Elon Musk Promotes xAI's Grok Amid Google Gemini Accuracy Concerns

Elon Musk, cashing in on the recent controversy surrounding Google's Gemini AI, has emphasised the crucial need for constant truth-seeking in the development of AI technology.

Taking to his social media platform, X, Musk recently shared a post highlighting the importance of Grok, an AI-powered chatbot unveiled by his AI company xAI in November 2023.

xAI's website describes Grok as a "Conversational AI for understanding the universe". Musk, known for his commitment to truth-seeking AI, admitted that Grok is still under development but promises rapid improvement. He confirmed that version 1.5 will be released in just two weeks.

The 52-year-old business magnate went on to emphasise the importance of relentlessly seeking truth in AI development, despite potential criticism.

In a separate post, Musk stated that the "woke mind" is killing Western Civilization. He pointed out that Google, Facebook, Wikipedia, and Instagram do the same thing with their search results.

In response to an X user who asked Musk whether it is too late to build a "woke-free" search engine and if search engines are likely to become futile with advancements in AI, Musk simply replied, "Grok". However, another user pointed out that Grok isn't available to everyone.

To those unaware, xAI's early access program is currently limited to X Premium+ subscribers, who pay a £16 monthly subscription fee. Musk's statement about his AI chatbot Grok follows reports of issues with Google's Gemini AI, including concerns about image accuracy and potential bias.

Google responds to criticism

Google has paused the image generation feature of the Gemini AI, acknowledging issues with historical depictions in its AI. Some users on social media raised concerns about the AI tool's portrayal of historical figures, including the U.S. Founding Fathers, who were shown as people of colour.

In an X post, the search giant explained that the AI feature is capable of generating a wide range of people, further noting that's a good thing since people around the world use it. On the downside, the company pointed out that the software feature is "missing the mark here".

Nevertheless, Google assured that it is "working to improve these kind of depictions immediately". In its updated statement, the company said it will pause Gemini's ability to generate images of people and will re-release an "improved" version soon.

To recap, the Sundar Pichai-led tech company launched its image generator tool at the start of February through Gemini, formerly called Bard. According to a CNBC reporter, Gemini did not generate any images at all on Thursday morning when tested.

Jack Krawczyk, a senior director of products for Gemini at Google said the company takes "representation and bias seriously" and will "continue to do this for open ended prompts (images of a person walking a dog are universal!)".

"Historical contexts have more nuance to them and we will further tune to accommodate that," Krawczyk added. As Google faces challenges with Gemini's image generation, OpenAI launched Sora, a new AI model capable of creating video from text prompts.