers advance robotics using human gaze

University of North Carolina Wilmington researchers published robotic perception and learning research showing how human gaze can improve robot learning and perception.

UNCW
Reddit · 19h #

Developer forks Chromium to fix browser automation with Claude Code

A developer forked Chromium to build a browser better suited to AI agents, addressing issues like throwaway profiles and poor cleanup in chrome-devtools-mcp.

OpenCode
Reuters · 19h #

Kamala Harris warns AI safeguard window is closing

Kamala Harris warned that the window for establishing AI safeguards is closing and urged US-China cooperation on the issue.

Stocktwits · 19h #

Qualcomm in advanced talks for $4B Modular acquisition

Qualcomm is reportedly in advanced talks to acquire AI software startup Modular for about $4B, a deal that would bolster its AI tooling stack.

Qualcomm
Bloomberg · 3h #

ASML raises outlook but cannot meet AI-driven demand

ASML raised its full-year sales outlook and posted Q4 guidance above expectations on strong AI chip demand, but said it will be unable to meet customer demand in 2027 and 2028. Shares rose.

ASML
Nasscom · 19h #

Nasscom report examines physical AI in automotive manufacturing

Nasscom publishes an analysis of how physical AI is being adopted in automotive manufacturing, covering robots and automation on factory lines.

Nasscom
CNN · 19h #

Grab announces Thailand initiative with OpenAI

Grab announced a new Thailand initiative in partnership with OpenAI, per CNN.

Grab
Reddit · 19h #

jitllm moves running model layers between GPUs mid-reply

jitllm is an inference engine that JIT-generates kernels per hardware and treats layers like movable processes; a demo shifts layers across T4 GPUs mid-reply in 8 ms without re-prefill.

jitllmOpenCode
매일경제 · 19h #

Galaxy Corporation unveils humanoid modeled on CEO's face

Galaxy Corporation unveiled a humanoid robot modeled on its CEO's face, with the company saying it will keep trying despite the risk of failure.

Galaxy Corporation
Dealroom · 13h #

AZOO raises $3.09M; Ambio raises €5.1M; FLÆ Robotics raises €1M+

Hotel OS startup AZOO raised a $3.09M Series A to automate lodging operations with AI. Helsinki's Ambio raised a €5.1M seed for AI matchmaking in recycled plastics procurement. FLÆ Robotics raised over €1M from Purple Ventures to put an AI receptionist in 60+ hotels.

semivision · 19h #

Nvidia adopts PTFE for next-generation AI rack technology

A report details Nvidia's PTFE strategy, describing the fluoropolymer material approach as a hidden enabler behind its next-generation AI server racks and high-speed interconnects.

Nvidia
Bloomberg · 20h #

Singapore's PDG seeks $1.2B loan for Malaysia data center

Asian data center operator Princeton Digital Group is seeking a syndicated loan of up to $1.2 billion to fund its Malaysia buildout, as AI infrastructure borrowing grows.

Princeton Digital Group
FinancialContent · 20h #

TourMind and Qwen partner on AI-powered travel platform

Travel tech firm TourMind says it will use Alibaba's Qwen models to power AI travel planning and booking features.

Qwen
Bloomberg · 20h #

Hitachi Energy CEO discusses AI power and grid demands

Hitachi Energy CEO Andreas Schierenbeck discusses whether electricity infrastructure can expand fast enough to support the AI boom and industrial electrification.

Hitachi Energy
https://www.ababnews.com/ · 20h #

OpenAI cuts ChatGPT and Codex desktop disk use by 60%

OpenAI says it optimized the ChatGPT and Codex desktop apps to reduce their disk storage footprint by 60%.

OpenAIClosedAppFreePaid
← NewerOlder →