ReadMultiplex.com: The Artist’s Canvas: GrokBot.Rocks and the Liberation of Results
43 minAudio
Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode! I have watched three revolutions reorder the way the world works. Each time, a cohort of capable people was left behind, not because they lacked skill, but because they loved the sequence of their work more than its results. The fourth revolution is here. They fell in love with the process of their work and not the results. Every great shift in computing has left that cohort behind. The Grok Bot shift is happening now, and it is moving faster than any I have seen. I Have Watched This Happen Before I watched the Personal Computer reorder the desks of the world. People who once typed on typewriters or wrote by hand moved to word processors that corrected spelling and moved paragraphs with a keystroke. The results arrived faster and with fewer errors. A group of skilled workers kept the old sequence anyway, because the feel of the keys and the sound of the carriage return still matched the work they recognized. The Internet then placed libraries and catalogs on a screen that anyone could query. Research that once required a trip and a stack of indexes finished in minutes. Some researchers kept walking the stacks, because the physical volume and the card catalog confirmed the thoroughness they expected. The iPhone put the calendar, the camera, the map, and the message box in one device that fit in a pocket. Meetings that once needed a phone call and a paper diary became entries that updated across every device. Some managers kept the paper diary, because the act of writing the time and place still felt like the meeting had been arranged. Each time the sequence improved, a cohort of capable people stayed with the methods they already knew. They fell in love with the process of their work and not the results. Read more at: ReadMultiplex.com Support this work by buying us a coffee: https://ko-fi.com/brianroemmele
Audio and video play straight from the show's own hosting. PodSubs stores and serves no media.
Transcript
From the show. Tap a time to play from there.
Have you ever stubbornly clung to an old way of doing something just because the process itself felt, I don't know, satisfying? Oh, absolutely. I think we all do that. I mean, really think about it. I'm talking about sitting at your desk, pulling out a physical notepad, writing down a to -do list with a pen you really like. A good pen makes a difference. It really does. And then getting that visceral, almost aggressive thrill of just crossing an item off. It's a great feeling. It is. Or maybe you're the type of person who manually organizes your digital files into perfectly
nested, color -coded folders, dragging and dropping them one by one. Even though you have a perfectly good global search function. Exactly. A search function that could surface literally any document on your hard drive in a millisecond. But you do it anyway. Right. You do it because those mechanical steps feel like actual work. Yeah, the friction of the process sort of tricks your brain into feeling productive. It's a deeply ingrained cognitive trap, to be honest. We are biologically wired to conflate the effort of the journey with the value of the destination.
Wow. Yeah. The effort equals value in our heads. Precisely. We feel that if something didn't require sweat or friction or just a lot of time, it somehow isn't rigorous. It isn't valuable. Which is a dangerous trap right now. It is. And right now we are sitting at the absolute epicenter of a technological shift that is exposing that psychological trap on a global scale. Okay, let's unpack this because this brings us right into the core of our deep dive today. We are looking at a massive structural shift happening in the world of work
right this very second. A total reordering. Yeah. And this is all mapped out in a fascinating series of essays and fieldwork from Brian Romley. He calls this the Grokbot era. It's a great term. It really is. And at the absolute center of his research on his Read Multiplex platform is this eternal underlying tension. The tension between loving the sequence of our work versus loving the actual result. Right. To really grasp the magnitude of what Brian Romley is laying out here, we have to establish the conceptual foundation
first. So where do we start? Well, his core thesis is that humanity is currently entering the fourth revolution of reordering how the world works. The fourth revolution. Yes. And the historical pattern here is ruthless. It's ruthless, but it's entirely predictable. Predictable how? What happens every time? In every single one of the previous three revolutions, a specific cohort of highly intelligent, highly capable people gets completely left behind. Just totally sidelined. Totally left behind. But the crucial insight
is why they get left behind. I mean, it is almost never a lack of intelligence. It's not a lack of skill, and it's certainly not an inability to learn the new technology. So what is it? Why do smart people fail to adapt? It happens purely because they fall in love with the mechanical steps of their profession rather than the finished product those steps were designed to create. I want to really dig into that. Because the immediate reaction most people have to the AI era, you know, when they think about it, is just pure fear. Fear of the complexity. Exactly. Fear that
