Reporting AI nonsense. Latest AI News on AI Agents, Model Releases, Tools, Leaks, and Rumors ποΈ
Berlin, Germany
Joined October 2015
- Tweets31.9K
- Following998
- Followers82.2K
- Likes100K
Google killed Jules AI agent π
Its planned functionalities are about to arrive to Antigravity instead.
> βWeβre working on bringing our learnings from Jules into Antigravity so you get the best of both worlds: autonomous remote background task execution and an agent-first development environment.β
Antigravity already has some traces of a new βChief of Staffβ agent spotted earlier.
I feel like it was unfortunate that Jules v2 got planned at a moment when AI super apps emerged and Google had to fight back with Codex.
Monitoring π
When we launched Jules, it was an experiment to learn how asynchronous coding agents can help developers ship faster. Seeing what you built taught us so much about where agentic development is headed, directly shaping several products across @Google.
Weβre working on bringing our learnings from Jules into @antigravity so you get the best of both worlds: autonomous remote background task execution and an agent-first development environment.
As we focus on this unified experience, Jules will be winding down.
Whether you used Jules to crush bugs, write tests, or launch featuresβ thank you for shaping the future of agentic coding! ππ¦
Canva now has its own AI vision-language model capable of editing existing designs, called Brush.
Users can describe a change, and Brush then plans and executes the edits across the design. According to Canva, Brush is 69% faster overall on general design editing tasks and 46% cheaper per task than frontier models.
π€ Paid partnership
Brush can be accessed inside the Canva AI sidebar while a design is open in the editor or attached in the chat. Changes made by Brush can be edited inside the design, like any other Canva file.
Brush follows Craft, Canva's first in-house image model, released about a week earlier.
Test it out π
canva.com/
GOOGLE π₯: Antigravity got a new Plugins marketplace with loads of different connectors!
Jira, Confluence, Bitbucket, Firebase, Notion, Stripe, Linear, Figma, Canva, Trello, Dropbox, Datadog, Postman, Vercel, and many more!
It is finally here π
Anthropic introduced self-assessment, a new technique that lets Claude give each conversation a preliminary safety assessment right after it responds.
> On Claude Fable 5, self-assessment flags 78% of the conversations that a stronger, prompted classifier flags as potentially harmful, and 94% of the high-priority ones, at a 1% false-positive rate.
Anthropic also shed light on how self-assessment and prompted classifiers help block attacks.
> "Each successful attack costs roughly four times as many generated tokens as with no safeguards."
We already live in a world where your security depends on the compute allocated to defend the systems you rely on.
Weβve trained Claude to flag potential misuse in conversations, a technique we call self-assessment. Right after it responds, Claude gives a preliminary safety assessment of the conversation.
Since we began using self-assessment in June, it has surfaced misuse that our other safeguards werenβt designed to catch.
Because Claude has already read the conversation, self-assessment is very efficient. And unlike classifiers, which target a discrete set of harms, it draws on Claudeβs own judgment, so it gets more accurate as our models improve.
Read more on our new Safeguards blog:
safeguards.anthropic.com/202β¦
A bunch of small Claude updates π
- Optimized view on iPad and iPhone Duo. Similar on Chromebooks, Googlebooks, and foldables - Claude on Android now works like a laptop app.
- Multi-account sign-in.
- Claude UI now supports 24 new languages.
- Voice mode answer speed has improved. Non-English voices are now more natural.
- CarPlay is now supported.
- "Pull down to search" feature on mobile is coming soon.
QOL improvements π₯
OPENAI π₯: Usage limits for all Pro 100/200/500 tiers and Business Premium users have been doubled!
Usage for Dots will continue to be separate for now as well.
Building time π
Day 25/
We are doubling all usage limits for all Pro 100/200/500 tiers and Business Premium, effective today.
We heard your kind feedback and wonβt be reducing usage limits for Pro 200, so your Work and Codex usage will stay the same. Weβre also going to continue offering extended limits for dots for now, and usage of your dot will still continue to be separate from your default usage.
And we'll of course also keep shipping better models and other efficiency improvements that increase your effective usage over time.
Snaptrude has launched Snaptrude Agent, an AI agent that turns a design brief into a construction-ready 3D model, including structure, MEP (plumbing, electrical, and HVAC), and drawings.
