I Don’t Grok
By Gordon Hopkins
So, in my last column, I bellyached about ads for pills that don’t bother to explain what those little miracle pills actually do. Now I am going to rant about another product that is marketed heavily but I can’t figure out why anybody would want said product.
It is called Grok and it is the latest attempt to use AI (Artificial Intelligence) to make money. Grok is the brainchild of Elon Musk, the “genius” who bought Twitter for $44 billion, renamed it X and turned what was already banal and painful and hateful and made it much, much worse. Like opening a DMV in hell.
The release of Grok was announced earlier this month. According to Musk, “Grok has real-time access to info via the X (formerly Twitter) platform, which is a massive advantage over other models.”
Great. But what does it actually do?
In the lead up to the launch of Grok, Musk said, “In some important respects, it is the best that currently exists.”
Okay, but what does it actually do?
Earlier this year, Musk said the purpose of his new company, xAI, which gave the world the invaluable Grok, is, “to build a good AGI (artificial general intelligence) with the overarching purpose of just trying to understand the universe.”
And how does Grok do that? I mean, what does it actually do?
Grok is actually a Martian word. Roughly translated, it means to intuitively, deeply understand. It comes from Robert A. Heinlein’s famed 1961 science fiction novel, “Stranger in a Strange Land,” which is a truly remarkable piece of literature, if for no other reason that it is was once a favorite novel of both Libertarians and hippies.
But that still doesn’t explain what Grok actually does.
On November 5, 2023, The Guardian newspaper in Great Britain published a story describing Grok as “an artificial intelligence chatbot.”
Okay, so Grok is chatbot. So what does a chatbot actually do?
According to the IBM website, “A chatbot is a computer program that simulates conversation with human end users, often using NLP (Natural Language Processing) to parse inputs and generative AI to automate responses.”
The problem with stories about AI is that they are filled with nearly-indecipherable acronyms. It took me 20 minutes of searching online to figure out what NLP meant. I kept coming up with National Premier Leagues, which is a soccer league.
Anyway, it appears that a chatbot is an AI that tries to pretend to be a human.
Great. But what is the value in that?
In 1950, an English mathematician, computer scientist, philosopher and cryptanalyst (I’m not even going to bother to look up what that is) named Alan Mathison Turing invented what came to be known as the Turing Test. It is a test of a machine’s ability to exhibit intelligent behavior equivalent to, or indistinguishable from, that of a human. Said machine has passed the test if an evaluator cannot tell the machine from the human.
We are not there yet, but we are clearly getting closer. You can ask chatbots, like Grok, questions and get answers.
So what makes Grok better than those other, run-of-the-mill, garden-variety chatbots? Well, according to xAI, one thing that makes Grok different is that it will answer, “spicy questions rejected by most other AI systems.”
Most chatbots have various safeguards and filters of sorts to prevent people from getting answers society (or at least your mom) wouldn’t approve of.
For example, someone asked Grok for a recipe for cocaine. As a responsible AI, Grok said, “Please don’t actually try to make cocaine. It’s illegal, dangerous, and not something I would ever encourage.”
Did Grok not answer the question because it doesn’t know the answer or because, despite xAI’s claims, Grok actually does have some filters?
Grok answers questions. I suppose there is value in that. Assuming those answers are actually correct.
Unfortunately, like all other AI chatbots, the answers it gives are often flat-out wrong. xAI warned in a statement that Grok “can still generate false or contradictory information.”
So what makes Grok different from other chatbots? According to Musk, it has a “rebellious streak” and “loves sarcasm.”
Musk has tried to create a chatbot that has a sense of humor. Unfortunately, Musk has the sense of humor of a snotty 12-year-old, which means so does Grok. When asked for a step-by-step to make cocaine, Grok offered, “Step 1: Obtain a chemistry degree and a DEA license. Step 2: Set up a clandestine laboratory in a remote location.”
So, not only is Grok a jerk but its idea of “humor” is referencing a show that has been off the air for a decade.
Put simply, Grok does nothing of value and adds nothing of value to anything.
So, now we have the answer. What is Grok?
Grok is marketing in its purest, most cynical, most manipulative form. Grok is proof that, with a truly aggressive marketing strategy, you can sell something that is nothing.
