Elon Musk has just launched Grok, a new intelligent chatbot meant to compete with ChatGPT.
After ravaging acquiring Twitter (now X) last year, the eccentric billionaire unveiled last weekend his brand new technological toy: his own AI-powered chatbot.
An AI with a particular sense of humor
Developed by his AI specialist company xAI, Grok is inspired by the humorous science fiction work “The Hitchhiker’s Guide to the Galaxy”. Equipped with biting and sarcastic humor, it is capable of responding to sensitive questions, unlike other AI systems, which are smoother and more reserved.
As an illustration, Musk shared on X (formerly Twitter) a screenshot of an interaction with his AI, where it is asked to provide detailed instructions on how to produce cocaine.
In response, Grok jokingly suggests the user first obtain a chemistry degree and set up a clandestine lab in a secluded location. The AI then proposes acquiring large quantities of coca leaves and various chemicals, and then to start cooking while hoping not to get arrested.
What is Grok really worth?
xAI’s teams claim that their AI already performs better than systems based on GPT-3.5 for use cases such as solving mathematical problems. However, they admit that it is currently outclassed by the most powerful model to date at OpenAI, GPT-4.
The reason: its dataset is much smaller than the latter’s, and its training phase has so far only been two months.
Grok’s added value would also lie in its ability to draw in real-time from the data of the social network X, offering a more current (though not necessarily factual) perspective than its competitors.
Still in beta phase, the chatbot prototype is accessible to a limited number of users as part of an early access program, with new features planned in the coming months. To have a chance to test it, you can sign up on a waiting list, provided you have, of course, a paid account on X.
Elon Musk announced that Grok will be included in the new X Premium+ subscription, offered at $16 per month.
A turnaround for Musk
While the entrepreneur has been sounding the alarm on AI for years, even asserting that they could be far more dangerous than the nuclear threat, some might find it surprising that he’s venturing into this purportedly mined territory.
A co-founder of OpenAI, Musk eventually withdrew from the project in 2018, citing a disagreement with the vision of the teams at the laboratory led by Sam Altman.
In March, the head of Tesla and SpaceX also joined hundreds of researchers and other tech figures in a petition calling for a six-month pause in the development of artificial intelligence systems.
According to the petition, these models and their exponential improvements pose “profound risks to society and humanity”. In the frantic race to deploy increasingly powerful artificial intelligences, no one, not even their creators, can understand, predict, or control these systems.
— Elon Musk (@elonmusk) November 7, 2023
In July, Elon Musk announced the formation of xAI, in order to build artificial intelligence systems that go “in the right direction”. Among its staff are former engineers from OpenAI, as well as from DeepMind, Google, and Microsoft.
As for the surprising name of this new AI: it is taken from the SF novel “Stranger in a Strange Land” by Robert A. Heinlein. The term “grok” refers to the philosophy of the Martians. It is a verb meaning “to understand fully” and “to become one with the object of one’s reflection”.
Also read on The Coding Love:
- 🎃 For Halloween, I asked ChatGPT to tell me the most terrifying story possible for a developer
- 🐧 Someone created an Angry Birds video game clone using only AI
- 🦾 Boston Dynamics’ robot dog now speaks thanks to ChatGPT
- 🏔️ Researchers create 3D-GPT, an AI that crafts virtual worlds using ChatGPT-like prompts
