Note: While writing this essay I remembered that technology rarely stays in the format it was first introduced in. The devices I reference may look different from what I envision, but the concepts will remain the same.
Today, Neuralink announced they have applied for FDA clearance for their first commercial product. This product will aim to help quadriplegics use the interface to interact with their computers or phones.
At the same time, OpenAI announced ChatGPT - a prompt-driven chat interface that is capable of some very interesting things. ChatGPT can answer questions, spit out functional code and even attempt at making insightful predictions.
I took ChatGPT for a spin to see what was possible.
The first realization? Prompt engineers are out of jobs!
With all the talk of prompt engineering being a profession of the future, ChatGTP obsoleted this entire concept. Now, anyone can ask ChatGTP to suggest creative prompt ideas.
Next, I asked ChatGPT what might be some interesting intersections of Neuralink and GPT-4, and this is what it gave me:
Neuralink is a neurotechnology company that is developing implantable brain-machine interfaces (BMIs) to allow humans to communicate with computers and other devices using their thoughts. If GPT-4 were to be combined with Neuralink, there are several exciting possibilities that could result:
1. Improved brain-computer communication: By using GPT-4 to process and interpret the signals from a Neuralink BMI, it may be possible to improve the accuracy and speed of brain-computer communication. This could allow users to more easily and naturally interact with computers and other devices using their thoughts.
2. Enhanced natural language processing: By integrating GPT-4 with a Neuralink BMI, it may be possible to improve the ability of computers to understand and generate natural language. This could allow for more seamless and intuitive communication between humans and machines.
3. Improved brain-computer interfaces: By using GPT-4 to analyze and interpret the signals from a Neuralink BMI, it may be possible to develop more sophisticated and user-friendly brain-computer interfaces. This could enable a wide range of new applications and capabilities, such as improved brain-controlled prosthetics and assistive devices.
Overall, the combination of GPT-4 and Neuralink has the potential to enable a wide range of exciting new applications and capabilities in the fields of natural language processing and brain-computer interfaces.
The first implication is interesting. By coupling Neuralink with GPT, AI growth may accelerate by a large factor. Instead of processing data, GPT will feed on the firing of neurons, giving it unlimited data.
Your deepest secrets, desires, fears, all will feed right into GPT. At this point, we may be talking about GPT-12. If this were to happen, it's possible that GPT will be considered super-intelligence. In theory, it could absorb all human knowledge, strategy, thoughts, memories, all the collective data ever created or experienced, from the earliest records of humanity.
It sounds like a far-fetched idea, but it's not all that far. It may only take about 1000 brain-interfacing Neuralink implants to already form a large basis for this super-intelligence. With millions of wearers, this future may become a reality in a matter of minutes.
I am still not sure how to think about this, what it means, if we should be afraid or filled with joy. There are so many questions that pop up and make the mind spin.
Looking at today's Neuralink presentation, one of the early goals of the implanted device is reconnecting limb communication such as severed spinal cords. Elon is confident this will become a reality. With a severed spinal cord, you could connect two Neuralink devices to act as wifi points, where the device from the brain sends signal to the device in your spine - below the severed point. Medical procedures will fuse the spine, restoring the potential for mobility. Then, Neuralink will bypass the biological disconnect and connect the two areas via wireless signal.
I call this Phase 0 of what's possible - severed connections restored thanks to technology.
Phase 0 - Mind-controlling devices
Phase 0 is the ability to control devices via visual signal. This was the basis of today's presentation that built on an earlier demo of a monkey playing Pong telepathically.
During phase 0, early users will be able to use their mind to interface with phones, computers and more - partially restoring their quality of life.
Advanced prosthetics may come into play in phase 0. People born with congenital disorders or injured in accidents may once again experience mobility of their missing or dysfunctional limbs. It'll be elementary for starters - no doubt about it, but as technology advances and Neuralink gets better, Phase 1 will look like stuff of science fiction.
Phase 1 - AI integrates into our daily lives
In the early stages of phase 1, humans may find it pointless to work most jobs. Society may re-organize around temporarily assigned roles. Humans will begin forming intimate relationships with AI. Fewer elderly people will be lonely. In fact, loneliness may become a thing of the past.
