📱 Humane replacement

🧵 In today's edition:

• 📱 Humane replacement

• 🎨 xAI’s multimodal model

• 🎥 Adobe wants to take the reigns

• 🛠️ AI Tools to Check Out

• 🤑 AI Fundraising News

• 🐦 Tweet Post of the Day

📱 Humane replacement

Let’s keep it real - the feedback on Humane’s Pin has not been good. 

• It hallucinates.

• The laser vision is blurry. 

• It takes ages to answer questions that it doesn’t have direct answers to.

• It overheats.

• It doesn’t connect to your phone (I realize that’s the point, but I highly doubt a world where we no longer use our phones).

• And to top it off, it costs $700 with a $24/mo OpenAI subscription (can’t really use the Pin without it).

But… don’t worry.

Dan Siroker's Limitless pendant, the latest AI wearable, might be a solid replacement. 

Here’s the tea:

• What is it? Limitless is a multi-platform application available on the web, macOS, and Windows, and as a wearable device. It's powered by everything you've encountered, spoken, or listened to.

• Specifications and features

  • It enables users to summarize and notarize all audio, accessible through the web app.

  • It boasts a 100-hour battery life.

  • It is equipped with multiple microphones for recording high-quality audio, even in crowded environments.

  • It utilizes Bluetooth and Wi-Fi for seamless syncing.

  • Mobile applications are in development.

• Privacy

  • It features a "consent mode," ensuring the device only records those who have verbally agreed to it.

  • User data is highly protected through their Confidential Cloud, which makes it inaccessible to the government and Limitless without your permission.

• Pricing

  • The pendant is priced at $99 -  immense value for money if you ask me…

  • There is an optional subscription plan costing $19.99 per month, billed annually, for unlimited AI capabilities.

P.S. Check out our podcast episode with Dan Siroker here!

Read more: The Verge | Dan Siroker

🎨 xAI’s multimodal model

Elon Musk's AI company, xAI, unveiled its first multimodal model, Grok-1.5V, capable of understanding text and images, documents, diagrams, charts, screenshots, and photos. 

Here's a quick summary:

• xAI asserts that Grok-1.5V competes with top multimodal models in various areas, such as multi-disciplinary reasoning and visual data understanding.

• The company demonstrated several examples of Grok-1.5V's abilities, including converting handwritten flowcharts to code, generating stories from children's drawings, explaining memes, and inspecting images for issues like wood rot.

• In performance tests against models like GPT-4V and Claude, xAI claims that Grok-1.5V outperforms its rivals on the newly created RealWorldQA benchmark. This benchmark evaluates spatial understanding using real-world images.

• xAI is open-sourcing this benchmark dataset, which includes over 700 images with associated questions and answers.

Overall: the launch of Grok-1.5V comes just weeks after xAI released its updated Grok-1.5 chatbot model and made the underlying Grok AI system open source.

While xAI has faced some controversies around potential misuse, it continues rapidly advancing its "beneficial AI" capabilities, promising significant multimodal upgrades in the coming months.

Read more: Venture Beat

🎥 Adobe wants to take the reigns

Adobe is determined to make big moves this year, starting with introducing GenAI video tools to Premiere Pro. Its proprietary Firefly video model will power these tools and may also integrate with third-party AI providers such as Runway, Pika Labs, and OpenAI's Sora model.

As a quick primer, the Firefly video tools enable users to lengthen video clips, add or remove objects using text prompts (akin to Photoshop's Generative Fill), and create b-roll footage.

Though Adobe's own generative video model exhibited some noticeable issues when removing objects, the company is smartly exploring collaborations with other AI firms to offer users more choices. Incorporating standouts like OpenAI's Sora could significantly enhance the generative video capabilities within Premiere Pro.

This move could benefit all parties involved: 

• Adobe can provide cutting-edge AI video generation by leveraging external models.

• AI companies can expose their tools to Adobe's extensive user base of creative professionals.

Adobe also intends to use its Content Credentials labeling to indicate which AI models were used on generated clips.

Overall: By integrating its own and third-party GenAI video tools directly into Premiere Pro, Adobe can retain users and position its software as a central hub for advanced AI video editing and creation capabilities.

Read more: The Verge

🛠️ AI Tools to Check Out

• PlayPlay: Describe your video needs in a sentence, and PlayPlay’s AI Video Assistant will combine text, media, and music into a compelling on-brand video. Check it out!

• Modelize AI: An all-in-one AI workflow tool with thousands of expert-built workflows to assist with research, analysis, content writing, image generation, and more. Use on web, mobile, and Chrome browsers. Check it out!

• GoSearch: Search across work apps with multimodal-AI enterprise search. Check it out!

_{Have anything you’d like to share with over 40k AI enthusiasts? Sponsor Not A Bot}

🤑 AI Fundraising News

• Lawhive raises £9.5M/$11.8M in Seed funding to provide an online legal platform that combines a tech platform with a regulated law firm.

• HumanX raises $6M in funding to serve as an incubator for making AI useful to organizations worldwide through its proprietary approach.

🐦 Tweet Post of the Day

Awesome rundown 👇️