1 of 25

Unveiling the Mysteries of Digital Mind Storage

Aneela Kaplan & Andrea Waltlova

Slido

2 of 25

Graph of how we hope this presentation will go

Andrea

Starting from the left, we begin with the "Intro & Facts." Here, we will set the stage by providing essential background information and context.

Moving on to the "Research," we will discuss the existing technologies and research in the field.

As we transition to "What is Possible” it will all be about exploring the futuristic possibilities—imagining what could happen if we achieve digital consciousness storage. It’s a realm of boundless imagination where the potential applications and benefits seem almost limitless.

Then, we address "Ethics." This is a crucial segment where we balance our excitement with the necessary gravity. We will discuss the moral considerations, the potential for misuse, and the need for robust ethical guidelines.

Following ethics, we move into "Storage," discussing the technical challenges and possibilities of how human consciousness could be stored digitally.

Next is "Environment," where we explore how these technologies could impact and be influenced by our ecological surroundings.

Finally, we conclude with "Advantages." Here, we wrap up on a high note, highlighting the potential benefits of digitally storing human consciousness.

3 of 25

What comes to your mind when you think about Digital Mind Storage?

ⓘ

Click Present with Slido or install our Chrome extension to activate this poll while presenting.

4 of 25

Background

Digital Mind Storage first appeared in a sci-fi short story “The Tunnel under the world” published in 1955.

Since then the concept has been explored in many movies & shows

Netflix’s Black Mirror or Altered Carbon
Amazon Prime’s Upload
American movies Total Recall, Transcendence
Czech movie Restore Point
…

Andrea

The implications of successfully storing human consciousness are profound. It could revolutionize fields such as medicine, psychology, and even the way we understand life and death. Imagine a future where we could preserve our minds indefinitely, or transfer our consciousness into new bodies or digital environments. It opens up possibilities for enhanced cognitive abilities, personalized education, and unprecedented ways to preserve human knowledge and experiences.

This idea first appeared in the science fiction short story "The Tunnel under the World," published in 1955. It was a visionary piece that laid the groundwork for many of the discussions and innovations we are exploring today.

A man wakes up feeling that something is off and experiences intense deja vu throughout the day. He realizes he's reliving the same day repeatedly. Upon investigation, he discovers that his entire town has been miniaturized and placed in an underground complex. An advertising agency is manipulating their minds and using them as test subjects for marketing experiments, resetting their memories each day.

Since then, the concept has been a popular subject in various movies and TV shows, capturing the imagination of audiences worldwide. If you're here looking for inspiration for your next sci-fi binge, consider checking out some of these titles

5 of 25

Concept of digitally storing human consciousness

Sensory input

Processing

Transmission

Storage

Recall

Andrea:

To understand this concept, let’s first break down the fundamental processes that occur in our brain when we experience and store information. As depicted at the top of the slide, the journey begins with sensory input. This is everything we perceive through our senses – sight, sound, touch, taste, and smell.

Once we receive sensory input, our brain moves into the processing stage. Here, the brain interprets and makes sense of the incoming information, converting it into electrical and chemical signals through a complex network of neurons. This leads us to the transmission stage, where these signals travel across synapses, forming intricate pathways and networks that encode the information. Next comes storage, the critical phase where information is stored in various parts of the brain, allowing us to recall and use it later. Recall, as the name suggests, is the ability to access and retrieve this stored information when needed, enabling us to learn, make decisions, and interact with the world around us.

Now, imagine if we could replicate this entire process digitally. The concept of digitally storing human consciousness aims to do just that. At the core of this idea is the belief that if we can accurately capture and simulate the processes of sensory input, processing, transmission, storage, and recall, we could theoretically store and possibly even recreate a person's consciousness.

Next Aneela will tell you little bit about current state of neuroscience and brain-computer interfaces

6 of 25

Current state of neuroscience and brain-computer interfaces

7 of 25

Notable Research

Dmitry Itskov’s “2045 Initiative”
University of Central Florida in Orlando
University of Illinois in Chicago
Neuralink
Metaverse Research
University of Oxford
Carboncopies Foundation
Nectome Start-up
…

Aneela: Even though it may sound like a dystopian idea, there are a lot of research projects and start-ups working on this idea from different lenses. For this Technology to exist essential advances are needed in areas of neurobiology, neurophysiology, brain imaging, micro- and nanoscale technologies, computational neuroscience, computing hardware, and even the philosophy of mind.. All the listed projects are working on small pieces to solve the big problem

2045 initiative is a project started in 2011 to create hologram/diagram like humans by 2045 in 4 phases, the first phase was to create a robotic copy of a human body remotely capable of interpreting commands directly from the mind, and sending information back to the mind in a form that can be interpreted via brain–computer interface. It is estimated to be popularized in or before 2020. This, however, has failed to transpire.

