After warning us of the danger of AI, Elon Musk launches his own chatbot
Elon Musk, who cofounded OpenAI but walked away from the company in 2018, has finally rolled out his long-threatened xAI chatbot to a small group of US X (formerly Twitter) users. The sassy, snarky Grok, like much of what Musk touches, has people equally excited and concerned.
"In some important respects, it is the best that currently exists," he posted on X, before its release. "As soon as it’s out of early beta, xAI's Grok system will be available to all X Premium+ subscribers.
Following four months of learning, Grok was tested on a range of tasks to determine its math and reasoning capabilities compared to current models LLaMa 2 70B, Inflection-1, GPT-3.5, GPT-4, Palm 2 and Claude 2. It performed strongly, though is still behind GPT-4 – OpenAI’s flagship ChatGPT model – for now.
“On these benchmarks, Grok-1 displayed strong results, surpassing all other models in its compute class, including ChatGPT-3.5 and Inflection-1,” the xAI team noted in a statement. “It is only surpassed by models that were trained with a significantly larger amount of training data and compute resources like GPT-4. This showcases the rapid progress we are making at xAI in training LLMs with exceptional efficiency.”
While still in its early days, Grok has a few points of difference. Accessing X information in real time, and promising – or threatening – to go where no chatbot has dared to go.
“A unique and fundamental advantage of Grok is that it has real-time knowledge of the world via the X platform,” the statement read. “It will also answer spicy questions that are rejected by most other AI systems.”
Those “spicy” questions include the one used to demonstrate the feature, in which Grok replies to a query about how to make cocaine. If you plug that question into GPT-3.5, for example, you’ll get the response: “I'm very sorry, but I can't assist with that."
“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!” the xAI release stated. “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!”
Or, as Musk modestly posted upon its launch announcement: “It’s also based & loves sarcasm. I have no idea who could have guided it this way.”
“We believe that AI holds immense potential for contributing significant scientific and economic value to society, so we will work towards developing reliable safeguards against catastrophic forms of malicious use,” the statement optimistically noted. “We believe in doing our utmost to ensure that AI remains a force for good.”
The launch comes eight months after Musk co-signed a letter in which tech developers and other influential figures demanded a pause on training AI models.
A month earlier, Musk posted of his former startup: “OpenAI was created as an open source (which is why I named it “Open” AI), non-profit company to serve as a counterweight to Google, but now it has become a closed source, maximum-profit company effectively controlled by Microsoft.”
While Musk has consistently criticized OpenAI, which he co-founded with Sam Altman in 2015, some could argue that xAI is hardly any better. It will only be available for X users paying US$16 a month for their Premium+ access, while the model used public posts on the platform to build its learning.
ChatGPT’s premium bot is also subscription-based, at $20/month.
While it’s in “very early testing” stages, and entering a crowded market that’s exploded since ChatGPT launched just under a year ago, on November 30, 2022, Musk certainly believes there’s room at the table for one more tech billionaire’s LLM.
“We believe that it is important to design AI tools that are useful to people of all backgrounds and political views,” the statement read. “We also want [to] empower our users with our AI tools, subject to the law. Our goal with Grok is to explore and demonstrate this approach in public.”
Meanwhile, the launch competed for airtime with other X news, with Forbes reporting that current X employees had spilled the beans on a new division at Musk's social media company named @HandleTeam. Documented over a string of internal emails and other correspondence, the department is reportedly setting up a marketplace for people to buy created but unused X handles, for the bargain price of up to $50,000.
While X still states, "Unfortunately, we cannot release inactive usernames at this time," Musk has hinted at a change to policies regarding X handles for some time.
"We’re purging accounts that have had no activity at all for several years, so you will probably see follower count drop," he warned in May.
Not surprisingly, it has attracted some criticism; while a purge is needed to free up usernames, cynics are not surprised Musk saw the potential in terms of dollar signs.
"Twitter is really going all-in on the monetization game, huh?" San Francisco investor Joe Maristela wrote in an X post. "It's going to be interesting to see how this affects the platform in the long run. On one hand, it could be seen as a way to help people get access to handles they may have always wanted. On the other hand, it's kind of a bummer that something that was once free is now being sold for such a high price."