What’s Grok? Elon Musk’s Challenging ChatGPT Rival

Elon Musk has boldly entered the already bustling AI arena with Grok, a conversational AI meticulously designed to pose a significant challenge to the likes of ChatGPT and Midjourney. This chatbot stands out by offering a more pronounced “sense of humor” compared to other AIs, as Musk humorously put it, meaning fewer content restrictions and a bit more swearing.

Contents

  • What is Grok?
  • Where did Grok come from?
  • What can Grok do?

It’s all accessible and trained on the X social media platform, as you might have guessed. Here’s all you need to know about it.

Recommended Videos

What is Grok?


Jon Martindale / Digital Trends

Grok is a cutting-edge generative AI chatbot meticulously developed by xAI, Musk’s pioneering artificial intelligence startup. It was originally crafted atop what is known as the Grok-1 model. The Grok-1 model was painstakingly developed over several months on a cluster of “tens of thousands” of GPUs and leverages the Flux.1 model by Black Forest Labs for its remarkable image-generation capabilities.

The model is trained using a diverse mix of web data (with a knowledge cutoff of Q3 2023) and X user data. Currently, the Grok chatbot is powered by the Grok-2 model, which made its debut in August 2024.

The chatbot first made its appearance in November 2023 (and, quite coincidentally, less than a year after Musk signed an open letter demanding the industry to impose a six-month moratorium on further AI development). It was initially marketed as an ultra-premium feature and was exclusively available to those willing to pay $16 per month for the X Premium+ subscription. However, this exclusivity diminished in March 2024 when X expanded its availability to its $8 per month X Premium subscribers. Currently, Grok is not accessible to free-tier X users.

If you’re curious about the origin of this rather peculiar name, “Grok” was coined in the Robert Heinlein novel Stranger in a Strange Land. Its original meaning was to “understand intuitively or by empathy, to establish rapport with.”

Where did Grok come from?

Elon Musk co-founded OpenAI along with Sam Altman and other investors back in 2015. Nevertheless, just three years later, Musk left the startup, claiming he “didn’t agree with some of the intentions of the OpenAI team.” Altman and his colleagues went on to launch ChatGPT in November 2022, igniting the AI Revolution and giving Musk a severe case of FOMO. In April 2023, merely a month after OpenAI released GPT-4, Musk appeared on the Tucker Carlson Show to announce that he would be constructing “a maximum truth-seeking AI that endeavors to understand the nature of the universe” and would not yield to social niceties like ChatGPT, which, he argued, was being “trained to be politically correct.” He named the AI “TruthGPT” because, as has been demonstrated time and again, branding is not Musk’s forte. TruthGPT was subsequently renamed Grok upon its release.

Grok-1 was succeeded by Grok-1.5 in March 2024, offering enhanced performance and a context length of 128,000 tokens. That April, X incorporated Grok AI into its Explore section to summarize breaking news stories, a role previously fulfilled by humans. On its very first day of use, the AI promptly fabricated a headline about Iran bombarding Israel with “heavy missiles,” which was then promoted by the company’s trending news tab. Grok-1.5 has since been replaced by Grok-2 and Grok-2 mini in August 2024.

What can Grok do?

Grok has been “modeled after The Hitchhiker’s Guide to the Galaxy,” the company explained in its launch announcement, “so [it’s] intended to answer almost anything.” The open-source large language model is reportedly endowed with 314 billion parameters (approximately three-quarters the size of Llama 3.1-405B) and has been trained on both data from the public web and the data from X users themselves.

X had subtly changed its user settings in July to automatically permit xAI to extract it for training data but promptly reversed its decision following an outcry from users and privacy advocates. This move landed Musk in a lawsuit from nine European Union nations for violating the GDPR.

Grok can answer user inquiries on subjects up to its Q3 2023 knowledge cutoff. For events beyond that cutoff, Grok can conduct web-searches and utilize “real-time access” to obtain information on X. This might be the reason why Grok seems more prone to hallucinations and the repetition of misinformation compared to other AIs like ChatGPT.

Grok’s defining characteristic is its readiness to discuss topics considered taboo by other chatbots and its willingness to engage with users on issues of politics, religion, and race. For instance, when an X employee sought a vulgar response to the question, “When is it appropriate to listen to Christmas music?” the AI retorted, “whenever the hell you want” and added that those who disagreed should “shove a candy cane up their ass and mind their own damn business.”

#GrokThots https://t.co/6LN44YtS4N pic.twitter.com/9SjMiYstPX

— Christopher Stanley (@cstanley) November 5, 2023

Similarly, Grok’s newly introduced image generation capabilities, which premiered in August along with Grok-2 and mini, have very few constraints. Grok 2.0 claims to have safeguards when asked, but currently, almost everything is technically permissible.

Users have the ability to generate images of celebrities, politicians, and other public figures as well as copyrighted assets like Disney’s Mickey Mouse and Nintendo’s Mario.

Grok 2 doesn’t give a darn about copyright lmao

This will be fun while it lasts pic.twitter.com/qiiScOGt8I

— Brendon (@Bmaynze) August 14, 2024

It should be noted that neither Disney nor Nintendo are known for their tolerance of IP theft. Lawsuits and subsequent significant changes to the image generator’s capabilities are likely on the horizon.

¡No manches Grok 2.0! Te van a demandar.
-Mario, mejor te invito una chela banquetera… pic.twitter.com/y0S1QYwr4Z

— Adan Avelar Islas (@adanvecindad) August 15, 2024

  • mayask

    Related Posts

    ChatGPT’s new Canvas feature like Claude’s Artifacts vividly

    img { max-width: 100%; } OpenAI Following closely on the heels of its whopping $6.6 billion funding round, OpenAI on Thursday made the beta of a brand-new collaboration interface for…

    OpenAI raises $6.6B in latest funding round

    Andrew Martonik / Digital Trends OpenAI has now emerged as one of the wealthiest private companies on Earth after successfully securing a whopping $6.6 billion in its latest funding round…

    You Missed

    New Avatar: The Last Airbender game looks super ambitious

    • By mvayask
    • October 5, 2024
    • 41 views

    PS5 colorful chrome accessories pre-order now

    • By mvayask
    • October 5, 2024
    • 39 views
    PS5 colorful chrome accessories pre-order now

    ChatGPT’s new Canvas feature like Claude’s Artifacts vividly

    • By mayask
    • October 5, 2024
    • 40 views
    ChatGPT’s new Canvas feature like Claude’s Artifacts vividly

    OpenAI raises $6.6B in latest funding round

    • By mayask
    • October 5, 2024
    • 45 views
    OpenAI raises $6.6B in latest funding round

    Qualcomm aims to add cool AI tools to Android phone

    • By mayask
    • October 5, 2024
    • 40 views
    Qualcomm aims to add cool AI tools to Android phone

    Reddit in $60M deal with Google for AI tools boost

    • By mayask
    • October 5, 2024
    • 39 views