Claude learns to make use of the pc

Welcome, AI enthusiasts.

Anthropic’s Claude is not just chatting anymore — it’s clicking, typing, and scrolling its way through computers like a human.

The AI agent dam appears to be breaking open this week, and AI capabilities are getting more hands-on by the day (literally). Let’s get into it…

In today’s AI rundown:

• Anthropic’s AI now navigates computers like a human

• Genmo drops open-source AI video model

• Master public speaking with ChatGPT

• Ideogram debuts AI Canvas workspace

• 5 recent AI tools & 5 recent AI jobs

• More AI & tech news

Read time: 4 minutes

LATEST DEVELOPMENTS

ANTHROPIC

🖥️ Anthropic’s AI now navigates computers like a human

The Rundown: Anthropic just introduced a brand new capability called ‘computer use’, alongside upgraded versions of its AI models, which enables Claude to interact with computers by viewing screens, typing, moving cursors, and executing commands.

The main points:

• Claude can now autonomously navigate computer interfaces, performing complex tasks across multiple applications and web sites.

• Anthropic said it taught the model ‘general computer skills’ as a substitute of making a standalone tool, helping it operate more like a human.

• The upgraded Sonnet 3.5 significantly improves coding and power use, outperforming other models (including o1-preview) on key benchmarks.

• A brand new Haiku 3.5 model matches the capabilities of previous high-end models at lower cost and better speed.

• Anthropic highlighted that computer use continues to be imperfect (including some hilarious examples), encouraging testing on low-risk tasks until skills improve.

Why it matters: While many hoped for Opus 3.5, Anthropic’s Sonnet and Haiku upgrades pack a serious punch. Plus, with the brand new computer use embedded right into its foundation models, Anthropic just sent a warning shot to tons of automation startups—even when the capabilities aren’t earth-shattering… yet.

TOGETHER WITH ASSEMBLYAI

 🗣️ Conversational AI that talks the talk

The Rundown: AssemblyAI is revolutionizing Speech AI with best-in-class accuracy and speed, empowering you to construct the following generation of voice-enabled products.

AssemblyAI delivers:

• Top-tier accuracy rates reaching 95%

• Significantly reduced hallucinations — as much as 30% fewer than industry leaders

• Blazing fast conversion with 63 minutes of audio processed in 35 seconds

• Hassle-free, code-free updates for continuous improvement

Put our API to the test. Start constructing without cost today.

GENMO

🎥 Genmo drops open-source AI video model

The Rundown: AI startup Genmo just launched Mochi 1, a brand new open-source video generation model that claims to rival closed competitors like Runway, Pika, and Kling — while being freely available to developers and researchers.

The main points:

• Mochi is built on a brand new 10B parameter architecture called AsymmDiT, making it the most important open-source video generation model ever released.

• The model focuses heavily on motion quality and prompt adherence, generating 480p videos at 30fps for as much as 5.4 seconds.

• Mochi surpassed top models like Kling, Runway Gen-3, Luma’s Dream Machine, and Pika in motion quality and prompt adherence during testing.

• A better-definition version, Mochi 1 HD, with 720p support and image-to-video capabilities, is planned for release later this yr.

• Genmo also announced that it secured $28.4M in Series A funding, with Mochi-1 being the corporate’s first step toward constructing ‘world simulators.’

Why it matters: Open-source AI video is officially competing with the highest of the market. Genmo’s Mochi is an especially impressive release that showcases how competitive the video generation landscape is about to turn out to be — especially with the key dominos (Sora, Midjourney?) still to return.

AI TRAINING

🎙️ Master public speaking with ChatGPT

The Rundown: ChatGPT’s Voice mode might be customized to simulate a live audience, providing real-time feedback and follow-up inquiries to improve your public speaking skills.

Step-by-step:

1. Download the ChatGPT app and access Custom Instructions in settings.

2. Set ChatGPT to reply with “mhm” during your speech until you say “Done”.

3. Start a brand new chat, activate voice mode, and supply the practice prompt.

4. Deliver your speech section by section, saying “Done” after each part.

Pro tip: Use ChatGPT’s custom instructions to make sure it doesn’t interrupt you or provide unnecessarily long responses.

PRESENTED BY SONAR

🛡️ Navigate AI code risks with confidence

The Rundown: Sonar’s AI capabilities improve the standard of each AI-generated line of code — helping you navigate the brand new risks emerging from the LLM developer landscape.

Take a look at Sonar’s guide to seek out:

• A deep dive into the OWASP LLM Top 10 and its implications

• Strategies to detect and forestall security flaws in AI-generated code

• The vital link between code quality and robust security measures

Download your free guide and stay ahead of emerging AI security threats.

IDEOGRAM

🎨 Ideogram debuts AI Canvas workspace

The Rundown: Ideogram just unveiled a brand new AI-powered workspace called Canvas, introducing advanced tools like Magic Fill and Extend to mix image editing and generation for brand spanking new creative workflows.

The main points:

• Canvas provides an countless digital board on which users can generate, organize, and seamlessly mix AI-generated and uploaded images.

• Magic Fill allows precise editing of chosen image areas, enabling tasks like object substitute, text addition, and background alteration.

• The Extend feature expands images beyond their original dimensions while maintaining style consistency, even with text.

• Ideogram also features an API, allowing developers to include the brand new features into their very own applications

Why it matters: The design industry is not any stranger to AI tools (Photoshop, Canva) — but Ideogram’s latest release looks like the precise variety of fastball that AI and design novices can really make magic with. The examples shown also illuminate how drastically creative workflows are changing within the AI era.

NEW TOOLS & JOBS

Trending AI Tools

• ⚙️ Softr for Notion – Turn Notion databases into portals and apps

• 📊 CapGo AI – AI-powered spreadsheet for market research and lead enrichment

• 📸 Pixyer – AI background generator for skilled product photos

• 💸 Hero – Use AI to scan, price, and list your stuff in seconds

• 💻 AIxBlock – Comprehensive platform to productize AI models with decentralized computing resources

Latest AI Job Opportunities

• 💻 Walmart – Senior, Software Engineer

• 🏗️ Palantir Technologies – Production Infrastructure, Product Manager

• 📊 Mistral AI – Data Quality Specialist, AI Tutor (Fixed term)

• 📞 Glean – Senior Manager, Customer References

• 🛡️ Coreweave – Senior Governance, Risk & Compliance Analyst

QUICK HITS

Runway debuted Act-One, a brand new feature that generates expressive character performances from a single video and image without motion capture or rigging.

Stability AI released Stable Diffusion 3.5, featuring Large and Large-Turbo models that improve customization, efficiency, and variety of outputs.

Cohere enhanced its Embed 3 model with multimodal capabilities, enabling enterprises to perform RAG-style searches across text and image content.

Chipotle launched a brand new conversational AI hiring platform called ‘Ava Cado,’ which the restaurant says can speed up the hiring process by as much as 75%.

Asana introduced AI Studio, a no-code platform for teams to design and deploy AI agents to automate business workflows.

Canva unveiled Dream Lab, a brand new image generator powered by Leonardo AI — alongside a series of recent AI features added to the platform’s Visual Suite.

Inflection AI launched Agentic Workflows, enabling its enterprise systems to take trusted actions for various business use cases.

THAT’S A WRAP

See you soon,

Rowan, Joey, Zach, and Alvaro—aka The Rundown Team