The agent starts by sharing its plan, then works on the model and schematics. Everything it adds can be edited further, letting architects take over at any point.
Prompt-to-building π₯
π€ Paid partnership
Today we're launching @snaptrude Agent, the most advanced agent for building design.
It does real design work with you, from a site to a construction-ready LOD 400 model with structure, MEP and drawings. Then can go back to your existing software with your families intact.
About a year ago we bet the company on AI. Honestly the insight came from software engineering. It wasn't long ago that autocomplete felt like a big deal, and now engineers direct agents that write a lot of the code. I think the same shift is coming to how we design buildings, and it's already started.
Most designers I meet aren't short on skill, they're short on time. Take the call every design team dreads, the client cuts the budget by 30%. Today that's a bad month for everyone. With the agent you're the director. You decide where the cost comes out, and it updates the layout, coordinates the structure and MEP, and redoes the drawings. So when design exploration gets this cheap, we go from 3 options to 300, and more options means better buildings.
Since August our users have gone through more than 140 billion tokens, and firms like Jacobs, Studio+, Kajima, VMDO, Dekker and hundreds of others use Snaptrude today. In our early access, more than 2,000 designers have run over 25,000 tasks through the agent since September.
We also built a benchmark for this, AECBench, because in AEC an almost correct answer is still wrong. It gives AI the same brief an architect would get, has its design in a real BIM model, and scores the result against the brief and against how architects review design. We're publishing the first results today, link below.
Thank you @prashanthp at @Accel and Shub and the team at @foundamental for backing us when this was still a bet, and the Snaptrude team, who built every bit of this. Really proud of you all.
Thanks to our investors, @scottbelsky, @andreasklinger, @thilokonzok, @immad, @CindyXBi, @DanilaVladi, @jmwass, rahulchaudhary, @adamecourt, Zach Ginsburg, Alex duff, Paul grossinger, really appreciate all the help along the way.
Snaptrude Agent can model a specific product from a manufacturer's spec sheet, down to the bolt positioning on a glass facade.
On active projects, designers can use markup to describe a change, and the agent tests options before addressing the request. Work can be imported and exported from/to Revit while keeping the teamβs components intact.
The agent is free to try, with detailed modeling and exports on the $60 Pro plan.
Test it out π
snaptrude.com
MISTRAL π₯: Le Chonk (Mistral Large 4) is now available on HuggingFace!
> "The new model also features substantial gains in reasoning efficiency that give developers greater performance while lowering inference costs."
> "Mistral Large 4 ranks as the best open-weight model for coding and among the world's most capable open models overall."
It now takes the top spot on Cognition's FrontierCode benchmark among open models!
We need new Mistral Small now π
Today, we release the weights of a new and improved Le Chonk, advancing the open frontier for agentic coding and cybersecurity.
π» Built to build: Le Chonk ranks #1 on @cognitionβs FrontierCode among open models, and sets a new benchmark for Pareto efficiency on @artificialanalysis.
π‘ Built to defend: State-of-the-art in cyberdefense, put through the paces by the world's leading cyber vendors, including @CrowdStrike, @cloudflare, @rubrik, @zscaler, and others.
Built for sovereignty: Make it yours. Download the weights on @huggingface , customize it with Mistral Forge, and deploy it wherever you need it.
Read more from our team and customers π:
DAILY AI BRIEF π β Oct 29
GOOGLE π₯:
- Gemini 4 Carbon tops the AA Intelligence Index at 59, rolling out first to cyber defenders.
- Alphabet Q3 revenue rose 25% to $127.6B, with Cloud up 103% and Gemini at 1B users.
- Connected Apps Platform is in preview for Search and Gemini, with Adobe, Calm and Peloton.
- Gemini Notebook will soon get Connectors, starting with Google Drive.
- XREAL AURA, the first Android XR glasses with Gemini, launched from $1,099.
OPENAI π₯:
- API docs reference an unannounced GPT Image 2.6 with markup-guided editing.
- ChatGPT Pro and Business users can claim up to $500/month in API credits.
- US Pro users can now message Dots via iMessage or RCS.
- ChatGPT and ChatGPT Work will merge later this year, per an early pop-up.
- OpenAI API, Codex and ChatGPT Work are now on Oracle Marketplace in the US.
MICROSOFT π₯:
- GitHub launched Copilot Fusion, nearly matching Claude Code at about half the cost.
