Musk has announced that Grok will be accessible to all users. This comes after the update to Grok 3 to bring its operation closer to how OpenAI’s ChatGPT and Google’s Gemini work. Musk disclosed that Grok is available for installation as its own dedicated app on Windows and MacOS.
Grok has been the default model that all X users use. However, the AI model will now have the chance to reach demographics outside X and its premium users. Miusk’s announcement brought excitement about XAI’s prospect of competing in the open market.
Grok will also be offered as a standalone application for the MacOS and Windows
— Elon Musk (@elonmusk) February 19, 2025
Grok has stood out because it has access to information in real-time due to its access to X platform data. According to reviews, it has a good sense of humor and performs strongly on many measures.
Grok 2 will be open-source
In addition, Grok 2 will become open-source. This means that developers will be able to dig into its code and customize it for use in entirely new applications. Some interesting AI tools might appear outside of the normal walled gardens.
For instance, smaller companies and independent developers could use it to make custom chatbots and personalized AI assistants.
Musk said, “Our general approach is that we will open-source the last version [of Grok] when the next version is fully out [..] When Grok 3 is mature and stable, which is probably within a few months, then we’ll open source Grok 2.”
However, how open the open-source version of Grok 2 will be remains unclear.
Meta has played with the open-source trend by making parts of its Llama models public since the models first came out. China-based DeepSeek, the new AI star, has also made strides in making AI open-source.
Grok 2 will be updated to Grok 3 to explain X posts
Elon Musk announced that Grok 2 would be updated to Grok 3 soon to explain posts on X. According to the tech giant, the rollout will take a day or two to complete.
Btw, Grok 2 is still being used to explain 𝕏 posts. That will update to Grok 3 in a day or two. https://t.co/iSv9CA3bjt
— Elon Musk (@elonmusk) February 19, 2025
This feature has stood out most with the Grok AI. This is mainly because it changes how people understand and interact with posts. It also gives users the freedom to explore material with confidence and clarity.
Musk’s announcement suggests that Grok 3 will enhance the AI’s ability to interpret and explain posts on the platform. He has been pushing for progress in AI through his company, xAI. Grok is now seen as a competitive option to models made by OpenAI, Google’s DeepMind, and others.
This comes as users continue to compare the new Grok 3 with other competitors. One user said that Grok 3 outperformed its competitors, including R1, O1 Pro, and Sonnet 3.5.
The post read, “Final Verdict: Grok 3 > R1 > O1 pro > Sonnet 3.5. Have done tests all day today, Grok and R1 can do in-context shape rotations with high-dimensional tensors, while O1 and Sonnet need for loops. XAI and DeepSeek have figured something out!”
This supports what the xAI team said during the launch: “Grok 3 is in a league of its own across the board. Even its smaller counterpart, Grok 3-Mini, is reaching the cutting edge compared to other competitors.”
All you need to know to understand which company will win a technology competition is look at the first and second derivatives of the rate of innovation https://t.co/rImcrpzfeY
— Elon Musk (@elonmusk) February 18, 2025
Along with the pre-trained model, xAI also released the Grok 3 Reasoning Model (Beta) and the Grok 3 Mini Reasoning Model. Unlike models that have already been taught, reasoning models are made to think like humans. They take a little longer to respond, but they are much better at answering complicated questions with more depth and accuracy.
The xAI team said, “At xAI, we believe that having the best pre-training model is not enough to create the best AI. The best AI needs to think like a human. If you’re using Grok 3, you may notice improvements almost every day because we are constantly working on enhancing the model. In fact, you might even see changes within 24 hours.”
They added, “You need to consider all possible solutions, self-critique, verify them, backtrack, and think from first principles. This is a crucial skill. We believe that by taking the best pre-trained model and further training it with reinforcement learning, we can enhance its reasoning abilities. This allows the model to improve significantly and scale, not just during training but also during testing.”
Cryptopolitan Academy: Want to grow your money in 2025? Learn how to do it with DeFi in our upcoming webclass. Save Your Spot