the technology is just too complex to learn. But you're saying, based on this research, the bottleneck isn't technical capability at all. Not at all. It's an emotional attachment to a routine. Yes, it's emotional. So to prove that, I feel like we have to look at the historical evidence. We have to prove this isn't some new unprecedented phenomenon brought on by neural networks. The sources argue that to see exactly where we are going with autonomous AI agents, you really only have to look at the oldest story in the history of tools. We have to go back. Way back. We have to examine the transition from
the desktop to the pocket and look at the exact moments when manual sequences were annihilated by new technology. Annihilated is a strong word, but it's accurate. The first example that comes to mind, and one that Rommel's work highlights, is the transition from the typewriter to the word processor. Oh, that's a classic one. Right. When you think about the late 1970s and early 1980s, when CRT screens first started replacing physical paper on a roller. The leap in productivity was staggering. It was. I mean, you could backspace without using whiteout. Right. No more messy
correction fluid. You could move an entire paragraph from page three to page one with a few keystrokes instead of retyping the whole document. Spelling errors plummeted, too. Exactly. The final result, which is a clean manuscript, arrived drastically faster. And yet there was a massive cohort of highly skilled writers, journalists, office workers who just aggressively refused to switch. And we have to ask why. Because it wasn't because they didn't want a clean manuscript. Everyone wants a clean manuscript. Right. Obviously. It was because they missed the tactile feedback
of the sequence. They missed the physical resistance of the keys, the smell of the ink ribbon. The loud, percussive sound of the carriage return at the end of a line. Exactly. That ding and the swipe. They had spent decades internally associating the physical friction of typing with the intellectual act of, quote unquote, real writing. So when the screen removed the physical friction, they felt like they weren't truly working anymore. They felt like frauds almost. They clung to the old sequence because the process was quite literally their identity. Wow. And we saw the
exact same psychological block a decade later. Right, with the rise of the internet. Specifically in academia and research. Yes. Suddenly, massive university libraries and sprawling physical card catalogs were just digitized. Overnight, practically. Anyone could query a database from a terminal and finish in minutes what used to require a physical trip across campus. Hauling a stack of heavy, dusty indexes to a desk. Yeah, spending hours manually cross -referencing authors. Yet, as the research points out, a significant number
of veteran researchers just... Kept walking the stacks. They insisted on physically roaming the library aisles. And again, the psychology is fascinating here. In their minds, the sheer physical volume of the books, the tangible flipping of the index cards, confirmed their thoroughness. And the physical exhaustion at the end of the day. Right. If they just typed a query and got the answer instantly, they felt they had cheated. Like they didn't earn the answer. Exactly. They equated the sequence of the search with the rigor of the findings. And then, you know, you shrink
that entire revolution down into the pocket with the iPhone in 2007. That was a huge cultural shift. I remember this vivid shift in corporate culture. Suddenly the calendar, the camera, the map, the Rolodex, the physical notepad. All of it. They're all absorbed into one glass rectangle. Meetings that used to require a phone call, a secretary checking a desk calendar, a physical paper diary. They became digital entries. Yeah. that magically, silently updated across everyone's devices instantly. But you still had managers,
often the most senior ones, who fiercely clung to their expensive, leather -bound paper planners. Why? I never understood that. Well, because the act of writing the time and place down with a fountain pen made the meeting feel arranged. Oh, I see. The digital update felt too invisible. It didn't feel real. It is the exact same dynamic playing out in a loop. The sequence improves, the mechanical steps between an idea and a finished product shrink. But a cohort of otherwise capable
people stays with the methods they already know. They fall in love with the process, not the result. Exactly. But I want to push further back into the history here because the sources trace this battle between sequence versus result way back. Long before silicon chips. Right. Long before even electricity. The historical timeline Brian outlines is incredible, particularly when you look at the printing industry. Okay, tell me about the printing industry. If we span from the 1880s to 1985, we see this exact cycle play
out in just brutal fashion. For centuries, since Gutenberg basically, the sequence of setting type was incredibly manual. It was a true craft. It meant a highly skilled hand compositor standing in front of a massive wooden case, physically picking up one tiny backwards metal letter at a time and placing it into a composing stick. And they had to read upside down and backwards. Yes. It was a craft that took years to master. Right. And they were covered in lead dust, working by gaslight. It was exhausting work. And then
