It’s hard to keep track of the many generative AI tools getting rolled out today.
But the latest on this front has to be Elon Musk’s xAI which just converted its large language model Grok into an open-source endeavor.
The fact that we’ve got chatbots running around garnering responses to questions and prompts thrown in their direction instead of real human replies is concerning. But that’s now the norm and we’ve started to accept it.
More discussions on this front spoke about how Musk’s company just rolled out internal codes for the Grok AI chatbot that gives Premium + users a chance to get funny and even make edgier replies to any questions, all depending on the app’s growing corpus for real-time content.
The fact that Grok’s internal codes and entire framework architecture entail a massive 314 billion parameters and now that’s up for the public to view is certainly mindblowing.
This release seems to be an integral part of the company’s drive to be more transparent about how its systems function and also to facilitate filtering out any kind of biases arising while exploring different systems under the ownership of third parties.
The rollout happens to be a part of the platform’s commitment to getting more publicity as an open-source endeavor in the world of AI. It all makes sense because Musk is known for criticizing other chatbots so making sure it is free from the same kind of faults that he sees in others means getting more flexible about the work.
There have already been so many chatbots in the market that continue to be slammed for their poor performances and errors like Google’s Gemini and OpenAI’s ChatGPT. Let’s not forget Llama by Meta as they were outlined as ‘massively woke’ to give out the right replies.
Musk has spoken in the past about how they are a serious threat to humanity and when you’ve got massive dangers like this posing as a threat for your own company, you better make sure you pull out all the stops to ensure that doesn’t happen.
Musk calls his chatbot non-biased in the world of AI and he also suggests that all training was done on posts from the X platform.
This matter is very interesting, to say the least. The platform already has a no-reach, freedom-of-speech rule that it likes to follow. For instance, it’s now leaving some harmful content on the app and limiting its reach if it feels is in violation of the entire policy. When it breaks the law, X vows to delete the post but if not, then users can still view it.
Musk feels that OpenAI has been tricking the masses for so long by being an enterprise that is called OpenAI but is not transparent about its functions to the world. Now the question that some users have on their minds is whether making Grok open-sourced would classify it as a contributor to the entire community for AI development or not.
Remember, Grok answers like other chatbots but in a sassier manner with plenty of access to data under the ownership of Twitter that keeps it separate from all others.
Experts are applauding the decision of Musk to make it open source as they feel the more access that the developer community has in this regard, the better. But the problem here lies with the concept of being open and modern and very functional as the best LLM.
But other critics are having issues with considering Grok as a modern-day LLM that’s open because that terminology tends to be abused to a great extent online. Hence, we’re excited to see what the tech world has to say about it as the word spreads.
Read next: Apple Eyes Google's Gemini for iPhone Intelligence Boost