Nectome is a start-up working on preserving Long term memories by trying to understand how they are generated by the brain in the first place. Interesting aspect they are researching into is how long term memories are generated across different neurons which make them outlast a single neuron.

Carbon copies is a foundation working on whole brain emulation in terms of preparing for the last step with focus on collaboration with earlier mentioned different areas, one of their research project is Braingenix and I linked their gitlab project later in the presentation, developers can check them out and probably contribute and learn more

8 of 25

What Neuralink does?

BIC - Brain computer interface

ultra-thin, flexible electrode threads implanted into the brain to record and stimulate neural activity

Neuralink does not enable memory storage in any way

9 of 25

Recent Achievements

2022 Nobel Prize in Physics awarded to scientists who proved that under quantum mechanics the universe we observe is “not locally real”
Nvidia Omniverse - Virtual world (Earth 2) that simulates real world physics to train robots

Aneela - There are some very recent achievements that can help propel whole Brain emulation. Have you ever heard of phrases like “Does the moon exist if there is no one to observe it?” 2022 Nobel Prize laureates proved that objects definitive properties like apple being red cannot be assumed to be true before the observation. This is a concept that cannot be fathomed by Human Brains and the advance of quantum mechanics can shed light on parts of the universe that human brains complete to make sense and can help in the research.

Another recent one is Nvidia’s June 3rd release of omniverse with which it is possible to simulate virtual world where real world physics like gravity etc exist which makes it easier for Robots to actually experience real world without being present in real world.

10 of 25

So … is it possible to store human mind?

Brain Complexity

Brain Neural Connections vs Chatgpt parameters

Size comparison with classical computing

Aneela: Before we go into if it is possible with current technology, let’s talk about how complex Human Brain is. Every Brain though intrinsically looks the same runs different algorithms not only based on physical/social environment but also the gene data. It has 86 billion neurons each with few thousands of connections which puts human brain at few 100 trillions of connections. Let’s put that in perspective, ChatGPT-4 has around 175 billion parameters, which are analogous to the weights in neural networks. That’s 1000 times less connections than human brain.

Researchers estimate that it takes close to 100 PB’s of storage space to store all the scanned data from Human Brain so, using classical computing it is going to take a lot of physical space to emulate one human brain. Although quantum computing still in earlier phases of utility, cannot be used to store huge amounts of data that is scanned from the brain. But because Quantum computing is efficient for solving probability problems so, can be used in the future for reading/interpreting the data. A hybrid approach need to be invented to efficiently emulate Human Brain. Quantum computing in addition with AI can help the cause.

11 of 25

Not yet.

Understanding different emotions by an AI is still not fully there.
Current technology cannot emulate any of the human senses.
Artificial neural networks try to replicate real neural connections but, are only the earliest phase of it.
Our brains are capable of only storing certain memories

Aneela:

Other Technical challenges include current technologies not being able to emulate any of the human senses. Them not operating as fast as human brain does A lot of human reflexes or interests are also based on DNA, we would need really complex machine learning models to correctly interpret based on different parameters. Human brains have multiple algorithms running at the same time in the background. Combining all these algorithms and using less space would be a problem, basically in openshift words, it would need millions of nodes running trillions of operations per second for a single brain and there needs to be high speed internet connection 24X7 everywhere which doesn’t exist yet. Brains are also capable of self-regulating what data is stored and what is removed. An algorithm that works for one person wouldn’t work for another.

From my research, The technology is pretty advanced in individual areas and the challenge comes with being able to combine all the areas and making it work seamlessly like our brains. Scientists are able to Scan Mouse’s brain for whole emulation but are not able to actually understand what some of the signals mean. MRI scans are able to scan which parts of the brain reacts to what environment they are in. Neuralink and some other researchers are able to stimulate brain to resolve neurological issues or replicate neurons that are severely damaged.