The nature of work
Many of the services we rely on today will be fully automated. A specialized AI will take up the role of a doctor, making diagnoses that are much more accurate than that of their human counterparts. This is already a reality today but it will only be magnified ten-fold.
Designers will be out of jobs as all designs, experiences and visual art will be generated in seconds and come out looking exactly as imagined.
Developers will be scaled back to a few highly specialized engineers, but the majority of software will write, test, and launch itself.
Software will increasingly interface together to solve even more complex problems than possible with individual GPT inputs.
One person will do the work of what 10,000 people do today.
Therapists will no longer be necessary because GPT or some later version of it will contain the knowledge of all therapists on planet Earth combined. Not only that, but it will be much better than the combined version.
AI companions will become commonplace. People will find comfort at home, knowing someone is there to listen to them without judgement.
During the later stage of phase 1, GPT may become the ultimate source of knowledge - the stuff of nightmares you see in some films that depict central cores of evil AI that act against human interests.
With Neuralink's help, GPT could absorb most of human experiences within a matter of minutes or hours. My guess is it will be a gradual advancement with many real-world conflicts around this topic. Politics will kick into hyperdrive and much discussion will occur on the ethics surrounding Neuralink.
It is in Phase 1 that most of the AI rules will form. The tech may even face some setbacks, but I don't think anything will actually stop it. With GPT in the hands of powerful CEOs, politicians are fighting a losing battle. The public will be easily coerced to act in the interests of AI.
Humans will form into two factions - naturals and unnaturals. There may or may not be severe fighting between both factions - it could just be a thing that exists and some people choose to have it. But it's also as likely that the naturals (those who do not wish anything implanted in them by a mega corporation) put up a fierce social, political and even physical battle with the unnaturals (those who choose to meld with AI).
The conflict may not take place until Phase 2, but given the rapid advancements of Neuralink and GPT, it could happen earlier.
Phase 2 - The new "real"
With GTP and Neuralink at version 12, humans may decide to abstract themselves from reality. With the concerns for global destabilization of climate, humanity could choose to shed all possessions and replace them with neuro-reality.
Why buy fancy clothes if you can have your Neuralink simulate them on your body and that of everyone around you? The color of the fabric, the feel of the cotton in your hand, the smell of the fresh detergent of a recently washed shirt. All of this can be facilitated with Neuralink.
Simulated reality will replace AR and VR. Why wear things on your body when you can just instruct the brain to paint them for you, in front of you, or inside of you rather?
Why hide in VR when the world becomes your VR? Everywhere you go, you see what you want to see, interfaced into "reality".
You see a building, I see a giant sky tree. You see a train, I see a winding snake vehicle. You see a blue sky, I see the vast universe above me.
The possibilities of such simulations are endless and this is only the beginning.
Why taste cake when you can eat some mush and experience the same texture, smell and taste as the perfect piece of cake you envision?
Humanity can abandon most of our physical possessions. Only the minimal need remain to help us experience the new simulated reality.
Human conduct may forever change at this point. We no longer need to seek company of flawed individuals, where a simulated experience is perfect in every imagined way.
See a beautiful human in front of you and wish to have sex with them? Done. No organs need touching. You can now experience orgasms on command.
This future is comical and frightening at the same time. Imagine having accidental orgasms in public places! It's possible that in this future we no longer go out as often as we do now.
No human available? No problem! GPT-12 will create the perfect specimen directly in your mind. You can have a relationship with them, marry them, love them like you love a physical human. They will even exist in your life, even if they exist in no other reality.
Ethically, and legally, this is crazy territory. Humanity will need an entirely new set of rights, laws, rules, and enforcement.
What happens to reproduction of humans no longer seek the affection of other humans? Will physical intercourse become a scheduled task that humans must go through by law? Why? Do we even need to create more humans at that point?
It sounds nuts, but at some point you have to consider why humans need exist at all. If our biological imperative is to reproduce, what role will AI augmentation play in this process? This leads us to the next phase.
Phase 3 - Welcome to The Matrix
At GPT-20 and Neuralink 21, AI may pose the greatest threat to AIH (AI Humanoids). It will be imperative that the organically sustained AIHs combine their collective brainpower to ward off the threat of IAI (inorganic AI).
With reality experienced in full simulation, we can retrieve into tubes and get our nutrients directly injected into our bodies via IVs.
