- AI Valley
- Posts
- Claude AI can now control your PC
Claude AI can now control your PC
PLUS: Runway just launched Act-One
Together with
Howdy! It’s Barsee again.
Happy Thursday, AI family, and welcome back to AI Valley.
In today’s edition:
🧠💻 Claude AI can now control your PC
🎥🌀 Animate expressive characters with simple video capture
🤖 Plus trending AI tools, posts, and resources.
Ready, set, go…
TOGETHER WITH FLOW
Tired of slow typing and endless edits? Wispr Flow lets you speak naturally and converts your thoughts into perfectly formatted text, saving you hours. Whether you're crafting AI prompts in ChatGPT, Cursor, or v0, or simply writing emails and messages, Flow adapts to your style, making everything seamless.
Professionals, students, and tech enthusiasts are calling it a game-changer.
Developers love how fast they can interact with AI. Product managers rave about how it turns messy thoughts into clear ideas.And for anyone juggling busy schedules, Flow's accuracy and speed give you more time for what matters.
With advanced voice recognition, auto-edits, and command mode, Flow captures your tone and polishes your words.
Ready to boost your workflow? Try Wispr Flow today and experience smarter, faster communication.
ANTHROPIC
🧠💻 Claude AI can now control your PC
On Tuesday, Anthropic released the upgraded Claude 3.5 Sonnet model and the new Claude 3.5 Haiku, along with a public beta for an experimental "computer use" feature.
What are the updates?
Upgraded Claude 3.5 Sonnet:
Claude 3.5 Sonnet has significantly improved in coding, vision, and reasoning tasks, outperforming GPT-4o and Google’s Gemini models.
On coding, the model boosts its SWE-bench verified score from 33.4% to 49.0%, outperforming major AI systems, including OpenAI's o1-preview and other agentic coding-focused models.
It also improves performance on TAU-benchmark, from 62.6% to 69.2% in the retail domain, and from 36.0% to 46.0% in the more challenging airline domain.
A new "Computer use" API:
It enables the AI model to interact with computers like humans, navigating screens, moving cursors, clicking buttons, and typing text.
Latest Claude 3.5 Haiku model:
It is built for low-latency tasks and precise tool use, making it ideal for user-facing applications, specialized sub-agent tasks, and managing large datasets like inventory records.
How to access it?
Both Claude 3.5 Sonnet and Claude 3.5 Haiku, along with the computer use feature, are available through Anthropic’s API, Amazon Bedrock, and Google Cloud’s Vertex AI.
RUNWAY
🎥🌀 Animate expressive characters with simple video capture
Runway has just introduced Act-One, a new tool that allows you to easily transfer the facial expressions of a person to an AI-generated character.
What makes it unique?
Act-One captures detailed facial movements, from eye shifts to subtle expressions, for more lifelike character performances.
It produces realistic facial animations across multiple camera angles, enhancing versatility for dynamic content creation.
And allows a single actor to perform multiple characters with different body types, using just a standard camera.
Also includes safeguards to prevent unauthorized use with public figures and verifies voice usage rights.
How to access it?
Act-One is already available to Runway users, with full access coming soon, but requires Gen-3 Alpha model credits to use.
Why does it matter?
This release simplifies character animation, allowing creators to achieve complex performance transfers without needing traditional animation tools, opening up new possibilities for storytelling and creativity.
QUICK HITS
Google released SynthID Text, which lets developers watermark and detect text generated by AI models. (link)
SynthID can also watermark videos by embedding a watermark in each frame’s pixels. (link)
Apple releases second wave of 'Apple Intelligence' features with ChatGPT integration via new developer betas. (link)
Asana launches AI Studio, a no-code tool for designing AI agents. (link)
Over 11,500 creative professionals, including notable figures like Julianne Moore, James Patterson, and Thom Yorke, have signed an open letter demanding a ban training AI without permission. (link)
ElevenLabs has introduced Voice Design, a new AI tool that generates a unique voice from a text prompt alone. (link)
Stability AI releases Stable Diffusion 3.5 open weights models with quicker image generation. Up to 1MP images can be generated using Stable Diffusion 3.5 Large. (link)
AI-related seed funding slows, but valuations remain high. First drop in AI seed funding since ChatGPT’s launch in Nov. 2022. (link)
Google AI Studio's new Compare Mode allows users to evaluate different Gemini models side-by-side, making selecting the best model for their use case easier. (link)
TRENDING TOOLS
Clueso > Create studio-quality videos and step-by-step guides to explain any product or workflow in minutes. (link) *
Paperguide > Discover, read, write, and manage research with ease. (link)
Chance: Visual Intelligence > AI-powered visual search engine, search by seeing with GPT. (link)
CapGo.AI > AI Spreadsheet for market research, and lead enrichment. (link)
Perplexity Pro Search > Break down complex queries into simple steps for more accurate and detailed results. (link)
COOL FINDINGS / RESOURCES
DAILY DOSE OF CONTENTS
1/ How an AI Bot became a Crypto Millionaire.
2/ Clone introduces Torso, a bimanual android actuated with artificial muscles. With 25 DoF hands and anthropomorphic shoulders, Torso is a truly one-of-a-kind creation. Human-level androids are almost here.
Introducing Torso, a bimanual android actuated with artificial muscles.
— Clone (@clonerobotics)
8:10 PM • Oct 23, 2024
3/ Just a day after Anthropic dropped computer use API, here is a very cool & creative use case!
AI agents will fundamentally reshape how tasks approached & how information is accessed, this will usher a new era of unprecedented autonomy and efficiency.
Anthropic computer use API + iPhone mirroring to a Mac = AI controlled phone.
Watch Claude control my phone and successfully look up stats in my Sports app.
I even got it to play a game in the Chess app against another AI - pretty crazy.
And this is the worst it’ll ever be.
— Mckay Wrigley (@mckaywrigley)
5:47 PM • Oct 23, 2024
THAT’S ALL FOR TODAY
That’s all for today’s issue, folks.
💡 Help me get better and suggest new ideas at [email protected] or @heyBarsee
👍️ Like what you see? Subscribe here
Thanks for being here.
HOW WAS TODAY'S NEWSLETTER |
REACH 100K+ READERS
Acquire new customers and drive revenue by partnering with us
Sponsor AI Valley and reach over 100,000+ entrepreneurs, founders, software engineers, investors, etc.
If you’re interested in sponsoring us, email [email protected] with the subject “AI Valley Ads”.