12 of 25

✅ Is possible

❌ Is NOT possible

Neural Data Recording
Brain Mapping
Basic Brain-Computer Interfaces
Deep Brain Stimulation and Modulation of neural activity

Understanding and digitizing the nuanced and personal nature of consciousness
Real-time monitoring of all neural activity at the necessary resolution

13 of 25

But let’s imagine we have the technology …

Are we ready in other aspects ?

14 of 25

Ethics

Self-Perception
Data Security/Informed Consent
Agency and Manipulation
Ethical Use and Abuse
Societal Impact
Impact on Relationships
Regulatory and Legal Issues

Aneela: There are so many ethical questions we don’t have answers to yet, some of them are the emulated being if gains consciousness, will they be a different person from then on and if so, did they give consent and did the person who gave consent to have their brain emulated give informed consent to using their brain knowing the full consequences like in the episode Joan is Awful of Netflix’s Black Mirror. Would they be able to choose to be deleted. With saving every data and unless the original copy which is the person’s brain is kept untouched by brain simulators, there are chances of manipulating and giving people different personality all together. What rights would the digital minds have? This can also exacerbate the existing social inequalities.

Is a digital copy of a mind the same person as the original?

What measures can ensure equitable access to these technologies?

How can international cooperation be achieved in regulating this technology?

15 of 25

Storage space

A range of histological features in 1 mm3 of human brain were rendered, including neuropil (A) and its segmentation (B) at nanometer resolution, annotated synapses (C), excitatory neurons (D), inhibitory neurons (E), astrocytes (F), oligodendrocytes (G), myelin (H), and blood vessels (I). A previously unrecognized neuronal class (J) and multisynaptic connections (K) were also identified.

https://www.science.org/doi/10.1126/science.adk4858

Interested? See their website

Andrea:

Alright, let’s say you somehow ethically obtain brain tissue – and that is a big 'if'. Now, let’s assume you are able to scan a brain. On the slide, you can see a diagram from a recent study published in May by Harvard researchers in collaboration with Google Research. They used electron microscopy to map 1 cubic millimeter of brain tissue, and yes, they obtained it ethically during surgery on a patient with epilepsy.

The map is in high resolution, revealing many fascinating details. Since most of us are not neuroscientists, we won’t dive deeply into the specifics. However, the interesting fact for our presentation – and the one we’ll have some fun with – is that it took 1.4 petabytes of data to store this map.

Before we explore some fun comparisons, it’s important to understand that the size really depends on the scanning resolution and the storage method. They stored the data as images from an electron microscope, which may not be the most efficient way. So, the size mentioned in the article is somewhat clickbait. You could compare it to saving a bitmap of a black square versus a vector image; the vector image, of course, would be much smaller.

16 of 25

How much space could we actually need?

10 million people = population of Czechia = 14 Zettabytes (14 billions TB)

1.4 PB per 1mm³ of human brain

You would need 700 of 2TB USB sticks to store 1.4PB

Or 14 pieces of 100TB SSD disk from Nimbus Data

7 million hours of HD video

Andrea:

To put this into perspective, 1.4 petabytes is the amount of data generated by mapping just 1 cubic millimeter of human brain tissue. That’s an enormous amount of data for such a tiny volume. But what does 1.4 petabytes really mean in more relatable terms?

First, if you were to use 2TB USB sticks to store 1.4PB, you would need 700 of them. Alternatively, you could use 14 of the 100TB SSD disks available from Nimbus Data.

Another way to think about this is in terms of video storage. 1.4 petabytes can hold approximately 7 million hours of HD video. That’s enough to watch cat videos continuously for 800 years without ever running out of content!

In terms of people, if we take population of Czechia which is around 10 million people, we would need 14 Zettabytes to store 1 cubic millimeter of every person in Czechia. It’s bigger than global internet traffic in 2022 which was ~4 Zettabytes (based on Cisco reports).

For even more context, consider that the Large Hadron Collider (LHC) generates about 1 petabyte of collision data per second during its experiments.

If we were to store the digital minds (mind that we are counting with the inefficient image storage!) of the entire world population of 8 billion people at the same data density, we would need approximately 11 yottabytes, which is around 11 trillion terabytes.

These comparisons help illustrate the massive scale of data storage required for brain mapping and digital mind storage, showing both the incredible detail captured and the significant challenges we face in managing and utilizing such vast amounts of data.

17 of 25

How about some future pricing model?

🥉Free Tier: "Basic Brain Bits"

Memory Storage: Up to 100 memories

Limited to minor events, random facts or songs from 90s
Just the most embarrassing memories allowed

Emotion Storage: Basic emotions

Happy, sad, neutral – no complex feelings allowed.

