AI Predicts Human Evolution

What will humans look like in the future? Given trends in biological, cultural, and technological evolution, how will humans appear in 10, 100, or 1000 years? I was curious what AI would predict. To figure that out, I used a combination of ChatGPT and Midjourney. ChatGPT takes text prompts and provides text responses. Midjourney takes text prompts and provides image responses. I chose to use Midjourney for images because they are generally higher quality than images from ChatGPT. Here are the text prompts that I used for ChatCPT: Given trends in biological, cultural, and technological evolution, predict what humans will look like in 10, 100, and 1000 years. Convert your predictions for human appearance into three Midjourney prompts. I used the results of #2 as prompts for Midjourney. Midjourney responds with four images per prompt. So, for each prompt, I chose to upscale the first of the four responses. The first wasn’t always my favorite, but this rule decreased my subjective influence on the process. Below are the results. Humans in 10 Years Here is the ChatGPT prediction: “In the short term, changes will be predominantly cultural and technological rather than biological. Technological enhancements such as wearable tech, advanced prosthetics, and increasingly sophisticated interfaces for communication will become more integrated into daily life. Genomic medicine might become more mainstream, allowing for personalized medical treatments that are more effective and less invasive. Biologically, the human species will remain largely the same, but we might begin to see the wider adoption of CRISPR and other gene-editing technologies to combat genetic diseases.” What do you think of this? I think ChatGPT’s prediction reasonable. Here is the prompt that ChatGPT generated for Midjourney: “A portrait of a modern human in 2034, wearing sophisticated wearable technology, such as smart glasses and health monitoring devices integrated into everyday clothing. The setting is an urban environment, bustling with interactive digital displays and green, eco-friendly buildings. The human is casually interacting with their environment using subtle gestures, displaying a seamless blend of technology and daily life.” Here is the Midjourney prediction: What do you think of this? I feel like Midjourney’s prediction is a bit much on the Matrix esthetic, although that may have resulted from an effort at depicting environmentalism. Humans in 100 Years Here is the ChatGPT prediction: “Over the next century, we might begin to see more significant integration of technology and biology. This could include biotechnological enhancements that improve physical and cognitive abilities, such as neural implants enhancing memory or learning speed. Culturally, society may become more accepting of these enhancements, possibly leading to a divide between those who choose to adopt such technologies and those who do not. Genetically, we may see more active management of human genetics to prevent diseases, enhance desirable traits, and possibly adapt to changing environmental conditions.” What do you think of this? I think ChatGPT’s prediction is probably too conservative. If humans are still around in 100 years, ChatGPT’s prediction for 1000 years may be more applicable. See below. Here is the prompt that ChatGPT generated for Midjourney: “A futuristic human from 2124, featuring biotechnological enhancements. This human has neural implants visible on the scalp and eyes glowing with augmented reality capabilities. They wear a sleek, form-fitting suit that adapts to environmental conditions and health needs, standing in a high-tech room with holographic displays and advanced AI interfaces. The overall aesthetic combines organic and technological elements, showcasing the integration of biology and technology.” Here is the Midjourney prediction: What do you think of this? Midjourney’s prediction seems dystopian. Who actually wants to look like that? Most people probably will prefer esthetics that aren’t sci-fi cliches. Humans in 1000 Years Here is the ChatGPT prediction: “Looking a millennium ahead, the possibilities become even more speculative. If humanity has not encountered insurmountable obstacles like severe climate change or catastrophic events, we might see a species that has heavily integrated technology with biology, possibly leading to what some might consider a new phase of evolution: transhumanism. This could mean anything from bodies that are highly adaptable to different environments (to accommodate interplanetary travel and colonization) to forms of consciousness that are partially or wholly uploaded onto computational substrates. Culturally and socially, the definitions of ‘human’ might expand to include a wide variety of bioengineered and technologically augmented beings. Genetic and cybernetic enhancements could lead to what might essentially be new human subspecies.” What do you think of this? I think ChatGPT’s prediction is much too conservative. As I mentioned above, this seems more like a prediction for humans in 100 years. Other than that, it seems reasonable – if anything can seem reasonable, given how little we can actually know about the consequences of imminent technological evolution. Here is the prompt that ChatGPT generated for Midjourney: “A visionary depiction of a human in the year 3024, heavily modified with advanced cybernetic and genetic enhancements. This being features skin that can adapt its pigmentation based on sunlight exposure, eyes capable of seeing multiple spectra, and limbs that can morph to perform different functions. The setting is a space colony with views of other planets, reflecting a society where biological and technological integration has created a diverse new form of post-human life.” Here is the Midjourney prediction: What do you think of this? I think Midjourney’s prediction is terrible. It’s even worse than the 100 year prediction. But, to be