in the 1880s, Otmar Mergenthaler invents the linotype machine. And this fundamentally breaks the old sequence. So how did the linotype actually work? Because it was a massive deal. To understand how disruptive the linotype was, you have to understand the mechanism. Instead of picking up individual letters, an operator sat at a 90 -key keyboard. Kind of like a massive typewriter. Right. And when they pressed a key, a brass matrix, which is basically a mold of a letter, dropped from a magazine into a line. Okay. Once the line
was full, the machine injected molten lead directly into the molds, casting an entire solid line of type at once. A linotype. A linotype, hence the name. That is wild. It effectively increased the speed of typesetting by, what, an order of magnitude? At least. Newspapers just exploded in size. They could print multiple editions a day because the sequence got radically shorter. Exactly. But the hand compositors were devastated, of course, because their manual dexterity was suddenly devalued. All those years of learning to read backwards just gone. But the cycle doesn't
stop there. Fast forward exactly a century to 1985, which is the dawn of the desktop publishing revolution. Right. This is the era of the Apple LaserWriter and Adobe PostScript. And Aldus PageMaker. In fact, Paul Brainerd, the founder of Aldus, is the one who actually coined the term desktop publishing. I didn't know that. Think about the disruption at that specific moment, though. typesetting houses that had survived the linotype era, companies that had defined this meticulous, expensive, heavy machinery craft for generations. They suddenly
had to watch a single person sitting at a single desk in a suburban office produce a completely finished, mathematically perfect layout on a screen. And just print it out locally. The manual setting of type, whether by hand or by molten lead, just vanished entirely. But here is the crucial takeaway from the source material, and I want to make sure we underline this for you listening. It's an important point. The skill that survived that 1985 revolution was not the mechanical ability to set type. No, it wasn't. The skill that survived was the eye for a good
page, the aesthetic judgment. The understanding of kerning, hierarchy, and balance. Right. The sequence was destroyed. But the judgment remained. That distinction is paramount. When the sequence is automated, the judgment is elevated. I love that. I want to look at another historical example from the sources because it introduces a slightly different angle on this. The story of the telephone dial in 1891. Ah, yes. Allman Strauger. Yes. Allman Strauger, who's a Kansas City undertaker.
Which is just the perfect weird historical detail. You could never make that up. An undertaker invents the rotary dial. The story of Strauger perfectly illustrates how frustration with a manual sequence drives innovation. In the late 1880s, if you wanted to make a phone call, you didn't dial a number. Right. You picked up an earpiece, cranked a handle, and spoke to a human operator at a central switchboard. You would say, connect me to the bakery. And the operator would physically take a patch cord and plug it into the bakery's jack on a massive wall of circuits. But Strauger,
our undertaker, had a problem. he noticed a sudden drop in his business. A very suspicious drop. Very suspicious. The story goes that he discovered the local human telephone operator was actually the wife of a competing undertaker in town. Talk about a conflict of interest. Seriously. So Strouger became deeply paranoid, and probably rightfully so, that whenever someone called the exchange asking for an undertaker, she was secretly routing his incoming business calls to her husband instead of him. Whether he was right about the conspiracy or not is almost beside the point, honestly,
because his reaction changed the world. He realized the human in the middle was the problem. Exactly. The manual sequence of the switchboard operator was a vulnerability. So he designed and patented the first commercially successful automatic telephone exchange. It was a complex electromechanical switch driven by a rotary dial on the user's phone. Right. The user dials the number, the electrical pulses control the switches at the exchange, and the connection is made automatically. The rotary dial entirely removed the manual sequence of asking a human to plug a wire into a board. The sequence died. But the result, which is the
actual conversation between two people, remained. In fact, it flourished because the friction was gone. The conversation was always the point, not the patch cords. And we see this friction removed again in 1972 with the shift from the physical slide rule to the pocket calculator. This is perhaps the most direct parallel to the cognitive shift we are experiencing today. I agree. For the vast majority of the 20th century, engineers, physicists, and architects lived and died by the slide rule. As the text notes, the
slide rule literally went to the moon. It did. The engineers at NASA, computing orbital mechanics for the Apollo missions, were using slide rules made by companies like Kufel and Esser. It was a physical instrument requiring intense manual manipulation to perform logarithmic calculations. But then, in 1972, Hewlett -Packard introduces the HP -35. It was the world's first scientific pocket calculator. It could perform trigonometric and exponential functions with just the press of a plastic button. The result was instantaneous.
