Elon Musk’s Grok vs Sam Altman’s ChatGPT: Which AI Chatbot is better?

Elon Musk, the CEO of Tesla and SpaceX, has officially revealed his highly-anticipated AI chatbot, Grok. This large language model (LLM) has been developed by Musk’s AI company, xAI, and takes inspiration from “The Hitchhiker’s Guide to the Galaxy.” Grok is known for its sharp wit and a hint of rebelliousness, as stated in an announcement made by xAI on a recent Saturday.

One of the standout features of Grok is its access to real-time data from X, formerly known as Twitter. This unique advantage sets Grok apart from other chatbots like OpenAI’s ChatGPT by allowing it to provide the latest information on current events and happenings.

While both ChatGPT and Grok are AI-powered chatbots, they exhibit several key differences:

Data Sources
ChatGPT relies on a diverse dataset from sources like Common Crawl, web content, books, and Wikipedia. In contrast, Grok is trained using real-time data from the X platform, ensuring it provides up-to-date and relevant information.

Ownership
Grok is created by Elon Musk and a team of AI experts within xAI, while ChatGPT originates from OpenAI, an AI research organization. These distinct ownership structures shape the development and trajectory of these AI systems.

Market Availability
ChatGPT became publicly accessible in November 2022, catering to a broad user base. Grok, however, is currently in its beta phase and limited to users in the United States. Musk has indicated that it will be available to all X Premium+ subscribers once it exits the early beta phase.

Pricing
ChatGPT offers users two versions – a basic, free variant with limited real-time data updates, and a premium version priced at $20 per month, providing continuous access to real-time information. Grok, on the other hand, is available to X’s Premium+ users at a lower monthly cost of $16, potentially making it a more cost-effective choice.

Communication Style
Grok is known for responding to queries with a witty and somewhat rebellious tone, setting it apart from ChatGPT’s communication style. This distinct communication approach aims to offer a unique user experience, catering to diverse communication preferences.

In addition to its clever and somewhat rebellious responses, Musk has emphasized that Grok is designed to inject humour into its interactions, adding a touch of amusement to user interactions.
Benchmarks and test scores
In a head-to-head comparison, Grok, the recently unveiled AI chatbot by xAI, has shown its mettle against established competitors like ChatGPT. Grok, although only two months into its training and currently in its early beta phase, has demonstrated remarkable performance in early benchmark tests.

In initial testing, which included middle school math problems and Python coding tasks, Grok outperformed all other models in its compute class, including ChatGPT 3.5 and Inflection-1, according to xAI. Despite its relatively short training period, Grok’s ability to surpass these established chatbots has garnered attention and positioned it as a promising contender in the AI chatbot landscape.

Grok’s early success in these tests indicates that it has the potential to offer advanced conversational AI capabilities and is poised to provide users with enhanced experiences, making it a chatbot to watch as it evolves further.

Based on benchmark data, Grok has shown promising results. On the GSM8k benchmark, tailored to middle-class math word problems, Grok achieved a score of 62.9%, surpassing GPT-3.5 and LLaMa 2 but falling behind Palm 2, Claude 2, and GPT-4. In the 2023 Hungarian national high school finals, Grok scored a C grade at 59%, outperforming Claude 2 but lagging behind GPT-4, which scored a B grade at 68%. This suggests that Grok-1 is more capable than GPT-3.5 but not as advanced as GPT-4.

OpenAI claims that Grok can outperform models trained on larger datasets and with higher computational requirements despite being trained on less data. Elon Musk even stated that, in some respects, Grok is the best available AI model.

Not infallible
However, like other Large Language Models (LLMs), Grok is not infallible and can still generate false or contradictory information.

What sets Grok apart from ChatGPT? Grok uses a custom training and inference stack based on Kubernetes, Rust, and JAX, running a custom LLM called Grok-1, trained on real-time data from the X social media platform and web scraping. In contrast, ChatGPT is built on GPT-3.5 or GPT-4.0 LLMs trained solely on publicly available internet data. The GPT-3.5 model in ChatGPT was last updated in September 2022, limiting its ability to provide accurate information beyond that point.

The beta version of Grok is currently accessible to a limited number of users in the US and will be made available to X Premium+ subscribers at $17 per month, making it more expensive than ChatGPT+.

Which one’s better, Grok or ChatGPT?
Choosing between Grok AI and ChatGPT depends on your specific needs and preferences:

Grok AI offers real-time access to information from the X platform and is designed to be intuitive and engaging. It is also currently in its beta stage and is available to a limited user base. However, it is currently behind a paywall and is unlikely to be ever released as a limited, albeit free version.

ChatGPT on the other hand is publicly available with a large user base and is trained on a vast dataset of text and code. Furthermore, ChatGPT is capable of generating various creative text formats. But, it lacks real-time access to information, at least the free version, and it may be more prone to generating biased or offensive responses.