Elon Musk just launched his own version of ChatGPT.
On Saturday, Nov. 4, Musk announced Grok, an AI agent that answers any question.
According to xAI, Musk’s AI company:
“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 X platform. It will also answer spicy questions that are rejected by most other AI systems.”
Musk says that Grok has “real-time access” to data from X.
He also said that subscribers to X’s Premium Plus plan will get access to Grok once it’s out of beta.
A quick note on the name:
Grok is a term coined in Robert Heinlein’s classic sci-fi novel, A Stranger in a Strange Land. In the book, “grok” is a Martian word that means you have a profound understanding of something or someone.
The term is used in geek and tech circles to say someone has a deep understanding of a topic.
In Episode 71 of the Marketing AI Show, Marketing AI Institute founder/CEO Paul Roetzer unpacked what’s going on with Musk’s latest venture.
We have very little insight into what Grok can actually do. xAI’s own analysis says it’s comparable to GPT-3.5, but that's unverified. By their own admission, they’ve only trained the model on 2 months of data so far.
But it does have Elon Musk and X’s data behind it, says Roetzer.
Having access to proprietary data is key to creating value with foundation models. So the fact they have a pipeline to real-time X data could be a unique advantage.
You can question whether that data is valuable, says Roetzer. But if it is, then Grok has the potential to become a major player very quickly.
This isn’t just about building a foundation model, argues Roetzer.
X’s recent drop in valuation (from $44 billion to $19 billion) is a reason to launch Grok. The fastest way to make X worth more is to make it a true AI company or have X’s data power a foundation model.
Musk also has a tumultuous history with OpenAI. He and Sam Altman have bad blood after co-founding the organization. Musk has repeatedly criticized OpenAI's direction.
“There’s no way they had time for red teaming this,” says Roetzer. By which he means, there’s just no way trust and safety is built into this model.
And it’s not clear if Musk has any desire to do that in the first place.
He’s repeatedly criticized current models for being “woke.” And he’s displayed disdain for safety measures at X around content moderation.
“I worry about this a little bit, to be honest with you,” says Roetzer.