And within just a few years, the great American slide rule manufacturers essentially collapsed. Kefla and Esser produced their last custom slide rule in 1976. Just four years later. Okay. I hear the historical argument. I really do. But I have to pause here. Go for it. Because I can't let go of the idea that we are losing the craft. The craft of the sequence. Right. When an engineer is forced to use a slide rule, the physical friction of manipulating the slip forces them to intimately
understand the underlying math. They have to keep the decimal point in their head. They have to understand the magnitude of the numbers intuitively. If a digital device just spits out the answer, Don't we just end up with a generation of operators who don't actually understand the bridge they are building? It's a common fear. It feels like we are comparing a slow, meticulously home -cooked meal to a microwave TV dinner. Sure, popping a tray in the microwave is faster. The sequence is shorter. But is the quality of the result actually the same? What's fascinating here is
that that is the most common counter -argument, but the historical record... firmly rejects it. And it's important to understand why the microwave dinner analogy fails mechanically. Why does it fail? Because the microwave automates the recipe. It replaces the judgment of the chef with a prepackaged mediocre standard. But the HP35 calculator did not replace the engineer's judgment. It replaced the grunt work. Oh, I see. To refine your analogy, the calculator isn't the microwave replacing the chef. The calculator is the sous chef. who
chops onions for four hours, freeing the head chef to actually design the menu. That's a great way to put it. When the slide rule vanished, the calculator didn't abandon the bridge. The bridge still got built, and it was actually engineered better. Better? How? Because the engineer could test a dozen different stress scenarios in the time it used to take to calculate just one. So the people who made the switch didn't lose their judgment. They simply applied their high -level judgment to a vastly larger volume of finished work. Exactly. Desktop publishing didn't abandon
the beautiful page. It allowed a single designer to iterate on 20 different layouts in a day, applying their aesthetic judgment to more work. While the holdouts ended up applying their identical level of judgment to significantly less work. The engineer didn't disappear in 1972. It stopped wasting their entire afternoon doing basic arithmetic. That reframing is powerful. The engineer didn't disappear. They just got to spend more time actually engineering. Right. And that bridges us perfectly from these physical tools to the exact software precedent that Brian Romley argues mirrors today's
AI revolution. The perfect software parallel. To understand the Grokbot, we have to look at 1979. We have to talk about the birth of the electronic spreadsheet. We have to talk about VisiCalc. VisiCalc is arguably the most vital precedent for understanding the Grokbot era. For anybody who doesn't know the history, VisiCalc was created by Dan Bricklin and Bob Frankston for the Apple II. A legendary piece of software. And to truly appreciate it, you have to picture what financial modeling looked like before 1979. We aren't talking about opening Excel. We are talking about massive, physical paper ledger
sheets spread across giant desks. Agonizing is the only word for the old sequence. Let's walk through the mechanics of it. Imagine a massive spreadsheet drawn on paper tracking a company's five -year financial projection. Okay, I'm picturing it. Lots of tiny boxes. Every row and column interacts. Now imagine the boss walks in and says, What if our materials cost goes up by 2 % next quarter? Panic. Utter panic. Exactly. Because changing that one assumption meant you had to physically get out a pencil eraser, erase the materials cost cell, manually calculate the
new number, write it down, and then erase and manually recalculate every single dependent cell all the way down to the bottom line. So you would have junior accountants working overnight, covered in eraser dust, running 10 key calculators, just to see how a single 2 % variable changed the profit margin. It was a massive manual sequence. And then Dan Bricklin is sitting in a classroom at Harvard Business School, watching a professor erase a blackboard to recalculate a model, and he realizes there has to be a better way. Enter
VisiCalc. The mechanism of VisiCalc was a revelation. It introduced instant recalculation. Instant recalculation. You change one single number at the top of the digital sheet, and the underlying logic tree instantly updates every dependent cell on the screen. The result was so profoundly transformative that accounting firms and businesses were buying $2 ,000 Apple II computers for no other reason than to run this $100 piece of software. It's fundamentally changed the behavior of an entire profession. Accountants stopped being
what the sources brilliantly call human recalculation engines. Human recalculation engines. When the manual sequence of erasing and writing was removed, they evolved into strategic analysts. Suddenly they could ask, what if a hundred times in a single afternoon? They could model a dozen different business scenarios before lunch. And that mechanism, the instant recalculation of a sequence, brings us directly to the modern parallel. On October 29th, 2026, Brian Rommel posted an incredibly bold claim. What did he say? He wrote, Grokbot is the most important business tool of the last
hundred years. That's a massive statement. Yeah, he's saying it's bigger than the commercial internet, bigger than the smartphone. How does he justify that? He justifies it based on what the technology scales. Think about the progression. The linotype scaled typesetting. The calculator scaled arithmetic. VisiCalc scaled financial modeling. Okay, I'm tracking. If VisiCalc was revolutionary because it instantly recalculated numbers, a Grok bot is revolutionary because it instantly recalculates the entire day. Recalculating the day. I want
to sit with that phrasing for a second because we all know that feeling. You open your laptop at 8 a .m. You look at your calendar. And you see an eight -hour stretch of pure sequence work. Just endless tasks. You have to triage 50 emails, pull data from three different SaaS platforms for a weekly status report, summarize a 40 -page PDF research brief, prep an agenda for a 2 p .m. meeting. It's a slog. It is the modern equivalent of erasing ledger cells. It's moving digital paper from one side of the desk to the other.
Precisely. And what is happening right now in the Grokbot era... is that an eight -hour stretch of that specific type of manual sequence work can now finish before a cup of coffee cools. Before a cup of coffee cools. Assuming an autonomous agent is executing the sequence. Correct. But let's clarify the mechanics here, because people throw around terms like AI and bot very loosely. When we say grokbot, what exactly are we talking about technically? We aren't just talking about
a chat window where you type a question and get a text response, right? No, that is a crucial distinction. A chatbot is reactive. It waits for a prompt, gives an answer, and stops. Right. A grokbot, in this context, refers to an autonomous AI agent. It is a large language model equipped with memory, loop logic, and access to external tools. So it can actually do things. Exactly. You give it an objective, like compile the weekly sales report, and it breaks that objective down into a sequence of steps. It thinks through the process. It navigates to the CRM, pulls the data,
opens a spreadsheet, formats the numbers, drafts a summary based on the data, and prepares an email draft. It operates on a loop. Reason, act, observe the result, and adjust. It is executing the sequence autonomously. Okay, but as powerful as that is... It's very easy to get swept away in the hype and assume these bots are just going to replace every single human function by next Tuesday. Right, which isn't true. The sources are actually very careful to highlight the boundaries. Let's talk about the so what of these limits, because understanding what a Grok bot cannot do is just as important as understanding what
it can. It is essential to be precise about the boundaries. First, Grok bots cannot handle unscripted physical work in the real world. They are digital entities. Sure. But more importantly for the knowledge worker, they cannot execute judgment that relies on unique, unshared context. What do you mean by unshared context? Think about being in a meeting. A bot can transcribe the meeting perfectly, but it cannot read the room. Oh, I see. It doesn't know that the VP of sales is wincing because of an unspoken political rivalry
with the marketing director. It doesn't know the unwritten cultural dynamics of your specific office. It lacks the human intuition of context. Right. And most crucially, a grokbot cannot and should not make decisions with real irreversible human stakes. Like what? Firing an employee, finalizing a massive financial merger, deciding the strategic direction of a company. That domain remains purely human. Going back to our analogy, the autonomous agent is the sous chef doing the
chopping. The human remains the head chef deciding what goes on the menu. The sequence is handed off. The judgment stays. So we have the theory. We understand the historical pattern. And we understand the mechanical definition of the AI agent. But how is this actually being deployed in the real world right this second? It's happening fast. Because if I'm a listener sitting at a mid -level management job, this might still sound like science fiction. Is this actually happening in corporate environments today? It absolutely is. Brian Roy has been conducting extensive on
-the -ground fieldwork on this exact transition. What's he saying? He's been spending time embedded with clients at large public companies, walking them through this operational shift department by department. And the paradigm he is introducing to them is staggering. Okay. He is showing them how to arm single, individual employees with an almost unlimited number of Grokbot employees. It's the ultimate leverage. It's like every single mid -level employee suddenly gets promoted to a department head overnight. Yes. And they are handed a staff of 100 tireless interns who never
sleep, never complain, process information instantly, and never make a copy -paste error. You can imagine what happens to the competitive landscape of an industry when that happens. One person armed with an AI army is incredibly powerful. Unbeatable. But an entire legacy company armed with those armies becomes a juggernaut that is almost impossible to compete against using traditional manual labor models. But wait, let's look at the IT and security reality of this. If everyone in a 5 ,000 -person company suddenly has hundreds of AI agents running
around doing tasks, accessing databases and drafting emails, how is that not just total unmitigated chaos? It could be. I mean, an army of interns could do an unbelievable amount of damage if they are all firing off emails to clients and deleting files without oversight. That is the exact bottleneck Brian identified, and it is where he introduces a crucial operational concept, the filter strategy. The filter strategy. He highlights a legacy public company that is currently going all in on deploying this AI fleet, but
they aren't just unleashing bots blindly onto their servers. Good. Every single Grokbot command, every action an agent attempts to take runs through a centralized audit trail governed by a strict traffic light filter system. A red, yellow, and green light system? How does that work mechanically? In a corporate IT environment, a red light means the bot physically lacks the API permission to execute a destructive action. Okay, give me an example. If a bot decides, logically, that the best way to clean up a hard drive is to delete
a master customer database, The system throws a red light. The action is hard blocked. Makes sense. What about yellow? Yellow light means the bot has prepared the action, but it requires human authorization before executing. So it pauses. Exactly. For example, a bot can read an incoming client email, access the CRM, draft a complex response, and attach a finalized contract. But it cannot hit send. Accuse the draft. It throws a yellow light and waits for the human manager to review it, tweak it, and click approve. And
green light. Greenlight means the bot is pre -approved to run autonomously in a safe, constrained environment. Sorting incoming vendor invoices into a specific SharePoint folder, parsing data from PDFs into an internal spreadsheet, these run on green. That makes total sense. Brian explicitly notes that this filter system is the defining difference between an army and a mob. An army follows orders. A mob just breaks things. The human managers are still very much there, but their day -to -day reality has shifted. They are now overseeing the work. They're managing
the filter, reviewing the yellow lights, rather than performing the manual sequence themselves. We actually have concrete case studies from Brian's own fleet of deployed bots to prove out the economics of this. Oh, like the print system example? Yes. On October 10, 2026, he detailed a scenario where a Grok bot was put in full control of a physical print system. Right, the package printing operation. And this wasn't just generating text. This was an agent controlling a physical workflow. Very different from writing an email. The bot coordinated
local AI models to run a massive printing operation for three solid hours, flawlessly. And the result was a 58 % savings in costs. Huge savings. But what exactly was the bot doing that a human operator couldn't? It comes down to dynamic optimization. A human operator usually batches print jobs based on simple rules, maybe by paper size or color. Right. But the Grok bot was able to ingest the entire queue of hundreds of print jobs and dynamically optimize the routing in real time. It optimized
ink flow, paper routing, and job batching simultaneously, reacting to millisecond delays in the printers. A human brain simply cannot compute those variables fast enough. The bot optimized the sequence, delivering the 58 % savings. And then there is the example from October 28, 2026, which Brian famously refers to as the 3 .27 a .m. disagreement. I love this example. This perfectly illustrates the yellow light concept and the boundary between sequence and judgment. I found this deeply fascinating
because it makes the AI feel much more collaborative. His bots were working autonomously on a complex task overnight, and they ran into a conflict in the data. They hit a wall. Now, a poorly designed automation script would just crash, or worse, guess and make a terrible mistake. Which happens all the time with legacy software. Instead, the bots paused. They texted Brian a single, perfectly crafted message in the middle of the night, presenting clear choices. A, B, C, and D. If we connect
this to the bigger picture of what work looks like in the 2020s, this is the ultimate proof of the human retaining the judgment. The bots did all the grueling sequence work. They assembled the facts, cross -referenced the data, structured the problem, and isolated the variables. They handed the human a multiple choice question. Brian didn't have to do the research. He just had to wake up, look at his phone, apply his unique human judgment, pick B, and go back to sleep while the bots resumed the sequence. It makes you wonder what the absolute limit of this
is. If a human only has to answer multiple choice questions, how small can a company get and still operate like a massive corporation? Brian tested the absolute limit of that exact question with his zero human company concept. This is where things get truly mind bending. He built a test company run entirely by AI platforms. Entirely. He used discarded data from a long bankrupt company, essentially their old emails, operational manuals and customer data, and fed it into a network of autonomous agents. What was the goal? The
goal was to see if the company could function, interact and perform tasks with almost zero human oversight. And it did. It proved out what he calls the coming agent -to -agent economy. We are moving toward a world where your company's AI agent negotiates a contract with your vendor's AI agent. They settle the terms, execute the transaction, and simply log a summary for the humans. The agents are actually becoming the primary actors in the sequence of commerce. It's a fundamental shift in how business is done. Okay, that is incredible on a macro scale, but
let's bring it back down to earth, right to the listener's desk today. Let's do it. We know the historical theory. We've seen the massive corporate deployment. But how does an individual actually implement this? It's a great question. If I'm a marketing director or a small business owner, I don't know how to code an autonomous reasoning loop. Where do these grokbots even come from? Am I just locked out of this revolution unless I hire a software engineer? That is the exact bottleneck the Read Multiplex project addresses,
and it brings us directly to the toolkit, specifically a resource called grokbot .rocks. Let's talk about the sheer scale of this directory. As of late October 2026, this directory tracks over 7 ,539 public Grokbot templates across 26 different categories, and it is updated weekly. It is essentially the new App Store. But the sources note that this marks the end of software being sold as rent. You aren't paying a monthly subscription for a SaaS tool to do these things anymore. You
are deploying an agent to do it. To explain how a non -technical person approaches a directory of 7 ,500 bots, Brian uses a brilliant analogy from the 1950s. The paint -by -numbers kit. No, I like this analogy. Think about what happened when those kits were introduced. Millions of people who had never held a paintbrush in their lives, who had no formal training in composition or color theory, suddenly became painters. Why? Because they started by filling in numbered outlines. They didn't have to face the terror of the blank
canvas. The grokbot .rocks directory is that numbered outline. You use these thousands of existing pre -built templates to learn the pattern. You deploy a bot someone else built to handle your calendar and you watch how it works. You color inside the lines. And once you understand the mechanism of how an agent operates, then you can step up to the blank canvas and begin building your own custom bots. But the way you interact with these bots is different. You don't even have to hunt through the directory manually to find what you need. This is where we get into a concept Brian calls the superpower prompt.