charitable, imagine the possibility that Midjourney’s prediction was just how future humans might look from our current perspective. Future Humans from Their Perspective What might future humans, 100 to 1000 years from now, look like from their own perspective? Maybe ChatGPT could generate a better prompt for Midjourney. I tried to do this in two parts. First, I asked ChatGPT: How will human appearance to each other be different than the way they might appear to us, if we could see them? Here’s the response: “In 1000 years, if we could observe humans from our current perspective, their appearance might be startlingly alien compared to what we expect today. The key differences in how they would appear to each other versus how they might appear to us could include: “Normalization of Enhancements: Future humans would likely be accustomed to a wide range of physical enhancements and modifications, seeing them as normal or even desirable. To us, these enhancements might appear extreme or unsettling, such as limbs designed for specific functions (e.g., different gravitational environments), eyes engineered for multiple spectrums, or skin with dynamic textures and colors. “Technological Integration: While future humans might view integrated technology as an intrinsic part of their identity and functionality, we might see it as a blurring of the lines between human and machine. This could include visible cybernetic enhancements, biotechnological implants, or even less distinctly human forms where technology and biology are indistinguishable. “Adaptive Physiology: Features that allow adaptation to various environments (e.g., thicker skin for radiation protection, gill-like structures for living underwater) might seem bizarre or even grotesque to us but would be perfectly rational and ordinary to future humans living in diverse habitats. “Sensory Perceptions: Future humans might communicate and interact through means imperceptible to us, such as advanced telepathy enabled by neural implants or pheromonal communication augmented by biological enhancements. What might seem like a silent, still gathering to us could be rich with communication and interaction to them. Digital and Virtual Manifestations: Some individuals might exist primarily in digital forms, with their physical manifestations being secondary or tailored for specific interactions. To us, these forms might appear transient or insubstantial, but to their peers, they could be as real and meaningful as physical presence. “Cultural and Aesthetic Differences: The aesthetic values that guide physical appearance in the future could be vastly different from ours, influenced by new norms, environmental needs, and the blending of diverse human and post-human cultures. What we might find aesthetically confusing or unappealing could be the pinnacle of beauty and functionality to them. “Overall, the future appearance of humans could challenge our current notions of identity, form, and aesthetics, presenting a rich tapestry of what it means to be human in a technologically and biologically advanced future.” This is a reasonably thought-provoking response, even if the timeline is too conservative, and despite our inability to conceive of superintelligent futures. So next, I asked ChatGPT: Generate a Midjourney prompt for an image that depicts how humans in 1000 years might see each other — rather than how we might see them. Here’s the response: “An imaginative portrayal of humans in the year 3024, interacting in a highly advanced, multi-planetary society. This scene includes individuals with a diverse range of enhancements and modifications: one with shimmering, adaptive skin that changes color and texture based on environmental signals; another with elongated, versatile limbs suited for low-gravity environments; and a third person appearing via a vivid, interactive hologram. The setting is a communal space in a space colony with architecture that integrates organic and synthetic materials seamlessly. Each figure communicates through subtle gestures that initiate complex augmented reality displays, visible only to them.” Here is the Midjourney prediction: I didn’t think it could get yet worse. I was wrong. It looks more like a retro sci-fi convention than a prediction of superhuman inter-personal perception. Meh. The Value of AI Predictions There’s at least one important reason why contemporary AI fails to provide effective visual predictions of future humans. AI is currently superintelligent in a broad sense, but not yet in a deep sense. It can more or less do only what humans can do, abstracting across everything it has learned from humans. It can do that at scale, which is superhuman compared to any one human, but still all-too-human when compared to humanity as a whole. So we shouldn’t be surprised that AI predictions, especially visual predictions, currently tend toward biases and cliches that are common among humanity as a whole. This is just how current AI works. That’s not great for prediction. But it’s still useful for reflection. Current AI “predictions” give us insights into the human imagination. If we look at AI predictions and see superficiality, that’s because AI has been trained on superficiality. If we want to see something less superficial, we have to offer that. But, like a good mirror, at least AI makes us more keenly aware of the weaknesses in our superficiality. My Prediction Here’s my prediction. Humanity within 100 years will approximate and surpass that which we now imagine Gods to be. This will apply to every level of our being: physical and cognitive. Esthetics will still matter, even more than they matter now. And the best technologies will be, as they are now, the most transparent technologies. Humanity within 100 years will be, if we could see and compare them to ourselves, pure esthetics. Function will be substantially pervasive and yet utterly invisible, except to the extent that it serves esthetics. Superhumanity will be art.

