Dodging the AI Bullet
The past month has been a whirlwind of exciting developments in Artificial Intelligence.
You've probably seen fantasy-inspired portraits of friends and celebrities pop up on social media from the app Lensa . If you're a techie, then you'll definitely have heard of OpenAI's ChatGPT doing everything from writing blog posts to Python code. In fact, MMN member Chris Rodgers just did a 20-minute live demo of this amazing chatbot during our community office hours.
What hit me wasn't just how cool these apps are, but their second-order social effects.
Solving the Metaverse Immersion Problem
I see Lensa’s popularity as a major stepping stone towards the Metaverse.
Lensa charges $7.99 for a starter 50-pack of portraits AI generated from 13-20 selfies. These are more realistic and diverse than anything I've tried previously.
The fact that everyone, regardless of how much you make, is willing to pay $8-50 to generate realistic avatars of themselves means a big part of the Metaverse equation is in the beginning phases of being solved: Immersion.
Immersion is a term to describe 1) in Virtual Reality (VR), fooling ourselves enough to believe that we’re actually inside a virtual world, or 2) in Augmented Reality (AR), seeing digital objects in the real world as if they were physical. Immersion is key to people eventually adopting the Metaverse.
Metaverse insiders tend to associate immersion with influencing the senses - hearing, touch, smell, taste and especially vision because up to 60% of our brains is devoted to processing visual information. If you can manipulate the senses enough, you can fool the brain into thinking that what's digital is real.
But something even more powerful is our own imagination. In order to imagine ourselves in a digital world is to see ourselves there - not just in first person perspective as most AR/VR demos show, but as realistically created avatars of ourselves residing inside the world.
Don't believe me? On Instagram, the most popular photos and videos are of people in the 3rd person rather than what they see (i.e. city/landscapes, sweeping scenery etc). We are naturally drawn (biased?) to the human face and body, in that order.
The purpose of immersion is for the user to feel attached to the digital world, which drives engagement and more time spent there, which in turn drives monetization. And there's nothing more we're attached to than our own self-image. In other words, our identity.
Lensa has shown that 1) the technology has matured enough to enable realistic, consistent and affordable reproduction of our likeliness without the need for expensive motion capture (aka mocap) equipment and 2) the willingness of the general public to pay for it.
Its success will drive commercial and research interest into accelerating both its performance and applications. Both startups and big tech are watching this development like a hawk. It's not unlikely that we will have realistic moving avatars posing against fantastic backdrops soon.
Parallel Technologies: Rise of ChatGPT
Big computing paradigm jumps (like from PC's to mobile) require an ecosystem of complementary technologies and parallel trends rather than any single app or platform. In this case, ChatGPT, the OpenAI chatbot that's been making headlines, has a role to play in 3 possible ways:
1) Enhancing NPC (non-player character) intelligence thereby making digital experiences more compelling with human-like AI characters. In other words, ChatGPT makes AI more human, while Lensa makes humans more digital. The future meets somewhere in the middle.
2) Making the creation of digital experiences cheaper by automating parts of the coding and design process (for starters).
3) The disruption (if not displacement) of knowledge worker jobs.*
There's a timescale and order involved, but let's assume these are plausible effects. We'll circle back to these factors in a bit.
ChatGPT summarized a 10,000-word transcript of our MMN office hours in 15 seconds. It's lacking in key insights but the ease of conversing with it is striking. The next important skill will be learning to phrase prompts in a way that the AI will yield desired results.
*Why will knowledge worker jobs be automated first? Because manual labor jobs are extremely difficult to replace. The fallout of the $10 billion self-driving car hype has shown us the tech, safety, and regulatory challenges of putting robots in the physical world.
On the other hand, anything that can be done on the computer can be digitalized, and hence be automated.
Adapting to this new paradigm will require both a mindset shift (that is framing these new technologies as opportunities rather than threats so that we can best use them) and societal system change (if you're curious, read on ;).
What's Next?
This then begs the question of what happens next. No one knows but I can paint one possibility.
Created by Midjourney, another AI that generates images based on text prompts. In this case, I used a simple one: "expanding the human mind, mind map, cinematic lighting, Miyazaki style"
Stage 1
More AI-generated photos of people and places will appear in our feeds, especially those with people in places. This will desensitize us to not only digital people in digital places but also IRL photos and video because:
1) We get used to it. This is the current state of landscape photography where stunning images are a dime a dozen that we now take them for granted.
2) It's cognitively tiring to differentiate between what's real and fake. Easier to bunch them together. This is why fake news is so effective (which ChatGPT could add fuel to fire).
Since many people travel for the gram, this decrease in social validation makes the expense, time and hassle of going to real places less attractive.
Simultaneously, digital experiences is improving with said advances, thereby making lower-cost digital activities relatively more attractive.
Stage 2
Moves people more toward the digital world i.e. the Metaverse. If mobile apps have taught us anything is that more engagement with them results in more immersion and attachment. If everyone in your circle is there and doing activities, then network effects will make it even stickier.
Meanwhile, our IRL dissatisfaction increases as the real world become more turbulent due to [enter the million things in the news we've already heard about]. Thereby increasing the demand for accessible digital experiences as a form of distraction and mental relaxation (they are two sides of the same coin).
Stage 3
Democratizes the Metaverse since everyone's beautiful and perfect as digital avatars. Most digital experiences are affordable. Everyone is happy.
Stage 4
Digital recreation of real world-class divide. How do you stand out when everyone's perfect? Spend more money on differentiation, whether that's buying more expensive digital accessories, experiences, and cliches.
All the while, we become increasingly attached to the digital world and detached from the real one.
The timescale of the above scenario is anyone's guess, but all the advances, trends and human nature make it very possible.
So how do we avoid this dystopian future?
The 10-Hour Work Week
Renowned economist John Keynes in 1930 predicted his grandkids would work just 2 days a week thanks to the rate of productivity growth and automation. The average American worker's productivity has indeed increased by 430% since 1950.
Therefore, it should take either less than 10 hours per week to afford the same living standards as someone in 1950, or our standard of living should be 4x higher.
Obviously, this isn't our reality. Keynes might have even considered our present a dystopian outcome.
I wrote a deep dive article explaining how this came to be, but the TLDR is that it's not a technology problem, it's a human and systems problem - a combination of our insatiable personal demand for more stuff plus inequitable wealth distribution.
The nature of the solution must match that of the problem. Therefore, the solution does not lie in our advances, but rather within ourselves. Since it's a human problem, we don't need to look to AI to solve it. I turn to the ancients instead.
Keynes and Aristotle existed 2,300 years apart but grappled with similar human problems.
Aristotle had a good answer as he saw leisure as the “The goal of all human behavior, the end toward which all action is directed.” But his definition of leisure wasn’t about relaxation or fun, but rather becoming “a complete, well-rounded human being”.
In other words, free time is used to develop our minds and character, of discovering and doing things that bring us lasting fulfillment - perhaps even informed and accountable members of society.
Our current work and cultural paradigm forces us in the opposite direction. To the extent that even our minuscule leisure time is used to recuperate through distraction and work disguised as vacation.
So if our AI advances will have any chance of saving us from a dystopian outcome, then we need to work towards building systems that actually give us back our time.
And in order to do that, we must have a handle on ourselves first.
The Stoic philosopher Seneca said “It is first to have what is necessary and second to have what is enough.” Because if we don’t know what’s enough, we’ll always feel poor and miserable. It's this very discontentment that drives much of our need to work more, make more, and consume more.
The change is in our own hands, but that's also the hard part.
Btw every word above was *not* written by AI ;)
AI might be able to write this article from my mind map in the future hmm...
