😺 New Claude, Who Dis?

PLUS: New AI drops from Canva, Runway, Stability, and more...

Welcome, humans.

Sponsorship slots in The Neuron are filling up very quickly, but we still have a handful of placements in 2024. First come, first served.

Quick form, few details, done. You'll get your product in front of 460,000+ of tech's most curious minds.

Just fill out this form and we’ll reach out to follow up.

Here’s what you need to know about AI today:

Claude released two new models and the ability to use computers.
Runway demo’d a new video-to-animation tool.
Asana released no-code AI workflow builder.
Qualcomm and Google announced new custom car voice AI.

Anthropic just dropped two new Claude models…oh, and Claude can now use your computer.

Claude | Computer use for automating operations

Just when we thought the AI boom was over (joking of course), something mind blowing just dropped: AI that can use a computer like you or me. Uh oh…

Actually, Anthropic just released three major updates: an upgraded Claude 3.5 Sonnet, a new Claude 3.5 Haiku, and yes, “Computer use” that lets Claude interact with our screens like humans do.

Here are the deets:

The upgraded Sonnet is available now, with Haiku coming later this month.
Computer Use is launching first in Beta via API (docs here, demo here).
Both US and UK AI Safety Institutes pre-tested these models…so ya boy’s safe!

Early reports suggest the upgraded Sonnet is crushing coding benchmarks on Livebench, outperforming ChatGPT-4o and ChatGPTo1-mini.

Not only that—Sonnet can write. And boy, can it… we messed around with it today, and even dialoguing with it, we were floored by how fluid and natural sounding it was to work with.

The real headline here is Computer use, though, which can take over your screen and do actions on your computer, like filling out a form. Several companies are already putting computer use to work, including Asana, Canva, DoorDash, and Replit.

Someone’s already used Computer Use to solve the Wordle in 3 guesses. Fun fact: I ALSO got the Wordle in 3 guesses yesterday. Take that, AGI.

Demo of computer use.

Ethan Mollick got early access to test out computer use, and his experience shows both the potential and limitations. For example, he had Claude create an entire high school lesson plan for The Great Gatsby by itself:

Claude downloaded the book.
Researched lesson plans online.
Built a spreadsheet.
And tied everything to Common Core standards (with a lil human help).

Here’s the catch: Computer Use is reportedly very expensive to run, and can be stubborn when it makes mistakes. When Mollick had Claude play a paperclip-making game, it persistently stuck to an incorrect pricing strategy even after being corrected multiple times.

And fun fact: in the middle of filming Anthropic’s launch demo, Claude decided to take a break and check out pictures of Yellowstone. AI—they’re just like us!

So why is Anthropic focused so much on coding? This post says it all:

FROM OUR PARTNERS

How to Build a Sales Dashboard in Less than 15 Seconds

What if you could save tens of hours on building sales dashboards, in just a few clicks? No, erase that, in just a few seconds?

Zebra AI is here to do just that.

Part LLM, part machine learning-powered financial expert, Zebra AI is an automated data visualization and analysis tool ready to build your sales dashboard and provide you with the specific answers you need—in less than 15 seconds.

All you have to do is:

Upload your data.
Let the AI do its job.
Ask your questions.

…Then share your Zebra AI stories with peers so they can pick up where you left off.

Made you curious? You can try Zebra AI for free here.

Around the Horn.

Introducing Act-One | Runway

Runway demo’d Act-One, a new tool that transforms basic video performances into expressive character animations (see above).
Asana announced AI Studio, a no-code builder that allows teams to design and deploy agents within workflows across their work management program (demo video).
Qualcomm announced that it’s working with Google to create an Android Automotive OS that lets automakers build their own custom voice assistants.
Microsoft + OpenAI dedicated $10M to a new fellowship program to support local journalism—a series of five outlets will get $500K to hire an AI specialist for two years to build new AI tools.
SAG-AFTRA, UMG, Penguin Random House, and 10.5K+ others have signed a statement protesting AI companies' unlicensed use of their work to train AI models.

Treats To Try.

Droptober 2024 | The latest Canva launches, designed for your greatest work

Canva released Dream Lab, a new image generator built with Leonardo.ai’s Phoenix model (which Canva owns), along with new features to auto-complete writing tasks, add animated captions to videos, create interactive charts, and collaborate on whiteboards with real-time reactions and summaries.
Stability released Stable Diffusion 3.5, a new version of its open-source image generator that’s easier to customize with improved prompt accuracy and image quality along—available in 3.5 Large and 3.5 Large Turbo (Hugging Face, Github).
TollBit helps you monetize your website's content by charging AI bots for access to your data while providing analytics on how they use it (raised $24M).
Hero lets you instantly scan and list items for resale by automatically generating prices, titles, and descriptions from your phone's camera.
Pixyer turns your basic product photos into professional studio-quality shots by automatically removing backgrounds and generating new ones.
Socket protects your software from dangerous code hidden in third-party tools you use to build apps—here’s the Github app to install (raised $40M).
Highlight turns your meetings, emails, and PDFs into instant summaries you can share anywhere (raised $10M).

See our top 51 AI Tools for Business here!

FROM OUR PARTNERS

Learn how NVIDIA, Databricks, Twilio, HP, and ServiceNow get their GenAI apps into production.

Don't miss GenAI Productionize 2.0 for GenAI best practices, including:

How to design an enterprise-scale GenAI stack.
Techniques for AI governance, evaluation, and observability.
Proven strategies for getting GenAI apps into production.

Free Registration.

Where do you #Neuron?!

Submit where you #Neuron here for a chance to be featured in our newsletter next week!

Ken from Tampa, FL: Ken starting his morning with The Neuron!

Ahmad from New Delhi, India: Coffee in hand, Neuron on the screen—just another day.

Craig, PIppa, Argie, and Smittens from Boise, Idaho: An ai reading up an ai, on the wing of a jet, with cats over Bogus Basin, Idaho.

Please everyone, take notes: THIS is how you should #Neuron!

This could be you, but you didn’t send us where you #Neuron…

Our fave reactions from the thread: “Put Rob out of his misery“, “You will live to see manmade horrors beyond comprehension”, and “It speaks calmly, but its eyes plead for an end to this waking death…”