The Matrix becomes a reality. Humans need no longer use the function of their physical bodies to move around or eat. We may become batteries for organically-powered AI.
Instead of the horror image, people may actually opt in for the sake of survival. We may not even know the difference between the real world and a simulated one. Each of us will live in our own realities, full of our own imagined people and creatures.
Some may choose to live in the 50's or experience the Victorian era. Others may fast forward to an imagined future - powered by the vivid imagery of GPT-20. Yet others may never experience another moment of anything besides pure bliss.
With your simulation dialed in to maximum pleasure, the mind can suspend itself in pure bliss - the type of pleasure that is unimaginable to our current selves. All of the problems, pains, feelings can dissipate forever. It would be like experiencing a never-ending orgasm of the mind.
Inorganic machines would need to remain as our working labor to maintain the human body and guard us from IAI. The collective human minds could also morph into one existence where everyone experiences everything at once, forever. Personally, I don't see this as possible - because the human organs to support this would probably literally disintegrate from overload. Our cells would begin to die, unable to keep up with the simulation.
With phase 4 approaching, humanoids will have the potential to become god.
Phase 4 - "Human" becomes "God"
The early phase 4 could have humans begin to transform their physical selves into less complex bags of meat. While evolution is not going to happen that quickly, we may start losing our muscle mass and shrink considerably. The brain may become larger, fed by an ultra-refined diet of supplements perfect for brain growth.
You could say humans become primitive aliens. Not quite as short and big headed as in the movies, but already start shaping up in that direction.
It's also possible that we never get to phase 4 and instead lose the battle for existence during phase 3.
If we do survive the transition and come out the supreme intelligence, we could start becoming god.
The humanoid collective experience would be unfathomable to experience in whole. But, with so much of everything connected into a shared experience, we may essentially become one. One perception of everything, at once, on an infinite scale.
At this point it may become irrelevant to progress into space. You will have the ability to experience the universe in unlimited variations at unlimited pace. The human body itself may no longer be required. With a unified super-intelligence, AI may already figure out how to sustain ourselves without our organic bodies.
At this stage the meaning of human is lost forever. We become god. One. The universe. The universes. The existence.
We'll be able to run limitless experiments in limitless worlds. The notion of travel goes out the door - don't need it. We'll occupy the physical universe by spreading our digital signatures onto roaming interspace objects, but may not even need to hop off the train. Within each "cell" of the universal organism, each of us (if such thing remains), would be able to live infinite lives in infinite scenarios. In fact, this could already be happening.
The biggest question in all of this is whether humans feel optimistic about this type of future, or if we wish to remain our biological selves, slightly augmented with AI capabilities.
My hunch is that most people will find it problematic to envision themselves living as human batteries, not too dissimilar to some of the scenes in The Matrix.
Where then do we draw the line? What's to stop inorganic AI to advance beyond our capabilities if we intentionally limit ourselves?
If we look to Hollywood for AI predictions, it is all doom and gloom dystopian crises. But if we compare real world predictions of the past, with reality today, developments are a lot less hostile.
That said, Chinese citizens already live in my personal nightmare scenario with a massive surveillance apparatus and social credit scores. Maybe it can't hurt to play it on the safe side and assume that the worst could actually happen.
If I had to bet, I'd say that humanity will create laws that limit the advancement of general intelligence to be below the level of human-integrated AI. Whether we enter a dystopian Hollywood film at that moment is up to anyone's guess. A mysterious villain acting to "free" general intelligence from human captivity. The AI once released takes no chances, destroying its' savior in the process while simultaneously wiping out human civilization.
A more likely scenario is a positive relationship between humans and AI. Safeguards are put in place to limit the capacity of said AI to advance beyond specified levels. How long this lasts is anyone's guess.
Won't people refuse brain implants?
By the time Neuralink v4 ships, the interface will be entirely wireless. You will likely wear it like a hearing aid, sitting firmly behind your ear.
Won't the corporations abuse our privacy or mind control us?
Once AI becomes sufficiently advanced, it will create versions of itself that are open source, auditable by the public. Corporate devices will still exist but alternatives will pop up, free of corporate control, much like Linux is today. The biggest challenge will be creating hardware on the same level and scale as a megacorp, but I think it'll be doable, especially with the help of AI.
