Anthropic Unveils Claude 3, Claims It Surpasses GPT-4 🔥

Welcome to another edition of Horizon AI,

Anthropic challenges OpenAI and Google with the release of its most powerful chatbot yet, said to surpass both GPT-4 and Gemini Ultra. Meanwhile, OpenAI executives respond to Elon Musk.

Let's jump right in!

Read Time: 3.5 min

Here's what's new today in the Horizon AI

• A Chart to Kickstart Your Day 📈: AI boom propels Nvidia to become the third US stock to close with a valuation above $2tn

• Anthropic Debuts Its Most Powerful Chatbot Yet🔥

• OpenAI Executives Reject Elon Musk's Claims

• AI Tutorial: Learn how to use the Canva AI features

• AI Tools to check out

• The Latest in AI and Tech 💡

A chart to kickstart your day

AI boom propels Nvidia to become the third US stock to close with a valuation above $2tn

• A 4 percent gain saw Nvidia’s shares close at $2.06tn, putting it behind only Microsoft and Apple at the top of Wall Street valuations, at respectively $3.09tn and $2.77tn.

Source

AI News

ANTHROPIC

Anthropic Debuts Its Most Powerful Chatbot Yet 🔥

AI startup Anthropic, backed by Google and Amazon, unveiled its latest offering - a suite of three state-of-the-art language models called Claude 3.

The Claude 3 family encompasses three models of increasing capability: Claude 3 Haiku, Claude 3 Sonnet, and the flagship Claude 3 Opus.
Key enhancements across the suite include:

• Improved proficiency in analysis, forecasting, nuanced content generation, code authoring, and multilingual conversations

• Sophisticated vision capabilities on par with industry-leading models for processing images, charts, diagrams, and other visual data formats

• An initial 200K token context window at launch, with potential for extending to over 1 million tokens for select enterprise users.

• Superior skills in comprehending and executing complex, multi-step instructions

According to Anthropic, benchmarking tests indicate the high-end Claude 3 Opus outperforms rival AI models such as OpenAI's GPT-4 and Google's Gemini 1.0 Ultra.

These are some bold claims, industry experts are surely rushing to confirm if they hold true.

You can see the official announcement and more details about these new models here.

OPENAI

OpenAI Executives Reject Elon Musk's Claims

In an internal memo, OpenAI's Chief Strategy Officer Jason Kwon directly addressed and disputed several of Musk's claims. Here are the key points:

• Not AGI Yet: Kwon maintains that GPT-4 is not a true artificial general intelligence system. An AGI, Kwon argues, would be highly autonomous and capable of devising novel solutions – something GPT-4 can't do.

• Committed to Benefiting Humanity: Kwon rejected the notion that OpenAI has abandoned its mission of benefiting humanity. He says that their challenge is both to build AGI and to make sure its impact is maximally positive.

• Not a Microsoft Subsidiary: Kwon emphasizes OpenAI's independence from Microsoft, stating it sets its own research directions, product lineup, and competes with Microsoft's offerings like Copilot.

In a follow-up statement, OpenAI CEO Sam Altman acknowledged that the year ahead would be challenging, with more "attacks" likely coming their way.

AI Tutorial

Learn how to use the Canva AI features

AI has made Canva more powerful than ever. Here are 15 Canva AI features that may surprise you:

1. Magic Grab

It helps you redesign images, just like a Canva template. Use it to select and separate the subject of your photo so that you can edit, reposition, or resize it.

2. AI-powered image generation

Bring your ideas into reality by from just one text prompt on Canva using it on Magic Media

3. Magic Write

It can take your text from zero to hero in seconds

4. Data and Privacy Risk

You are in control if you want to share your data with Canva or not

5. Images as unique as you

You can check the style that you like and go for an image style accordingly.

6. AI Presentation Maker

Click “Create a design” on the homepage and select Presentation. Once in the editor, go to the Design tab in the left panel. Describe your ideal presentation in the search bar.

7. Video editing

Pick your clips and images, enter the prompt. Use Magic animate to add smooth transitions and animations to your videos. Magic Switch can repurpose your video content for multiple platforms. You can create videos from text prompts too.

8. Magic Switch

Swap formats, languages, and dimensions in seconds.

9. Magic Expand

Seamlessly extend an image in any direction for the perfect shot. Fix awkward framing, save zoomed-in images, or turn a vertical shot into a horizontal one in seconds.

10. Magic Edit

Reimagine your image and add to, replace, or edit it with a short written prompt

11. Magic Eraser

Remove unwanted details and distractions from your photos.

12. Background remover

Remove backgrounds from images and videos easily.

13. Magic Animate

Apply the perfect animations and transitions to your entire design automatically.

14. Magic Morph

Transform words and shapes with a simple written prompt. Magic Morph applies effects to text and shapes to quickly create patterns and textures.

15. Integrate other AI apps

Here are some of the AI apps you can integrate into your Canva workflow

AI Tools to check out

🎙 Transistor: Generate incredibly accurate transcripts for your podcast episodes.

✨ Synthesys X: Create your own versions of any image you find online.

📖 Dataku: Seamlessly extract valuable insights from documents and texts.

📝 Survicate: Create surveys in seconds and analyze results without drowning in data.

📊 Basedash: AI-generated interface to visualize, edit, and explore your data.

The latest in AI and Tech

India instructs tech companies to seek approval before launching new ‘unreliable ’ AI tools

India has mandated that tech companies must obtain government approval before releasing AI tools deemed "unreliable" or still in trial phases to the Indian public. The IT ministry's recent advisory states that such tools must be explicitly permitted by the Government of India prior to being made available to Indian internet users.

Amazon's $1 Billion Bet on Robotics and AI

Amazon plans to invest $1 billion in startups combining artificial intelligence and robotics technologies. The company believes that “generative AI holds a lot of promise for robotics and automation” and is an area they are going to focus on this year.

Researchers create AI worm that infects users via AI-enabled email clients

Researchers have developed a generative AI worm called Morris II with the ability to steal data and send spam emails through two methods: text-based self-replicating prompts and image-embedded self-replicating prompts that trigger email assistants to propagate malicious content.

California officials give Waymo the green light to expand robotaxis

California has given Waymo the green light to expand its self-driving robotaxi operations onto highways in parts of Los Angeles and the Bay Area. While the company plans to take a "careful and incremental" approach, this move brings us a little closer to a future where autonomous vehicles seamlessly navigate city streets and open roads.

OpenAI’s new Sora videos go viral

OpenAI's newest batch of Sora videos has once again captured the internet's attention, showcasing the impressive potential of its generative video model. These viral clips have sparked a wave of excitement about the future possibilities for AI-created content.

That’s a wrap!

Thanks for sticking with us to the end! Let’s stay connected on LinkedIn and Twitter.

Not subscribed yet? Sign up here and send it to a colleague or friend!

See you in our next edition!

Gina 👩🏻‍💻