- Copilot Projects orchestrates multiple agents across repos.
- Copilot Automations runs jobs on demand, on a schedule or on GitHub activity.
- GitHub added a Copilot Security extension and Copilot dreaming.
- Microsoft FY27 Q1 revenue hit $92.3B, with Azure up 48%.
SPACEXSI π₯:
- Grok API now accepts guidelines to steer handling of content categories.
- Grok Bot launched referrals with a two-week free Premium trial for friends.
ANTHROPIC π₯:
- Claude for Legal added ready-made artifacts for contract review and cite checks.
PERPLEXITY π₯:
- Voice mode now runs on OpenAI's gpt-live-1, and Computer supports multiple sessions.
MISTRAL π₯:
- Mistral's Responses API is in public preview for building agents.
NVIDIA π₯:
- Nemotron 3.5 Transcribe is an open-weight model for transcription in 40 languages.
AMAZON π₯:
- Amazon unveiled five new Echo devices powered by Alexa+.
* Used Grok to compose this brief, cherry-picking the news and doing some post-editing.
DAILY AI BRIEF π β Oct 28
GOOGLE π₯:
- Google DeepMind announced Genie 4, its most realistic interactive world model with natural-language steering and native audio, rolling out first to Trusted Testers before Google AI Ultra and Gemini API access.
- Gemini Robotics ER 2 is now in public preview on Google Cloud for enterprise robotics deployments.
- Gemini Notebook on mobile can now record up to an hour of live audio and use it as a source.
ANTHROPIC π₯:
- Anthropic shared early results from drug-development research partnerships with six organizations using Claude.
- Anthropic is letting all US Claude Max subscribers off the Claude Duet waitlist this week, with US Pro users next in line and no timeline yet for access outside the US.
OPENAI π₯:
- ChatGPT added iCloud Mail and Calendar plugins on all plans across web, desktop and iOS.
- Codex CLI 0.165.0 adds a built-in Google Cloud Vertex AI provider and GPT-6.1 Terra in the model catalog.
MICROSOFT π₯:
- MAI-Image-2.7-Preview launched as Microsoft AI's new image-editing model, debuting at #3 on Arena Edit and #2 on AA-Image-Editing.
- MAI-Image-2.6 and MAI-Image-2.6-Flash are now generally available in Microsoft Foundry.
- MAI Cyber 1.1 Flash is live on MDASH, matching MAI Cyber 1 Flash at 18% lower cost per scan.
- GitHub Copilot now supports server-to-server authentication for GitHub Apps in public preview, ahead of GitHub Universe opening today.
SPACEXSI π₯:
- Grok Bot now supports Memory Import to bring over preferences and context from another AI assistant.
- Cursor Design Mode gained Annotate for editing elements, drawing on the page and reviewing designs at different screen sizes.
META π₯:
- Muse is now available in Singapore on iOS, Android and web.
NVIDIA π₯:
- NVIDIA and Meta are expanding their collaboration on Muse models, with broader Vera Rubin deployments planned at Meta.
- Vera Rubin NVL72 and Vera CPUs are now generally available on Oracle Cloud, with OCI Superclusters scaling to multi-gigawatt size.
ARTIFICIAL ANALYSIS π₯:
- Intelligence Index v5 launched with new mostly private benchmarks, putting Claude Opus 5.5 on top at 58.
LIQUID AI π₯:
- Liquid AI released LFM3, an open-weight family of six 2B-class text, vision, audio and omni models on a new tri-hybrid architecture.
* Used Grok to compose this brief, cherry-picking the news and doing some post-editing.
NVIDIA released Nemotron 3.5 Transcribe, a new open-weight, 30B MoE model with ~4B active parameters.
> Nemotron 3.5 Transcribe can transcribe in 40 languages, perform speaker attribution, translate, handle contextual prompting, and understand audio.
> It ranked first in VoiceArenaβs 14-system diarization + ASR comparison.
Meet NVIDIA Nemotron 3.5 Transcribe ποΈ
An open-weight, promptable 30B MoE model with ~4B active parameters that brings transcription in 40 languages, speaker attribution, translation, contextual prompting and audio understanding into one checkpoint.
Watch it keep track of who said what when everyoneβs talking at once.
1/ π§΅
Articul8 launched Bivia, its own decision model that can be connected to AI agents and hold back when confidence is low.