Personality Backup: Basic traits

Just the essentials: introvert/extrovert, favorite color, etc.

Access: Read-only

No editing or re-experiencing stored memories.

🥈Silver Tier: "Nostalgia Pack"

Memory Storage: 10,000 memories

Includes childhood memories and significant life events.

Emotion Storage: Full spectrum emotions

Fear, joy, anger, surprise, and more.

Personality Backup: Comprehensive traits

Includes hobbies, preferences, and quirks.

Access: Read and Write

Edit and annotate memories.

🥇Gold Tier: "Life Story Vault"

Memory Storage: 1,000,000 memories

Entire life history, including mundane daily routines.

Emotion Storage: Enhanced emotions

Nuanced feelings like nostalgia, euphoria, and melancholia.

Personality Backup: Detailed personality map

Includes detailed habit patterns and decision-making processes.

Access: Read, Write, and Share

Share memories with friends and family (with permission).

Additional Features

Emotion tagging for memories.
Memory playback with emotion simulation.

Monthly Cost: Free

Monthly Cost: $?

Amazon S3 Glacier Flexible Retrieval is $0.0036/GB per month

18 of 25

Environmental Concerns

Energy Consumption
Cooling and Heat Management
Electronic Waste
Life Cycle Emissions
Raw Materials
Land use and Habitat Disruption
Heat Generation and Waste Heat Utilization
Local Ecosystems Disruption
Over-population

Chat-gpt 3.5

285000 CPUs

10000 GPUs

400 GBs/s connectivity

2% of global energy usage

4% of global greenhouse emissions

1 Brain

~

=

Aneela - With the amount of storage space used and gpu’s or Mind Processing Unit’s(mpu’s) that may evolve in the future, these are the environmental concerns that exists and must be addressed while working on WBE process, AI once was only available to researchers and today with chat-gpt, millions of operations are being performed every second. Just the supercomputer that Chat-gpt 3.5 is trained on is equal to 100,700 MacBooks + all the millions of users using their own devices.

26 football field size can store data of 1000’s PBs range and to save a single whole brain it would need a data center the size of 2600 football fields

The largest one right now that manages all Big4 companies traffic and data is 100 times smaller than what we would need for 1 brain and it would use 4% of energy that US consumes and will generate 6% of world’s Green-house Emissions today everyday

19 of 25

“Amazon, Microsoft, Meta and Google are the four largest purchasers of corporate renewable energy power purchase agreements, having contracted almost 50 GW to date, equal to the generation capacity of Sweden.”

(International Energy Agency - source)

20 of 25

What would you do if you can create a digital clone of your brain?

ⓘ

Click Present with Slido or install our Chrome extension to activate this poll while presenting.

21 of 25

Potential Advantages

Medical and Psychological Benefits
Enhanced Learning and Creativity
Cultural Preservation
Technological Advancements
Create Simulations
Understanding Consciousness

Aneela: We can see the most benefits in the Medical field, imagine being able to create whole human that replicates you and doctors can perform surgical simulations or lifestyle simulations to understand and solve disorders quickly. Digital backups of mind could be used to restore lost cognitive functions due traumatic brain injuries or alzheimer’s. Researchers can collaborate and work on solving problems beyond life. We may get closer to understanding the universe and beyond when we try to emulate human brain and all other patterns in the world. Even without humans learning and passing down skills we will be able to preserve and evolve cultural knowledge and scientific significances without knowledge getting lost in the translation and then becoming superstitions. Can create simulations and understand the pitfalls. Computers used to be retrieval based but, with AI they are turned into generative based, answers are generated for questions even without it being pre-loaded but it is still you ask a question and AI answers. With Brain Emulation it does stuff without being asked to, like it’s their job is to do stuff.

We will be able to get real close to understanding consciousness or the universe.

22 of 25

Let’s say we are able to emulate human brain, will it actually be conscious and if it were to be conscious will it actually be self-aware?

23 of 25

As a developer where can you get involved

h ttps://braingenix.org/Challenge/Overview/
https://gitlab.braingenix.org/carboncopies
https://sites.google.com/carboncopies.org/ethics/register-an-ethical-concern
Thinking about donating your brain? Take a look here

24 of 25

Audience Q&A Session

ⓘ

Click Present with Slido or install our Chrome extension to show live Q&A while presenting.

25 of 25

Thank you for your attention

Resources