GrokAI is accused of stealing OpenAI's code (Express Photo)Elon Musk’s GrokAI, a ChatGPT and Bard rival is finally rolling out to X Premium+ subscribers and people have already started to compare it with GPT-4. ChatGPT has posted a screenshot of GrokAI on X, stating “it goes against OpenAI’s use case policy,” and ChatGPT writes ‘we have a lot in common.”
we have a lot in common pic.twitter.com/jxRsJg8qPr
— ChatGPT (@ChatGPTapp) December 9, 2023
Elon Musk has commented on this post, stating “Well, son, since you scraped all the data from this platform for your training, you ought to know.”
Several users have also reported instances of GrokAI mentioning it as a model of OpenAI. According to Igor Babuschkin, xAI’s co-founder, the web is currently full of ChatGPT outputs, and Grok is trained on a large amount of web data, which is causing this issue. He also clarifies that no OpenAI code was used to train Grok, and this will be resolved in the future version of GrokAI.
The issue here is that the web is full of ChatGPT outputs, so we accidentally picked up some of them when we trained Grok on a large amount of web data. This was a huge surprise to us when we first noticed it. For what it’s worth, the issue is very rare and now that we’re aware…
— Igor Babuschkin (@ibab) December 9, 2023
When compared to other prominent large language models, GrokAI has been trained to be funny and witty in its responses. According to the company, the earlier version of GrokAI already outperforms the competition and will soon gain capabilities like image recognition and image generation. Unlike most generative AI models, Grok can access real-time information from X and offer up-to-date responses.
Grok-1 has already surpassed the performance of models like GPT-3.5 and LLaMa 2. However, it is still not as capable of GPT-4 or the Palm 2. Right now, GrokAI is available exclusively to X Premium+ subscribers and the company is currently improving and adding new features to the generative AI model.