> Bivia weighs each decision against the goal of the whole task, instead of judging each step on its own.Β
> It supports structured replies and can be used by AI agents mid-step.
> It is multimodal and was specifically trained on active production workflows.Β
> Agents can include context from the task they are running along with each question.
> Bivia currently costs $0.042 per million input tokens with no subscription.Β
Small mistakes add up over long tasks. At 95% accuracy per step, an agent finishes a 20-step task only about 36% of the time.
π€ Paid partnership
Introducing Bivia.
A multimodal AI decision model built to complete entire workflows, not just answer individual questions.
We've spent 3+ years building and deploying AI in production across semiconductors, energy, manufacturing, and other complex industries.
Today, we're bringing that experience to everyone.
Articul8 also develops domain-specific AI models for industries such as energy, manufacturing, and aerospace.
Test it out π
articul8.ai/bivia
π€ Paid partnership
OPENAI π₯: ChatGPT and ChatGPT Work will merge this year, according to Tibo. Some users may've seen a pop-up highlighting this change on ChatGPT; however, it was a mistake.
> Tibo: "Team is working on this but it's not rolling out just yet, sorry for any confusion."
> Tibo: "Plan is to bring together best of both Chat / Work capabilities together at some point this year, but we're going to proceed carefully here and take our time."
OPENAI π₯: ChatGPT Pro and Business subscribers can now claim monthly API credits!
- Pro plans get $100, $200, or $500 per month by tier.
- Business plans get $20 per paid Standard seat or $100 per paid Premium seat, pooled up to $500 per workspace.
That's exactly what is needed for testing π
GitHub dropped a slew of announcements at its UNIVERSE'26 event.
> Copilot Fusion - Can coordinate multiple models and workflows when one model alone isnβt enough.
> Copilot Projects - One place to see where the work stands and direct what Copilot takes on next.
> Security extension for GitHub Copilot - Helps review what you and your agents just wrote alongside your repository's code scanning alerts in one Security canvas.
> Copilot dreaming - It suggests improvements to your personal memories, repository instructions, agent skills, and custom instructions for Copilot code review.
> Copilot Automations - Define the job once, then run it on demand, on a schedule, or in response to activity in GitHub or a connected service.
Github Copilot as a Superapp πͺπ
Frontier performance at a fraction of the cost.
In the @ArtificialAnlys Coding Agent Index, Copilot Fusion Max scored within two points of Claude Code with Opus 5.5 Max at about half the API cost per task.
Get 30% off through Nov. 30. Try it now. β¬οΈ docs.github.com/copilot/concβ¦
ALT Chart titled "Coding Agent Index vs. Cost per Task" by Artificial Analysis. The chart displays Artificial Analysis Coding Agent Index vs. average pay-per-token API cost per task (USD). It compares various agents, including some from Claude Code, Codex, Antigravity CLI, DeepSeek Harness, Grok Build, Opencode, Muse Code, Kimi Code CLI, and GitHub Copilot. The x-axis represents higher cost per task (USD) to lower cost per task (USD), and the y-axis represents the coding agent index. The chart highlights a "Most Attractive Quadrant" in green and uses dots to differentiate agents. Copilot Fusion (max) sits within the most attractive quadrant.
ALT Bar chart titled "Artificial Analysis Coding Agent Index". Incorporates 3 benchmarks: FrontierCode v1.1, Terminal-Bench 4.0, SWE-Atlas-QnA-AA. A higher scroe denotes higher average intelligence across becnhmarks. Copilot Fusion Max from GitHub Copilot has a score of 60. The chart also includes various coding agents from Claude Code, Devin Fusion CLI, Antigravity CLI, Codex, Grok Build, DeepSeek Harness, Muse Code, Opencode, Antigravity CLI, and Kimi Code CLI, with scores ranging from 62 to 41. Source and logo of Artificial Analysis are included at the bottom.
Google XREAL officially launched XREAL AURA, the first Android XR Glasses!
> Dramatically lighter - "Users can comfortably keep them on for hours, whether watching a movie or getting work done on multiple screens."
> "Gemini works across everything in view, including your laptop window, so you can ask it to tidy up your windows or brainstorm about whatever you're looking at."
> "XREAL AURA is available for users 18+ to order today in the U.S., Canada, and Korea, and starting November 2 in Japan, with launch pricing starting at USD 1,099 through the end of the year."