Grok of X revolutionizes social with AI: new frontiers or risks?

X has announced the release of internal code for its AI chatbot called “Grok.” This chatbot enables X Premium+ users to get sarcastic and explosive answers to questions, based on X’s vast corpus of real-time updates. This article will explore the details of Grok LLM and the impact it could have on social media platforms.

What is Grok LLM?

Grok LLM is a large-scale language model developed by X. With its 314 billion parameters, Grok LLM was trained from the ground up by xAI. It is important to note that this model has not been optimized for specific applications such as dialog, but is a starting point for further development. X decided to release the source code and weighting data of Grok LLM as part of its commitment to transparency and openness.

According to xAI, the release of the source code of Grok LLM will allow for the identification of bias and exploration of the operation of the system by third parties. However, it is important to note that Grok LLM has been trained using X’s posts, which could lead to misleading, offensive, or conspiracy theory-based responses that are harmful.

X’s commitment to AI development.

X founder Elon Musk has expressed concern about the use of generative AI in social media applications. He has criticized chatbots from other companies, such as OpenAI’s ChatGPT, Google’s Gemini, and Meta’s Llama code, calling them “too smart” to produce accurate responses. According to Musk, this could pose a risk to humanity.

Musk ‘s vision is to promote his own AI chatbot, which would be neutral and unaffected by bias. However, given that Grok LLM is trained on X’s posts, the responses produced are likely to be far from an “awake” point of view.

Impact of X’s policies on content

X has adopted a “free speech without reaching” policy that allows many offensive and harmful posts to be displayed within the app. However, if such posts violate X’s policies, their visibility is reduced. Only illegal content is removed completely.

Considering that theLLM is trained on the entire corpus of X posts, it is possible that the chatbot will produce misleading, offensive, or incorrect responses based on conspiracy theories and harmful stereotypes. This raises concerns about the spread of misinformation and conspiracy theories within social media platforms.

Limited accessibility to Grok LLM

Currently, only a very small percentage of X users have access to Grok LLM. Less than 1% of users have subscribed to X Premium, and Grok LLM is only available in the “Premium+” package, which costs twice as much as the basic subscription. This limits the amount of information available about the actual responses produced by Grok LLM.

However, it is reasonable to assume that Grok LLM is susceptible to the same “woke” responses as other AI tools, depending on the questions asked. At the same time, it is likely to produce misleading responses based on X’s posts as input.

The future of social media and AI bots

The introduction of AI b ots such as Grok LLM raises important questions about the future of social media. Currently, AI bots can be used to create posts on platforms such as Facebook, LinkedIn, and Snapchat. This allows one to create an alternate persona that is completely powered by AI tools, but does not represent the user’s true identity.

This could lead to a situation where more and more content on social media is generated by bots and human interaction is eliminated. This leads to a shift away from the main purpose of social media as tools for social connection.

In recent years, there has been a significant decline in the number of people posting on social media, with an increased focus on private conversations. This trend was triggered by TikTok, which shifted the emphasis from followers to AI recommendations based on user activity. As a result, social media platforms are increasingly becoming entertainment platforms rather than connection tools.