AI News - Live AI News & Trends | Blockchain.News

OpenAI boosts COBOL, Fortran, Visual Basic migration tools

According to OpenAI via its blog and X post, its latest model advances improve building and migrating software in COBOL, Fortran, and Visual Basic to help enterprises modernize legacy systems while preserving business logic. (Source)

Share More from OpenAI 10-19-2026 14:30 UTC

Red Queen Bio uses AI to design broad spectrum antibodies

According to The Rundown AI, Red Queen Bio is using AI to develop antibody drugs aimed at preempting future pandemics, including novel viral threats. (Source)

Share More from The Rundown AI 10-19-2026 14:14 UTC

ARK Invest: Tesla Cybercab aims 40% lower robotaxi costs

According to ARK Invest via Sawyer Merritt, Tesla added the Cybercab to its robotaxi fleet and could run ~40% lower cost per mile than Waymo’s next-gen robotaxi, while ARK’s 2031 model projects robotaxi as ~80% of Tesla’s enterprise value. (Source)

Share More from Sawyer Merritt 10-19-2026 14:04 UTC

ARK Invest: Tesla hits 1M driverless miles, eyes FSD V15 ramp

According to ARK Invest via Sawyer Merritt, Tesla logged over 1 million unsupervised driverless miles across seven cities in three states in ~7 months, expects acceleration with FSD V15, and frames it as key to scaling robotaxi service. (Source)

Share More from Sawyer Merritt 10-19-2026 14:02 UTC

ARK Invest projects Tesla $17.5T valuation by 2031: AI, robotaxis

According to ARK Invest via Sawyer Merritt, Tesla’s 2031 base case values the company at $17.5T driven by robotaxis and Optimus, with cumulative AI infrastructure spend rising to $225B. (Source)

Share More from Sawyer Merritt 10-19-2026 13:58 UTC

Meta unveils Muse Glimmer 30B open weights for agentic tasks

According to AI at Meta, the Muse Glimmer 30B open-weight model runs at practical speeds on a single consumer GPU at 4-bit precision and leads peers in tool use and multi-step planning. (Source)

Share More from AI at Meta 10-19-2026 11:33 UTC

OpenClaw v2026.10.5 Release: Sidebar, Undo, Public Links

According to OpenClaw, version v2026.10.5 adds a Sidebar to keep agents accessible, introduces Undo in Skill Workshop 2.0, enables sign-in–free public conversation links, and credits 2,250 PRs from 637 contributors. (Source)

Share More from OpenClaw 10-19-2026 11:11 UTC

Inception unveils Mercury 3 diffusion LLM with 1,000 tps

According to The Rundown AI citing Inception, Mercury 3 is a 40x larger diffusion language model than Mercury 1 that delivers 1,000+ tokens per second, matches or beats GPT-6 Luna and Claude Haiku 5.5, and targets low-latency use cases. (Source)

Share More from The Rundown AI 10-19-2026 10:54 UTC

Grok Bot voice access via iPhone Action Button: Latest

According to Sawyer Merritt on X, users can now trigger and talk to Grok Bot using the iPhone Action Button, enabling faster voice access to the chatbot. (Source)

Share More from Sawyer Merritt 10-19-2026 10:42 UTC

Meta details Prometheus supercluster powering Muse Wonder

According to AI at Meta on X, Meta trained its Muse models, including Muse Wonder, on the Prometheus supercluster, reporting 85%+ pretraining goodput, 90%+ RL goodput, and about 47K GPU hours mean time between interruptions. (Source)

Share More from AI at Meta 10-19-2026 10:14 UTC

Anthropic signals possible AI intelligence ceiling at IMF talk

According to The Rundown AI, Anthropic co-founder Jack Clark told an IMF forum that CEO Dario Amodei has publicly floated setting a ceiling on AI intelligence, warning the window to act is only a few years. (Source)

Share More from The Rundown AI 10-19-2026 09:43 UTC

SoftBank eyes $18B robot guard deal amid legal and civic tests

According to The Rundown AI, SoftBank is weighing an $18B acquisition in robot security as telepresence, delivery robot biometric lawsuits, and micro-robot sewage fixes advance. (Source)

Share More from The Rundown AI 10-19-2026 06:30 UTC

Microsoft expands Project Perception to secure AI agents

According to Microsoft’s official blog, Microsoft expanded Project Perception with continuous agent risk assessment, adaptive permissions, and threat containment to treat AI models as insider risks. (Source)

Share More from Satya Nadella 10-19-2026 05:05 UTC

Anthropic unveils public pay plan amid junior staff downturn

According to The Rundown AI, Anthropic introduced a public-facing compensation plan while internal data shows junior employees losing ground in the AI job market. (Source)

Share More from The Rundown AI 10-19-2026 02:30 UTC

DeepSeek V4.1 Flash scores 77% on DeepSWE under $0.20

According to AI News on X, DeepSeek V4.1 Flash achieved 77% on the DeepSWE coding benchmark at under $0.20 per task, signaling cost efficient software automation. (Source)

Share More from AI News 10-19-2026 01:50 UTC

DeepLearning.AI launches agentic coding lab with live preview

According to DeepLearning.AI on X, the course Building AI Assistants with On-Device Memory now includes a browser-based agentic coding lab with live preview and no installs, available to enroll for free. (Source)

Share 10-18-2026 08:05 UTC

Anthropic CEO forecasts near-AGI in 2–3 years: Analysis

According to a post by Timnit Gebru referencing Dario Amodei’s remarks, Anthropic’s CEO said AI could surpass humans at most tasks within 2–3 years, highlighting vast opportunities and risks. (Source)

Share More from timnitGebru (@dair-community.social/bsky.social) 10-17-2026 13:53 UTC

Yann LeCun to unveil AMI Labs research at AI Dev NYC

According to DeepLearning.AI, Yann LeCun will present his latest AMI Labs work at AI Dev NYC in New York from Nov 30 to Dec 1, highlighting advances in convolutional neural networks. (Source)

Share 10-17-2026 08:05 UTC

Elon Musk says AI and robots needed to fix US deficit

According to Sawyer Merritt on X citing an upcoming Walter Isaacson podcast, Elon Musk said the only way to solve the US national debt is with AI and robots due to limited political will to cut spending. (Source)

Share More from Sawyer Merritt 10-17-2026 08:00 UTC

DeepSeek V4.1 Flash 1017 boosts speed and benchmarks

According to DeepSeek on X, DeepSeek-V4.1-Flash-1017 delivers stronger benchmark performance, faster task completion, and shorter outputs, and is live on its website, app, and API with optimizations for DeepSeek Harness v0.2.1-rc.1, Tencent CodeBuddy, and WorkBuddy. (Source)

Share More from DeepSeek 10-17-2026 00:35 UTC