Grok can now "see"

PLUS: What is the definition of "Open source AI"? Now we got one and its something you want to know

In partnership with

I was confident that they’ll add vision ability to our baby Grok from xAI.

And now it can see and analyze whatever you throw at it.

I recently an image of a celebrity from Insta (everyone’s there a celeb). And it exactly shared who that person was because they were also active on 𝕏.

Let me share what we got in this post for you in 3+ mins.

(here’s how Grok identified Elon Musk and roast him)

Read the full post:

Grok can now 👀

Here’s everything you need to know:

Elon Musk’s xAI just rolled out an update that lets Grok handle image understanding, and I think it’s a big step forward.

Now, paying users on X can upload an image and ask Grok questions about it. It’s a feature that’s in its early stages, but Musk promises rapid improvements, and I think the possibilities here are huge.

Here’s why I’m excited about it:

  • Multimodal Capabilities: Grok can now analyze images. Whether it’s understanding the joke in a meme or breaking down complex visuals, this new functionality adds serious versatility.

  • Context-Rich Replies: This feature means we’re no longer limited to text-only responses. Grok’s visual analysis opens up new uses — I can see applications for content creation, real-time trend analysis, and so much more.

  • Building on Grok-2’s Updates: This adds another layer to the Grok-2 model, which introduced FLUX.1 for image generation. Now, Grok can interpret images, not just generate them, keeping it competitive with other multimodal AI tools out there.

  • Future Document Understanding: Musk hinted that document support is next, so PDFs and other files may soon be on the table. I love that xAI is moving fast on this — according to Musk, they’re achieving in months what takes others years.

  • Boosting X’s Premium User Appeal: With the recent launch of tools like Radar for Premium+ subscribers, X is adding real value to its paid tiers. These kinds of features make the platform a lot more compelling, especially if you’re looking for insights into trending topics.

For xAI, these updates could set Grok apart in the crowded multimodal space. I think it’s a smart move that could solidify Grok’s role on X and make it an even more valuable tool for users.

The “official“ definition of Open source AI

ngl OpenAI made the definition a mess.

The official definition:

An open-source AI model must fully disclose its design and training details, enabling others to substantially recreate it. It provides unrestricted rights for anyone to use, modify, and build upon, ensuring transparency, community oversight, and the freedom to innovate on top of the model without requiring permission

Here’s the scoop:

For years, there’s been no clear consensus on what exactly makes an AI model "open source." But as of now, we have one. The Open Source Initiative (OSI), which has long set the standards for open-source software, has just released the Open Source AI Definition (OSAID) 1.0. This long-awaited definition is the result of years of collaboration with academics, tech experts, and policymakers. Its purpose? To standardize what qualifies as truly open-source AI.

  • Clarity for Policymakers & Developers: Regulators and developers need common ground on what makes AI models open source. OSI EVP Stefano Maffulli noted that the OSI worked closely with diverse stakeholders to ensure the definition supports future regulation and fair industry practices.

  • Transparency & Trust: According to the OSAID, an open-source AI model must allow others to "substantially" recreate it. This means the model’s design, training methods, and data sources should be openly accessible to build community trust.

  • Usage Rights for Developers: The OSAID sets out clear user rights — anyone can freely use, modify, and build upon an AI model without needing special permissions, a hallmark of traditional open source.

Yet, this new definition isn’t enforceable. The OSI itself has no authority to force companies to follow it, but it can publicly flag models that claim to be open source yet fall short of the OSAID standards.

This definition is also contentious. Meta, for example, has challenged the OSI’s criteria, especially around its Llama model, which Meta calls “open source” but limits access to platforms with over 700 million users.

So, the definition is this:

  • Transparent Development: An open-source AI model must disclose design and training details that enable others to substantially recreate it. This includes sharing code, training methods, and, ideally, making training data accessible.

  • Freedom of Use and Modification: Under the OSAID, developers should have unrestricted rights to use, alter, and build on the model for any purpose.

  • Community Oversight: By providing transparency and unrestricted usage rights, the OSAID allows the broader AI community to monitor, enhance, and call out misuses of the “open source” term.

While some might argue for more robust licensure requirements, this 1.0 version is an attempt to set a baseline for open-source AI that can adapt as the tech — and its applications — evolve.

This is what it is. 🤷

Writer want to say something and I bet you’ll love it.

Sponsored by WRITER:

Escaping AI POC purgatory: Techniques for enterprise AI engineers

  • From POC to production: practical strategies for bringing AI to production successfully

  • Expert insights: Led by Sam Julien, Director of Developer Relations

  • Real-world examples: Drive business impact with AI

Whenever you’re ready, there are 2 ways we can help you:

  1. Help you promote your product and service to 150k+ engineers, AI enthusiasts, entrepreneurs, creators, and founders. Sponsor us

  2. Help you build a irresistible brand on 𝕏/Twitter and LinkedIn in less than 6 months. We’ve helped 100+ creators from YouTube, entrepreneurs, founders, executives, and people like yourself. Contact us here: [email protected]

Thanks for reading…

That’s a wrap.

typing typewriter GIF

What's on your mind?

Share your best ideas with us at [email protected].

We'll bring your ideas to life. Send them our way, and we'll get to work on making them a reality.

Did you find value in our newsletter today?

Your feedback can help us create better content for you!

Login or Subscribe to participate in polls.

I hope this was useful…

If you want to learn more then visit this website.

Get your brand, product, or service in front of 150,000+ professionals here.

Follow us on 𝕏/Twitter to learn more about AI: