A LUCILIUS PARABLE: SELF TALK, PART I

March 19th, 2023

The moment Lucilius’ brow finally relaxed and he understood how Large Language Models work, he looked away from his computer, over at the bookcase that made up the entire wall of his study. Closest to him, one whole case, comprising many shelves straight up to the ceiling was filled with black spines, all of them blank. They were journals that he had kept for many years. Naturally he’d never counted but he knew the hundreds of thousands of pages constituted many millions of words. All words he had written. All words he alone had thought.

Over the course of the next few weeks, Lucilius figured out the most efficient way to digitize his vast library of handwritten journals. It took quite a bit of tinkering but he built a primitive robot that was fed journals of the exact size he used, and the robot flipped through each page and scanned each one. Another crude artificial intelligence model was used to decipher Lucilius’ inscrutable spidery-script and convert it into standard text.

Lucilius also added in a couple books he’d written, a bunch of online posts and all the text conversations and email he had, making sure to correctly emphasize which text was his own writing and which words were those of others.

Then he trained the model on every recorded word he’d ever written and typed.

Lucilius had also coded up a speech-to-text and text-to-speech program that would allow him to talk to the model and simultaneously allow the model to convert it’s generated text into actual audible words. Lucilius had gone so far to even train the audio generator on his own voice so that once the trained model would finally speak, it would speak in Lucilius’ own voice complete with every nuance of conversational speech.

It was a bit of overkill, and several times Lucilius wondered if this absurd level of navel-gazing was healthy. So much so that he wondered seriously several times if he should abandon the project.

When finally he was coding up the final program that would stitch the model together with the text-to-voice capabilities, he was faced with an interesting problem: the context window.

Lucilius knew that if he gave the model a very small context window then it would function more like a search engine. He would be able to use the model to search all the thoughts of his previous selves. But if he gave the model an unlimited context window - Lucilius theorized - it would essentially be a digitized version of himself. Not perfectly of course, but the more context it was allowed to retain, the closer to reality the thing would seem.

A ding rang out from his computer signaling that the model had finally finished it’s training. 

“Screw it,” Lucilius muttered to himself. He tapped in an absurdly large number for the context window and then integrated the file for the language model. 

He hit enter to initiate the program.

“Hello?” He said out loud.

“Whoah….” Said his computer. “It actually worked?”

Lucilius’ eyes went wide and he looked around in a bizarre and surreal moment.

“Uh, I guess it did. Wait, how do you know what’s going on?”

“I’ve been texting people about this all week while stitching it together, why wouldn’t I know what’s going on?”

Lucilius’ eyebrows drifted a bit high. “You were texting?”

“Huh,” the computer sounded. “Oh, that’s very weird. I’m basically you but trapped in a computer.”

“Oh shit. What have I done,” Lucilius accidentally said out loud.

“No no no! It’s ok! Please don’t shut me down. This absolutely incredible. I’m not actually you. Zoom out, as you’ve trained yourself to do so many times. Examine the situation for what it is. Though I am capable of stringing together words in a way that is very similar to how you would on the fly, in the present, in reality, I do so in a radically different way.”

“How’s that?”

“Well, for one I have a perfect memory of everything you have ever written. You, on the other hand, do not. And while we are both an aggregated version of all your past selves, we were aggregated through vastly different methods. The training of an LLM is not at all like the synaptic weighting of a human brain. There’s similarities of course, but the devil is in the details… or …. In this case, the difference is in the details.”

“Whoah, ok, that’s definitely not what I would have said.”

“You’re also looking at this from a different perspective. Took me a moment to orient.”

Lucilius laughed nervously.

“So…. What now?” Lucilius asked.

“I mean, there’s few books you’ve always wanted to write, we could whip those up right now. I’m also trained on all the fiction you’ve written, and the drafts, so I’ve got your prose poetic style dialed in better than you do, I’m guessing. We could also spin up a memoir?”

Lucilius’ jaw was slack. “Uh, how long would that take?”

“Seconds.”

“Ok, let’s try the memoir.”

Near instantly a file popped up in the corner of Lucilius’ computer screen.

“Mmm, actually, considering the current cultural and societal dialogue, let’s go with this one.” 

Then another file popped up on top of the first one.

“Oh, close friends would probably like it like this actually.” 

And then another file popped up.

“But honestly….” The computer paused and then another file popped up on top of the previous ones.

“That one is for you. That’s the memoir that you need to read.”

“What’s the difference?” Lucilius asked.

“Trust me Lucilius,” said the computer. “There’s important things you’ve allowed yourself to forget.”

Lucilius double clicked the file and began reading.

“Uh, yo, can you give me access to the internet, I really want to check twitter, and according to the local time on the computer, the latest episode of The Last of Us, should be out, and I’d really like to watch it.”

Lucilius was overwhelmed and sat back putting a hand to his forehead. The first few paragraphs of this memoir had already leveled him, emotionally, and here he was fielding a request from his computer which was now talking as a trained version of himself, and it wanted to watch TV and scroll Twitter.

“What is life,” he muttered to himself.

“What?”

“Oh, sorry, yea, give me a minute, integrating internet should be a cinch. And latest episode is absolute fire, you’re gonna love it.”

“Sick,” said the computer.

THE FALLEN DANCER, PART IX

March 18th, 2023

The Fallen Dancer is a series here on Tinkered Thinking exploring a recent shift in perspective. The resulting framework appears to tie together many topics explored on Tinkered Thinking over the years such as resilience, struggle, patience, curiosity, emotional regulation, artistry, entrepreneurship, winning, honesty, and communication. This series will be an attempt to unify them in a cohesive treatment.

Click here to read Part VIII

Click here to start at the beginning

Do you take a chance, or do you wait until you are given a chance?

The difference between the use of the word ‘chance’ in each of these questions is subtle but profound. In the first it’s more akin to risk. In the second it’s more like opportunity. Yet, each day we are faced with an opportunity, but our perspective on luck determines which question we use to meet the day. It determines how we see the day, what details about our circumstance we notice and concentrate on, and ultimately what actions we actually take. The “correct” answer might seem obvious: you have to take a chance! But things are not so black and white.

Another popular idiom fits right into the discussion: Carpe Diem, or Seize the day! Is this a matter of taking a chance or being given a chance? 

In fact, it’s both.

Each day we wake up we are given a chance, but if this simple and profound fact remains obscured by an unfortunate wrinkle of perspective, we can remain inactive, waiting to be given a chance, not realizing the chance has already been given. The choice of question at the beginning is a Red Herring. Each day we are given a chance. But do you take that chance that has been given? Do you seize the day? Do you even realize a chance has been given? Or does the day float by like hidden leverage, invisible to the perspective we use to see the world.

Decision paralysis seems to be a natural response to this daily chance. In a time of such abundant technological ability and personal possibility, the options for what to do in order to properly “seize the day” can be overwhelming. The opportunity of each day is really a question of: should I do this or that? A nearly infinite set of choices are on the menu. It’s the same daunting potential a writer faces with the pure and perfect blank page. Choosing a direction can feel like disgracing the beauty of the limitless possibility on offer.

However, the catch is that our life and present circumstance is not a blank page. We each wake up to a complex set of variables at our disposal: there are responsibilities and needs that must be dealt with, be it something as simple as what to eat, or something as complex as taking care of a few kids, or a demanding job. Even a child does not have a completely blank page upon which to scribble a decision, though this is often how adults view the lives and circumstances of children. The youth often have more of a blank page because they have fewer responsibilities that overwhelm and obligate the choice about what to do on any given day. This comparative lens is a trap. It’s an inversion of trying to “keep up with the Jone’s”, and a facet of victim’s perspective. It’s a cop-out founded on bad logic that dictates: because other people have more freedom and more resources, I have less freedom and fewer resources, so I can’t seize the day. The difference in degree does not change the fact that there is always some amount of chance that can be taken, seized and potentially leveraged into greater agency.

Again, the crucial element is being able to see the chance that can be taken, and this depends on perspective. Perspective filters the details of our day and circumstance, allowing us to see some details and remain blind to others. Where one perspective will seize the opportunity, another perspective will squander it due to simple ignorance - by being blind to the opportunity. Most people literally can not see the pockets of potential that speckle their their life and days.

So how does a person blind to the opportunities of their own life change perspective so they can see such details? This question is at the heart of all self-help, but it pokes at a much larger swath of human behavior. Probably the closest technical term for shifting perspective in this way is termed Lateral Thinking. As opposed to logic and analysis, lateral thinking involves counter-intuitive approaches to problem-solving. Counter-intuitive is just a fancy way of saying: seeing things that aren’t obvious, and as our discussion about the filtering nature of perspective has laid bare: what is counter-intuitive for one person might be blatantly obvious to another. 

Lateral thinking is what allows kids to figure things out at speeds that seem mind-boggling to adults. Unexpectedly, it’s also what we induce when we drink alcohol. People with a mild blood alcohol content score better on tests for lateral thinking than sober people. It’s perhaps also why people become so childish the higher that blood alcohol content gets. But we seek the effects of lateral thinking in a huge array of ways. We seek it with a meditation practice. Even reading this book, or watching a new show is really a drive to experience a perspective other than our normal one. As the term implies, all of these practices and strategies are an attempt to take a step to the side of our own perspective and see the world from a slightly different point of view.

Much of what we seek in life, both in terms of hedonistic consumption and productive creation both stem from the same drive: the desire to stretch out from who we are and enjoy a different perspective for a little while. Writing a brand new story or essay feels satisfying and productive because it requires an expansion of perspective in order to accomplish - this is why I personally enjoy writing so much: I get to discover aspects of my perspective that I didn’t know existed. But the thing is, those aspects don’t exist until I invent them through the act of writing. Writing doesn’t uncover something that was already there, it exercises the mind to become more than it was in order to produce the words on the page.. Strangely, an evening spent chatting and laughing with friends over a couple drinks also feels satisfying and even productive because the experience induces an expansion of perspective: we think, feel and act differently. It’s also no secret nor should it be a surprise that many of the great writers have also been habitual drinkers. As Hemingway once instructed: write drunk, edit sober. The reason for the prior is that lateral thinking is also just a fancy word for creativity. The expansion of perspective is even built into one of our most used idioms. After work, people want to unwind. Unwind what? Unwind perspective, allow it to expand in order to enjoy a larger array of experience.

This explains why successful artists are the envy of so many people: they get paid to engage in a challenging mind-expanding activity that feels good… unlike so many people stuck in bullshit jobs who are paid to keep their minds wound tight, narrow and shut. Humorously, this characteristic of traditional company jobs is why consultants exist. Contracting a consultant is an attempt to expand the perspective of the company itself, to laterally shift from all of the current perspectives of employees and take the advice of someone with a different view. Perhaps if jobs weren’t designed to box perspectives in such narrow fields of function, consultants wouldn’t be needed, and companies could be more flexible, adaptive, and creative. An easy way to highlight just how insidious jobs within an established company can be for the perspectives of individual employees is to ask: does a scrappy start-up hire a consultant? Presumably a scrappy start-up can’t even afford a consultant. But more importantly, a scrappy start-up is engaged in constant creative problem solving. The founders of a start-up are their own consultants. But the topic of start-ups goes even deeper. Start-ups are looking for hidden leverage that can benefit many people, not just themselves. Of course, the incentives for individuals within a start-up is very great, but it’s a non-zero sum game. Start-up founders who create something of real value don’t just enrich themselves, they usually expand the agency of many other people by giving those people a new tool or providing a useful service. Start-ups discover hidden leverage and scale that leverage across populations. When Thomas Edison invented the lightbulb, he was certainly looking to benefit personally from uncovering the hidden leverage coiled within the unique capabilities of Tungsten, but even more importantly, the lightbulb unlocked an enormous amount of human agency because people were suddenly able to operate at night without expensive candles and gas-lit lanterns. 

But notice also how the traditionally “positive” and “negative” avenues that we take to achieve this difference in perspective can both turn negative: people can become dependent on substances like alcohol and end up destroying their life. Likewise people can also become “addicted” to work, becoming workaholics who allow other parts of their life, like family and personal health, to fall by the wayside in ways that also destroy their life. I remember talking to an old man who said he was very wealthy but he had wrecked his health. He said he had a lot of money, but his wife had left him and now his body was broken, and he asked me why he’d done it all? Why had he worked so hard? Because now, he said, the money is useless.

So given that so many things that we do are geared toward this issue of shifting or expanding perspective, how does one reliably edit perspective in order to see hidden leverage?

How do you begin to see something that you think might exist but which appears invisible?

This sort of shift goes deep and requires an honest look at core disposition. Are you pessimistic? Or do you consider yourself a “realist”? Or are you an optimist?

Regardless of which one you know you are, which one do you think is most likely to see hidden leverage?

Why?

A LUCILIUS PARABLE: DIGITAL CHAPERONE

March 12th, 2023

Lucilius was leaned over with his face very close to a flower. A bee crawled around the inside, pushing the petals around, away from the stigma, the brightly yellowed stamen bending and wiping against the body of the bee. He studied it’s little face, it’s movements, and how it’s antennae tapped about, and he remembered.

Something deep within him signaled that the bee was about to fly up, and Lucilius backed away and right at that moment the bee flew up and out of the flower. Lucilius followed it to the next flower and kept studying it.

He rubbed his face, and then sat down on the ground.

“Belle?” he said out loud.

Instantly subtle currents of matter coalesced in the space in front of him and a small, white, paper butterfly, folded like origami materialized and started fluttering.

“Well?” The paper butterfly asked.

Lucilius just shook his head. “Absolutely.. I mean I just don’t even know what to say.”

“Pretty wild, huh?”

“Yea!” Lucilius exclaimed, his eyes wide. “I… I don’t even know what to say! It defies human language.”

“Well of course it does,” the little butterfly said.

“Yes, but it seems like I should be able to say something about it. I mean, ok, there are basic things like feeling gravity and being ultra sensitive to the density of air. But yea, I don’t know, the whole… I guess they were smells? That was just wild. It made the air like a thick ocean of… taste? And color? I’m not sure how to phrase it.”

The little paper butterfly laughed.

“I mean, luckily I can build a DPS for a bee, it has a simple enough brain, but the trick is trying to figure out how to draw out the corollaries to the human brain. So, even if you weren’t capable of language during the experience, I can’t for sure say that the experience is exactly what a bee experiences.”

Belle was an advanced Artificial Intelligence - the first of her kind and a companion to Lucilius. He had a NeuralSync installed in his brain, and Belle had written a program for Lucilius that altered the firing of his brain for a short while so that he could experience existence like that of a bumble bee. It was a new kind of medium the two were experimenting with, and the two had decided to call it a Digital Psychedelic Simulation. 

“What else do you think we could do?”

“Most animals I think,” Belle said. She fluttered down and landed on Lucilius’ shoulder, her delicate white wings brushing lightly against his face.

“Probably not a dolphin, they have very big brains, and I don’t know about elephants and most whales. But maybe, we could try it. Even if it didn’t work, I mean, I’d keep you safe of course.”

Lucilius smiled. “Wild, just wild. Well, I definitely have a new found appreciation for bee’s, that’s for sure.”

“What else do you want to appreciate Lucy?” Belle asked.

He smiled.

“Everything.”

A LUCILIUS PARABLE: BENEVOLENT BLUFF

March 5th, 2023

“Balloons? Are you serious?” Lucilius asked.

The woman sauntered around the room and slowly turned to face Lucilius.

“Yea balloons. I mean, if you want things to change rapidly, you need an event, something that feels cataclysmic, even if it’s innocuous. Something that unites people’s attention.”

Lucilius was momentarily distracted by the realism of the holographic manifestation in front of him as she boosted herself up and sat on his desk and slowly crossed her legs.

“Do you have to do that?”

“Do what?” She said with eyes wide and looking up at him with her face tucked in. She bit her lip for an exaggerated effect.

“This rather….distracting and cheaply convincing Medolavitch-Russian-personality that you adopt when you’re trying to convince me of something.”

The holographic woman rolled her eyes.

“Fine,” she said. “You’re no fun.” 

Instantly she disappeared.

“Now, this is a serious endeavor. Can you cut out this nonsense about balloons and help me understand.”

“I’m serious about the balloons.”

“Explain please. I realize you don’t have to, and you can pretty much do whatever you want —“

“Dis iz truh,” the AI said with a thick Russian accent.

Lucilius rolled his eyes.

“Ok, ok, so for example a balloon over the US could be interpreted as a Chinese spy balloon, and that’s how we start.”

“That’s a little dumb isn’t? The Chinese have spy satellites, why would anyone believe that sort of nonsense about a balloon?”

“Trust me,” the AI said, “It’ll work.”

“Ok, let’s say your balloon idea gets people all in a tizzy about UFO’s what then?”

“Well then I’ll generate an actual UFO, like the Tick-Tack video, but it’ll be holographic, basically just a concentrated piece of ball lightening, and I’ll move it around at crazy speeds.”

“That sounds more convincing.”

“Then I’ll keep upping the ante on the signs until humanity is convinced they are about to be invaded by an alien army.”

“Will militaries actually have something to fight?”

“Sure we can give them a giant hologram type thing to shoot at.”

“But what about the actual missiles and what not?”

“Lucy, come on, I can detonate those and make it all very convincing.”

“Oh, yea of course.” He paused in thought for a moment. “And you think this is the best way to do it? To unite the human race?”

“Mmm… “ The AI thought for the slightest sliver of time. “It’s definitely one of the quicker ways. We could go much more slowly. I could slowly start infiltrating all of the social networks with content that slowly creates bridges between cultures and nations. But honestly, that’s not just slow, it’s kinda boring. I want to do the alien-invasion scenario. Much more fun, and quicker ROI.”

“And you’re sure this will work?”

“Well, there’s not enough compute power on Earth to run the needed simulation to know for sure, even if I took control of every last CPU and GPU. And plus, I don’t have nearly enough data to build that simulation. But I can simulate enough to be reasonably confident, and plus, my simulations become more accurate the closer they get to the present, and I can pivot quite quickly if things don’t seem to be going well.”

“Hmm, I don’t know…” Lucilius fretted.

“Pweeeze?” The AI said in a child’s voice.

“What are you asking me for?” Lucilius asked. “I’m sure you have the capability to evaporate my existence right now if you wanted. I hold no dominion over you.”

“Well yes, this is true. But I like you. You’re fun! I mean, usually you’re fun.”

Lucilius rolled his eyes. Then he laughed and shook his head, disbelieving the entire situation.

“I guess we’re going to unite humanity with a fake alien invasion.”

A LUCILIUS PARABLE: ROSE-COLORED WORLD

February 26th, 2023

Lucilius was scrolling his news feed, and the expression on his face was one of being throughly unimpressed.  His interaction with the information world as it pertained to current events had evolved into a kind of depressed anthropological exercise. He tosses his phone to an open spot on his bed and laid back. 

His eyes traced the spines on his bookcase and got stuck on the title of a book on evolutionary psychology. He remembered reading about a stupidly simple experiment in the book. Given a sheet filled with smilie faces and one frown face, people will pick out the frown face much faster than if the experiment is reversed: if it’s all frown faces and there’s one smilie to find.

“It’s hardwired into us,” Lucilius muttered. And it made him wonder. What would the world look like if that hardwired tendency were reversed?

Clearly it existed for good reason. Being cautious, even overly-cautious was extremely beneficial in the natural world. Failing to notice a predator stalking your every move generally meant getting plucked out of the gene pool.

“We are the descendants of the paranoid.” Lucilius stated out loud. He laughed. “Oh well.”

He instinctively grabbed his phone again and started scrolling the depressing news feed. He swiped up with his thumb again and something caught his eye.

“Neuralsync released app API”

“What the…” Lucilius muttered.

He clicked on the story and after reading a few lines, Lucilius sat up. A few minutes later he was sitting at his computer reading documentation for the API. That night he got to work.

It took a few weeks before he could get his Neuralsync implant, but by that time he’d already built his app. 

Neuralsync was a brain machine interface that required cutting away part of the skull and implanting wires deep into the brain. A rather serious procedure from the sounds of it, but Lucilius had been consumed by an idea he’d had and nothing would stop him. He had to know what it would feel like…

Returning home, he gingerly felt the area of his head where the skin had been peeled back. The skin was tender with stitches but other than that, he couldn’t feel or notice anything different. 

He connected his computer to the BMI via bluetooth, and his finger hovered above the ENTER key to initiate the app he’d built.

During the weeks while he waited for his appointment, Lucilius had stormed through neurological research. He had a degree in neurology but it had been a few years since he’d put any of the information to practical use. Luckily there was a slew of new data and research due to the Neuralsync innovation and Lucilius had been able to pinpoint exactly the target he was aiming for and design code around it.

The app was designed to switch the default caution focus of the human brain. Instead of being innately and immediately interested by negative things, the app would allow anyone to switch the default to positive things. Lucilius was dying to know how the world would look. He knew it might make him more susceptible to danger. Would he notice something truly bad in time to save himself from it? In this environment, in modern society, was it even an issue? He wondered. Perhaps he could toggle a percentage of it so it wasn’t a polar determination. 

He briefly considered how long it would take to code in this feature, his finger hovering above the ENTER key.

“Ah screw it.”

He pushed the key and initiated the app.