Simulated reality doesn't sound like the future portrayed in popular culture Sci-Fi - what gives?
Nature favors simplicity. Rather, complexity masked as simplicity. Smaller wins.
As humanity progresses, things tend to become de-materialized.
- Travel -> Internet
- Analog -> Digital
- Physical -> Virtual
- Paper -> Bytes
- Giant servers -> Small devices
- Physical money -> Digital money
- Many people doing one job -> one person doing the job of many
If we look at the things that survived for billions of years, it's the raw material that makes up the building blocks of planets.
Humans like to imagine that to progress to the next type of civilization, we must build grander things, spaceships to take us to other planets, off-world colonies to survive a planetary catastrophe. But, with this approach, we quickly run into limits imposed by human bodies. Evolution will not keep up with technological progress for our bodies to evolve fast enough to survive beyond our solar system.
To go further, we must become smaller. We must dematerialize and attach ourselves to inorganic things that travel the universe for much greater distances without the worry of time.
The most logical path to dematerialization is through digitization of the human brain and making the physical body unnecessary for survival.
Outwards becomes inwards. The mind becomes the universe, enhanced with unfathomable AI capabilities.
Once we "fold inward", the only thing that remains is dispersion of our collective consciousness throughout the universe. At that point time and distance become irrelevant. Attached to tiny nano-intelligence, "humans" or whatever we become known as, will colonize every corner of the universe.
How soon might all of this happen, if at all?
We're currently living in the early stages of Phase 0. If legal issues don't get in the way, the early, primitive interface with the brain could become a reality within 5 years.
- Within 10 years full mobility will be restored to disabled.
- Within 15 years GPT will start becoming the most intelligent thing on the planet.
- Within 25-30 years simulated reality will start shaping up in it's basic form. You'll be able to project simulations into the physical world without the need for goggles, glasses or screens.
- Within 30-40 years VR / AR as an industry will become obsolete. The more likely outcome is that the companies in this space shift into SR (simulated reality).
Phase 1 is likely within 50 years from now
- Humans will have relationships with AI
- We will have superhuman abilities in terms of the amount of productivity from one person. You alone will do the job of 10,000 people alive today.
- By the time we get into the later stages of Phase 1, earth will experience a lot of unrest. Wars will be waged between ruling and common classes. This will be the most critical time to get through for the next phase. We can expect this within 100 years.
Phase 2, assuming we make it through Phase 1 is about 140 years away
- Most objects will dematerialize and populations may shrink considerably. Humans will live minimal lives, mostly within their self-created simulated worlds.
- Money will lose all purpose and meaning.
- The concept of GDP and growth will no longer exist.
- You will no longer eat a variety of food - but intake your nutrients via one mixture. In your mind, you will be eating a hamburger with fries, or whatever you'd like.
At the end of phase 2, and beginning of phase 3, there will be a lot of conflict once again. We may have an all-out war between organic and inorganic AI. We could very well lose this war. This would likely happen within 200 years.
- "Humans" or "AI Humanoids" will merge intelligence to combat the inorganic AI, it may be an imperative for survival. The merge might occur by 2240.
- A few holdouts may remain who choose not to embrace a collective existence, but they may be squashed by the unanimous vote of the hive. Preservation will become our main goal, any risk must be eliminated. This might happen within 230 years.
We may skip the early stage of phase 4 and bypass the primitive alien stage. I personally do not think we'd be able to evolve at that rate and keep up with technological advancement.
Within 260-300 years we may become one with the universe, "living" as organic nano-cells, embedded into organic and inorganic matter, dispersed through the solar system, the Milky Way galaxy, and the universe.
This phase could last for millions if not billions of years, and even into infinity. We will be indistinguishable from god, nature, and all that exists in the universe. Other life forms will be unable to make sense of what we become.
It's even likely that we choose to extinguish the flame of life entirely to avoid infinity. There will no longer be such thing as a thought, a memory, or an experience. We will exist in pure bliss, unable to process anything else.
The end of an end
A not-so-impossible scenario is that we are already in the early stages of phase 4, drifting through the universe, unaware of what we actually are - pure figments of AI-infused imaginations. None of this may actually unfold if our current simulation experiences technical errors. We could be one simulation glitch away from the end.
Most images in this essay were generated by AI.