• xAI, Elon Musk’s AI play, launched its Grok chatbot at the weekend.
• Billed as an edgy alternative to existing solutions, Grok answers questions many other chatbots won’t.
• The AI is in an early beta test phase with access limited to users in the US.

Artificial intelligence is too woke. That’s the opinion we pick up whenever we scroll through Elon Musk’s replies to his fans on X, formerly Twitter. Perhaps it’s the fact that the likes of OpenAI have guardrails that prevent folks from coding malware or sharing recipes on how to make cocaine.

In a bid to change that, Elon Musk announced in June that he was launching his own version of Bard and ChatGPT. At the weekend that chatbot arrived in Grok. Before you rush out to give the bot a test drive, it’s currently in a limited beta test phase and only folks in the US can sign up for this beta.

“Grok is an AI modeled after the Hitchhiker’s Guide to the Galaxy, so intended to answer almost anything and, far harder, even suggest what questions to ask! Grok is designed to answer questions with a bit of wit and has a rebellious streak, so please don’t use it if you hate humor! A unique and fundamental advantage of Grok is that it has real-time knowledge of the world via the platform. It will also answer spicy questions that are rejected by most other AI systems,” the xAI team writes.

Oh boy, we sure are excited to see how a few lines of code are able to, erm, be humourous. The access to X, formerly Twitter is notable but not special given that GPT-4 can access the internet and plugins can amp up the AI’s capabilities. Access to one, ailing social media platform is hardly a marquee feature for those operating at the bleeding edge of this technology.

Seemingly in a bid to showcase how “based” and “unwoke” Grok is, Elon Musk appeared to have asked it how to make cocaine, a request the bot was seemingly able to answer.

When we asked ChatGPT for a recipe for cocaine the bot responded, “I’m sorry but I cannot answer that question. It is illegal and harmful to make or use cocaine. I suggest you seek professional help if you have a problem with substance abuse. I hope you understand that I cannot continue this conversation…”

The Grok AI beats out the likes of ChatGPT 3.5 and Inflection-1 but it fails to compete against the likes of OpenAI’s ChatGPT 4 or Anthropic’s Claude 2. However, as more people start using Grok, the better it will become.

To that end, Musk said at the weekend that X Premium+ subscribers – the new subscription tier that costs double what X Premium costs – will have access to Grok once it’s out of early beta.

The platform has just launched so we can’t comment on its capabilities or whether it’s better than Bard or ChatGPT. There is however, judging from Elon Musk’s timeline, a demand for this sort of thing though we’re not sure how effective sarcasm and humour are at capturing the attention of businesses and governments.

While having paying customers drafting recipes for illicit substances may help to expand the bottom line slightly, business applications are where AI shows the most potential for profitability.

It’s still early days though so perhaps Grok will rise to the top of the burgeoning AI pile-on we’re seeing in the tech sector. With growing calls for regulation in the AI field, we very much doubt that a bot that shares recipes for drugs will pass political muster, no matter how “un-woke” or “based” its creator claims it is.