And we need to define this because it isn't just typing a question into a chat box. Right. A super prompt is fundamentally different from a standard query. When people use a basic chat bot, they give it a task like, write an email to my boss. That relies entirely on the AI's generic default training. Which is usually pretty mediocre. Exactly. A super prompt, as Brian defines it, delivers a program and a map. A program and a map. You are injecting a multilayered set of logical constraints, a cognitive framework that teaches the AI how
to behave before it even attempts the task. So it's teaching it the rules of the game. It's the difference between handing someone a wrench and handing them a wrench along with a master class in plumbing and a strict set of safety regulations. So how does that work practically with a directory? You deploy a super prompt by pointing your Grokbot directly at the grokbot .rocks URL. You explicitly instruct it to study the skill sets that are already established in those 7 ,500 templates. You tell it to match those existing proven skills to the specific
goals you want to achieve for your business and then integrate those working recipes into its own behavior. You're essentially telling your AI to go read the instruction manuals of thousands of other AI agents and teach itself how to do the job you need done. Yes, but, and this is vital, with strict safety rules included directly in the prompt. You build the yellow light system into the agent's instructions. You explicitly tell it, never send, post, buy, or delete anything without asking me first. The person who used to spend their morning deciding which manual
sequence of clicks to perform, opening the email, copying the data, pasting it into the CRM, now spends their morning deciding which high -level results to pursue, while the bot handles the clicks safely. Let's make this incredibly concrete. The sources map out 10 specific handoffs, 10 manual sequences that you can hand off to a Grok bot today. Let's walk through them. I want to look closely at a few of these because it paints a vivid picture of how our days are about to change. Let's look at email triage. This is perhaps
the most universal pain point for knowledge workers. Think about how you handle email right now. You open an inbox with 150 unread messages. You manually scan the subject lines, trying to separate the newsletters from the internal company memos from the urgent client requests. It takes forever. You open a client email, read the history, open your calendar to check your availability, draft a response, and hit send. It takes hours. If you hand this off to a Grok bot, the mechanism is entirely different. The bot is connected to
your inbox via API. It reads the 150 emails. Instantly. Instantly. It uses its semantic understanding to categorize them. It archives the newsletters. It highlights the three urgent client emails. It accesses your calendar, drafts three perfect responses offering available meeting times, and leaves them in your drafts folder. You wake up, review the drafts in three minutes, hit approve, and you just saved two hours of your life. And we see a similar transformation with report assembly. No, report assembly is the worst. In the old
sequence, an operations manager might spend every Friday afternoon pulling numbers from Salesforce, downloading CSV files from a marketing platform, and manually copying and pasting those metrics into a PowerPoint slide deck for Monday's executive meeting. Such a waste of human potential. When handed off to a grokbot, the agent logs into the platforms using secure credentials, extracts the specific KPIs, formats them into the exact PowerPoint template you've provided, and even writes a brief executive summary highlighting anomalies in the data. The human manager doesn't
build the report. They review the insights generated by the report. And the sources list several others that follow this exact pattern. Summarizing dense research briefs. Managing local file inventories. Monitoring specific industry topics on social platforms like X. Drafting content in your own established voice. Checking vendor invoices for mathematical errors before payment. summarizing tax and legal paperwork, and tracking project status across multiple software tools like Jira
or Asana. Every single one of these is a job someone in an office is doing by hand this week. And every one of them can finish before your coffee cools if handed off to an agent. If you are currently doing these by hand, you have to ask yourself why. You are walking the dusty library stacks when the search engine is right there in front of you. You are attached to the sequence. you are mistaking the friction of the copy and paste for the value of the final report. Exactly. Now, all of these tools, these frameworks, these
super prompts sound absolutely incredible, but they obviously require a highly specific map to navigate safely and effectively. You can't just stumble into this. You can't just blindly stumble into an agent -to -agent economy without a guide, or you will end up causing chaos in your organization. Right. Here's where it gets really interesting. For those trying to figure out how to actually deploy this without breaking their company's IT protocols, the source of our deep dive today becomes incredibly relevant. Right. Brian Rommel has published a massive operational