What will humans look like in the future? Given trends in biological, cultural, and technological evolution, how will humans appear in 10, 100, or 1000 years? I was curious what AI would predict.

To figure that out, I used a combination of ChatGPT and Midjourney. ChatGPT takes text prompts and provides text responses. Midjourney takes text prompts and provides image responses. I chose to use Midjourney for images because they are generally higher quality than images from ChatGPT.

Here are the text prompts that I used for ChatCPT:

Given trends in biological, cultural, and technological evolution, predict what humans will look like in 10, 100, and 1000 years.

Convert your predictions for human appearance into three Midjourney prompts.

I used the results of #2 as prompts for Midjourney. Midjourney responds with four images per prompt. So, for each prompt, I chose to upscale the first of the four responses. The first wasn’t always my favorite, but this rule decreased my subjective influence on the process.

Below are the results.

Humans in 10 Years

Here is the ChatGPT prediction:

“In the short term, changes will be predominantly cultural and technological rather than biological. Technological enhancements such as wearable tech, advanced prosthetics, and increasingly sophisticated interfaces for communication will become more integrated into daily life. Genomic medicine might become more mainstream, allowing for personalized medical treatments that are more effective and less invasive. Biologically, the human species will remain largely the same, but we might begin to see the wider adoption of CRISPR and other gene-editing technologies to combat genetic diseases.”

What do you think of this? I think ChatGPT’s prediction reasonable.

Here is the prompt that ChatGPT generated for Midjourney:

“A portrait of a modern human in 2034, wearing sophisticated wearable technology, such as smart glasses and health monitoring devices integrated into everyday clothing. The setting is an urban environment, bustling with interactive digital displays and green, eco-friendly buildings. The human is casually interacting with their environment using subtle gestures, displaying a seamless blend of technology and daily life.”

Here is the Midjourney prediction:

What do you think of this? I feel like Midjourney’s prediction is a bit much on the Matrix esthetic, although that may have resulted from an effort at depicting environmentalism.

Humans in 100 Years

Here is the ChatGPT prediction:

“Over the next century, we might begin to see more significant integration of technology and biology. This could include biotechnological enhancements that improve physical and cognitive abilities, such as neural implants enhancing memory or learning speed. Culturally, society may become more accepting of these enhancements, possibly leading to a divide between those who choose to adopt such technologies and those who do not. Genetically, we may see more active management of human genetics to prevent diseases, enhance desirable traits, and possibly adapt to changing environmental conditions.”

What do you think of this? I think ChatGPT’s prediction is probably too conservative. If humans are still around in 100 years, ChatGPT’s prediction for 1000 years may be more applicable. See below.

Here is the prompt that ChatGPT generated for Midjourney:

“A futuristic human from 2124, featuring biotechnological enhancements. This human has neural implants visible on the scalp and eyes glowing with augmented reality capabilities. They wear a sleek, form-fitting suit that adapts to environmental conditions and health needs, standing in a high-tech room with holographic displays and advanced AI interfaces. The overall aesthetic combines organic and technological elements, showcasing the integration of biology and technology.”

Here is the Midjourney prediction:

What do you think of this? Midjourney’s prediction seems dystopian. Who actually wants to look like that? Most people probably will prefer esthetics that aren’t sci-fi cliches.

Humans in 1000 Years

Here is the ChatGPT prediction:

“Looking a millennium ahead, the possibilities become even more speculative. If humanity has not encountered insurmountable obstacles like severe climate change or catastrophic events, we might see a species that has heavily integrated technology with biology, possibly leading to what some might consider a new phase of evolution: transhumanism. This could mean anything from bodies that are highly adaptable to different environments (to accommodate interplanetary travel and colonization) to forms of consciousness that are partially or wholly uploaded onto computational substrates. Culturally and socially, the definitions of ‘human’ might expand to include a wide variety of bioengineered and technologically augmented beings. Genetic and cybernetic enhancements could lead to what might essentially be new human subspecies.”

What do you think of this? I think ChatGPT’s prediction is much too conservative. As I mentioned above, this seems more like a prediction for humans in 100 years. Other than that, it seems reasonable – if anything can seem reasonable, given how little we can actually know about the consequences of imminent technological evolution.

Here is the prompt that ChatGPT generated for Midjourney:

“A visionary depiction of a human in the year 3024, heavily modified with advanced cybernetic and genetic enhancements. This being features skin that can adapt its pigmentation based on sunlight exposure, eyes capable of seeing multiple spectra, and limbs that can morph to perform different functions. The setting is a space colony with views of other planets, reflecting a society where biological and technological integration has created a diverse new form of post-human life.”

