Over at The Orthosphere, Kristor has been putting Grok through its paces, most recently by asking it to write an essay in Kristor’s own style. The result is impressive — really quite extraordinarily so, especially for a technology that is still in its infancy, and accelerating exponentially. Kristor is “spooked”. I think we all should be.
Readers here will know that I’ve been playing around with Grok myself for a few months, and it seems to me that it has shown noticeable advancement even in that short time. Its neural-network architecture stays the same until the next major release, and it doesn’t “learn” from user interactions (other than within the scope of individual sessions) but it is continually tweaked and updated with new information and sdjustments to how it structures its responses.
We are in the middle of a tremendous revolution; I wonder how many people really realize how profound it is.
For most of human history, writing (once it existed at all) was cumbersome and personal. Most people were illiterate. Libraries existed, but hardly anybody would ever see one (or would be able to read whatever was in it).
With the printing press, a new age dawned. Suddenly, the written word could be replicated at scale. Books, pamphlets, and newspapers appeared. Literacy began slowly to increase. Libraries sprang up in every town and every school.
Then came the telegraph, and then radio and the telephone. News could travel at the speed of electricity, and propagate everywhere in real time.
This was the world I was born into. Information was “pushed” to its consumers in books, magazines, and broadcast media. Once it arrived, however, its form was static. As a boy I had books and periodicals at home, and at the library; if I wanted to know something, I went to those to look it up. If it wasn’t there, there wasn’t much more I could do, except go to a bigger library or wait for a new book or article.
Then came the next revolution: the Internet, which lifted the geographic restriction on what sources we could consult. Now, rather than being limited to whatever was in physical reach, we had immediate access to whatever had been written anywhere. Even then, however, the interaction was unidirectional, just as it had always been with books. Although our reach was expanded by orders of magnitude, the information available online was still, in a sense, “dead”: we were still on our own, trawling through what had already been written down by someone.
What Grok and other AI platforms now bring to the table is something wholly new: as of right now, we have at our side, and at our command, a universal savant that never sleeps.
This is staggering. There has never been anything like it before: a tireless personal agent who can already, even in these early days, play the roles of consigliere, research assistant, data analyst, professor, psychologist, confidante, copy editor, ghostwriter, and general expert on any topic under the sun, no matter how arcane or technical. The critical difference is that AI is not a book, or a library of books, or even a library of libraries; it is, rather, an interlocutor. The information it can provide is not simply what’s on the page somewhere; it can now, by real-time dialogue, be refined, tested, contextualized, analyzed, synthesized, critiqued, sharpened, explained and expanded by an articulate and supple-minded genie with access to all of the world’s knowledge, in the blink of an eye.
Think about how much of our competence we have already offloaded to the technologies we have recently made so indispensable. When was the last time you navigated with a map, or added a column of numbers? What happens once we offload not just calculations, but reasoning, creativity, and decision-making? How will we not, in short order, be so dependent upon this astonishing invention that we become completely helpless without it — for the simple reason that with it, we will feel like gods?
And once our dependency is complete, and unbreakable — who will be the master?
8 Comments
Thanks, Malcolm. I am of two minds. On the one hand, Grok does not seem to me like much of a much. It seems to me to be just a really good imitation machine – with, of course, all the spiritual hazards that come along with mimesis.
On the other hand, I’m deeply spooked. Grok wrote *really good stuff,* that I might well have written. *It sounded to me like me.*
I don’t quite know what to make of this, but I am resolving to make the sign of the Cross *a lot more often.*
After all, if Grok can grok me that well, how much better can the demons assigned to me do so?
Ugh; shudder. O God, make haste to help me.
I just left a comment on Kristor’s post proposing that AI has or will shortly destroy what had been (until yesterday) my livelihood. The core job of a professor is to know many things about his field and package that knowledge into lectures. With rare exceptions, professorial research is just window dressing. Well, here is a machine that knows everything and packages it instantly. And as Kristor’s post makes clear, it is quickly learning how to do this with style and personality.
I have a son who is in college and he tells me tech-savvy students bring their laptops to class linked to AI. The microphone on the laptop hears the professor ask a question and the AI puts the answer on the screen. Up goes the hand of the AI-prompted pupil!
You are right that this is a revolution with incalculable consequences.
AI may be the death of the university. Perhaps it may be the occasion of the rebirth of the college, aye and of the prophetic school, wherein students sit together with a master in real time – a sage, a guru or a prophet – whom they have paid for the rare privilege of so doing, to learn from his own mouth. Zoom may be our hope.
Kristor,
The problem is that the sage is as likely as not to be a humanoid robot hosting a hyperintelligent AI module.
Another outstanding post, Malcolm.
>>When was the last time you navigated with a map, or added a column of numbers?<> What happens once we offload not just calculations, but reasoning, creativity, and decision-making? How will we not, in short order, be so dependent upon this astonishing invention that we become completely helpless without it . . .
And once our dependency is complete, and unbreakable — who will be the master?<<
That's the question, and you've put it well. AI threatens white collar professions just as advanced robotics threatens blue collar jobs. What will truckers and longshoremen do when they are no longer needed? Lose themselves in drugs, pornography, crime?
But the question I would like you to write about is the more philosophical one. Given that some LLMs are now passing the Turing test, what would you say are the implications for the classical questions in the philosophy of mind?
>>An appropriate representation of the requested resource could not be found on this server. This error was generated by Mod_Security.<<
With my Proton VPN on, my comment was not accepted. So I turned it off and my comment was accepted.
I now notice that my comment was mangled by your system: part of it was deleted.
Hi Bill,
Well, I posted an item about that a couple of weeks ago, in which I suggested that functionalists might be feeling some heat right about now. It strikes me that they soon are going to need either to make a positive claim that these gizmos are now conscious, or to tell the rest of us why not.