playbook on his Read Multiplex platform. This isn't just theoretical musing. It is an actionable blueprint for deployment that sits behind the members line. For Read Multiplex members, you get access to the exact master prompt Brian uses. The super prompt with all the complex safety rules and cognitive frameworks already built in. You don't have to guess how to code the yellow light system. He gives you the text. You get 10 ready -to -paste prompts specifically designed for those exact handoffs we just discussed, like email triage and report assembly. The operational
value is significant. The platform includes a comprehensive seven -step playbook to help you hand off your very first sequence this week safely. And for larger organizations trying to scale this, there is a detailed department -by -department handoff map covering sales, support, finance, operations, marketing, and HR. It maps out exactly which tasks should have red, yellow, and green light permissions for each specific domain. And then there is the insider guide at the bottom of grokbot .rocks, which is constantly updated. Members get Brian's 12 top bot picks and the
nine best bot stacks, meaning which specific bots to chain together in a sequence. Whether you're running a small business, operating as a solo creator, or just trying to manage a busy household with privacy in mind. There are 62 practical tips from actual AI builders, a five -point safety checklist to run before adding any community bot to your system, early access to Brian's newest bots before they hit the public list, and even 25 % off custom consulting work with him. It is important to understand the larger philosophy driving this research. Brian Rommel isn't just building these tools for the sake
of corporate efficiency. What's his ultimate goal? He views this moment as documenting humanity's greatest technological leap. The mission is to prepare individuals and businesses wisely for what he calls a multi -planetary future. A multi -planetary future? He is analyzing and sharing these hard -won lessons so that people aren't left behind like... the hand typesetters or the slide rule engineers. He wants people to survive the fourth revolution. And this brings us to an important point. This is the Read Multiplex
podcast. Be sure to support Brian and his work and become a member to see the rest of the story. It's vital work. This is entirely independent research. There's no massive tech conglomerate or AI company secretly funding this field work behind the scenes. If this conversation has ignited your imagination about what is possible for your career, your business, or your daily life, it's highly recommended to support the person mapping the territory. It's about value for value. It is the only way independent, on -the -ground fieldwork of this depth can continue in an industry
dominated by massive tech giants. So, to synthesize all of this, what is the core takeaway for the listener? The core lesson we have to take away is that our digital tools are no longer just passive calculators waiting for us to press buttons. They are no longer just screens we stare at. They are tireless, autonomous armies waiting for direction. The only limiting factor left in the modern economy is our own emotional attachment to the old manual sequences of our days. Which leaves us with a final lingering question for you, the listener. Think about your desk right
now. Think about the routine you're going to log into tomorrow morning. The sequence you are so used to. If a grokbot recalculates your entire workday tomorrow, stripping away every single mechanical step, all the email triage, all the data entry, all the report assembly. All the friction you've used to define a hard day's work. What does your judgment alone actually look like? When the comfortable, mindless sequence is entirely gone, are you ready to be the artist of your
own career? That is the true challenge of the Grokbot era. It forces us to figure out what we are actually worth when the busy work is gone. Thank you for joining us on this deep dive into the Read Multiplex research. Stay curious, question your sequences, and don't get left behind. leads the way with research deep and true technology
and history he connects them all for you Brian Romer leads the way with research deep and true technology and history he connects them all for you ancient sparks to modern code he traces every threat go exploration knowledge widely spread read multiplex go but human moves on love equation
for AI zero human companies build Thank you. Thank you.
Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode! Every cell in your body has lived its whole life inside a magnetic field. The Earth is a dynamo, and it has been...
Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode! The 78 RPM Record Hour, Episode 3: From Steel to Hillbilly (Early Country). From steel to hillbilly: Frank Ferera,...
Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode! The 78 RPM Record Hour, Episode 2: The Knife and the Steel (Hawaiian). The knife of the blues slide guitar and the...
Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode! Somewhere in your house there is a computer you stopped thinking about. Maybe it is a 2017 laptop in a drawer, its...
Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode! The 78 RPM Record Hour, Episode 1 (Pilot): Blues. Fourteen original 78 RPM records from 1917 to 1925 that trace...