Here is the Midjourney prediction:

What do you think of this? I think Midjourney’s prediction is terrible. It’s even worse than the 100 year prediction. But, to be charitable, imagine the possibility that Midjourney’s prediction was just how future humans might look from our current perspective.

Future Humans from Their Perspective

What might future humans, 100 to 1000 years from now, look like from their own perspective?

Maybe ChatGPT could generate a better prompt for Midjourney. I tried to do this in two parts. First, I asked ChatGPT: How will human appearance to each other be different than the way they might appear to us, if we could see them? Here’s the response:

“In 1000 years, if we could observe humans from our current perspective, their appearance might be startlingly alien compared to what we expect today. The key differences in how they would appear to each other versus how they might appear to us could include:

“Normalization of Enhancements: Future humans would likely be accustomed to a wide range of physical enhancements and modifications, seeing them as normal or even desirable. To us, these enhancements might appear extreme or unsettling, such as limbs designed for specific functions (e.g., different gravitational environments), eyes engineered for multiple spectrums, or skin with dynamic textures and colors.

“Technological Integration: While future humans might view integrated technology as an intrinsic part of their identity and functionality, we might see it as a blurring of the lines between human and machine. This could include visible cybernetic enhancements, biotechnological implants, or even less distinctly human forms where technology and biology are indistinguishable.

“Adaptive Physiology: Features that allow adaptation to various environments (e.g., thicker skin for radiation protection, gill-like structures for living underwater) might seem bizarre or even grotesque to us but would be perfectly rational and ordinary to future humans living in diverse habitats.

“Sensory Perceptions: Future humans might communicate and interact through means imperceptible to us, such as advanced telepathy enabled by neural implants or pheromonal communication augmented by biological enhancements. What might seem like a silent, still gathering to us could be rich with communication and interaction to them.

Digital and Virtual Manifestations: Some individuals might exist primarily in digital forms, with their physical manifestations being secondary or tailored for specific interactions. To us, these forms might appear transient or insubstantial, but to their peers, they could be as real and meaningful as physical presence.

“Cultural and Aesthetic Differences: The aesthetic values that guide physical appearance in the future could be vastly different from ours, influenced by new norms, environmental needs, and the blending of diverse human and post-human cultures. What we might find aesthetically confusing or unappealing could be the pinnacle of beauty and functionality to them.

“Overall, the future appearance of humans could challenge our current notions of identity, form, and aesthetics, presenting a rich tapestry of what it means to be human in a technologically and biologically advanced future.”

This is a reasonably thought-provoking response, even if the timeline is too conservative, and despite our inability to conceive of superintelligent futures. So next, I asked ChatGPT: Generate a Midjourney prompt for an image that depicts how humans in 1000 years might see each other — rather than how we might see them. Here’s the response:

“An imaginative portrayal of humans in the year 3024, interacting in a highly advanced, multi-planetary society. This scene includes individuals with a diverse range of enhancements and modifications: one with shimmering, adaptive skin that changes color and texture based on environmental signals; another with elongated, versatile limbs suited for low-gravity environments; and a third person appearing via a vivid, interactive hologram. The setting is a communal space in a space colony with architecture that integrates organic and synthetic materials seamlessly. Each figure communicates through subtle gestures that initiate complex augmented reality displays, visible only to them.”

Here is the Midjourney prediction:

I didn’t think it could get yet worse. I was wrong. It looks more like a retro sci-fi convention than a prediction of superhuman inter-personal perception. Meh.

The Value of AI Predictions

There’s at least one important reason why contemporary AI fails to provide effective visual predictions of future humans. AI is currently superintelligent in a broad sense, but not yet in a deep sense. It can more or less do only what humans can do, abstracting across everything it has learned from humans. It can do that at scale, which is superhuman compared to any one human, but still all-too-human when compared to humanity as a whole.

So we shouldn’t be surprised that AI predictions, especially visual predictions, currently tend toward biases and cliches that are common among humanity as a whole. This is just how current AI works. That’s not great for prediction. But it’s still useful for reflection.

Current AI “predictions” give us insights into the human imagination. If we look at AI predictions and see superficiality, that’s because AI has been trained on superficiality. If we want to see something less superficial, we have to offer that. But, like a good mirror, at least AI makes us more keenly aware of the weaknesses in our superficiality.

My Prediction

Here’s my prediction.

Humanity within 100 years will approximate and surpass that which we now imagine Gods to be. This will apply to every level of our being: physical and cognitive. Esthetics will still matter, even more than they matter now. And the best technologies will be, as they are now, the most transparent technologies.

Humanity within 100 years will be, if we could see and compare them to ourselves, pure esthetics. Function will be substantially pervasive and yet utterly invisible, except to the extent that it serves esthetics. Superhumanity will be art.